Interleaving Execution and Planning for Nondeterministic, Partially Observable Domains
Info-Note-5-Doc9803alltext航线安全审计.en
Approved by the Secretary General and published under his authorityLine OperationsSafety Audit (LOSA)First Edition — 2002Doc 9803AN/761AMENDMENTSThe issue of amendments is announced regularly in the ICAO Journal and in the monthly Supplement to the Catalogue of ICAO Publications and Audio-visual Training Aids, which holders of this publication should consult. The space below is provided to keep a record of such amendments.RECORD OF AMENDMENTS AND CORRIGENDA AMENDMENTS CORRIGENDANo.DateapplicableDateenteredEnteredby No.Dateof issueDateenteredEnteredbyTABLE OF CONTENTSPage PageForeword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .(v) Acronyms and Abbreviations . . . . . . . . . . . . . . . . .(vi) Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .(vii) Chapter 1.Basic error management concepts. .1-11.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . .1-11.2Background . . . . . . . . . . . . . . . . . . . . . . . . .1-2Reactive strategies. . . . . . . . . . . . . . . . . .1-2Combined reactive/proactive strategies. .1-2Proactive strategies . . . . . . . . . . . . . . . . .1-41.3 A contemporary approach to operationalhuman performance and error. . . . . . . . . . .1-51.4The role of the organizational culture . . . .1-71.5Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . .1-7 Chapter2.Implementing LOSA . . . . . . . . . . . . .2-12.1History of LOSA. . . . . . . . . . . . . . . . . . . . .2-12.2The Threat and Error Management Model.2-1Threats and errors defined. . . . . . . . . . . .2-1Definitions of crew error response . . . . .2-4Definitions of error outcomes. . . . . . . . .2-4Undesired Aircraft States . . . . . . . . . . . .2-42.3LOSA operating characteristics . . . . . . . . .2-5Observer assignment . . . . . . . . . . . . . . . .2-7Flight crew participation. . . . . . . . . . . . .2-72.4How to determine the scope of a LOSA . .2-72.5Once the data is collected. . . . . . . . . . . . . .2-82.6Writing the report . . . . . . . . . . . . . . . . . . . .2-82.7Success factors for LOSA. . . . . . . . . . . . . .2-8Chapter3.LOSA and the safety changeprocess (SCP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-13.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . .3-13.2 A constantly changing scene. . . . . . . . . . . .3-13.3One operator’s example of an SCP . . . . . .3-2 Chapter4.How to set up a LOSA —US Airways experience . . . . . . . . . . . . . . . . . . . . . .4-14.1Gathering information. . . . . . . . . . . . . . . . .4-14.2Interdepartmental support . . . . . . . . . . . . . .4-14.3LOSA steering committee. . . . . . . . . . . . . .4-1Safety department . . . . . . . . . . . . . . . . . .4-1Flight operations and trainingdepartments . . . . . . . . . . . . . . . . . . . . . . .4-2Pilots union . . . . . . . . . . . . . . . . . . . . . . .4-24.4The key steps of a LOSA. . . . . . . . . . . . . .4-2Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2Action plan . . . . . . . . . . . . . . . . . . . . . . .4-24.5The keys to an effective LOSA . . . . . . . . .4-4Confidentiality and no-jeopardy. . . . . . .4-4The role of the observer . . . . . . . . . . . . .4-54.6Promoting LOSA for flight crews . . . . . . .4-5 Appendix A — Examples of the various forms utilized by LOSA . . . . . . . . . . . . . . . . . . . . . . . . . . .A-1 Appendix B — Example of an introductory letterby an airline to its flight crews. . . . . . . . . . . . . . . .B-1 Appendix C — List of recommended readingand reference material. . . . . . . . . . . . . . . . . . . . . . .C-1FOREWORDThe safety of civil aviation is the major objective of the International Civil Aviation Organization (ICAO). Consider-able progress has been made in increasing safety, but additional improvements are needed and can be achieved. It has long been known that the majority of aviation accidents and incidents result from less than optimum human per-formance, indicating that any advance in this field can be expected to have a significant impact on the improvement of aviation safety.This was recognized by the ICAO Assembly, which in 1986 adopted Resolution A26-9 on Flight Safety and Human Factors. As a follow-up to the Assembly Resolution, the Air Navigation Commission formulated the following objective for the task:“To improve safety in aviation by making States more aware and responsive to the importance of Human Factors in civil aviation operations through the provision of practical Human Factors materials and measures, developed on the basis of experience in States, and by developing and recommending appropriate amendments to existing material in Annexes and other documents with regard to the role of Human Factors in the present and future operational environments. Special emphasis will be directed to the Human Factors issues that may influence the design, transition and in-service use of the future ICAO CNS/A TM systems.”One of the methods chosen to implement Assembly Resolution A26-9 is the publication of guidance materials, including manuals and a series of digests, that address various aspects of Human Factors and its impact on aviation safety. These documents are intended primarily for use by States to increase the awareness of their personnel of the influence of human performance on safety.The target audience of Human Factors manuals and digests are the managers of both civil aviation administrations and the airline industry, including airline safety, training and operational managers. The target audience also includes regulatory bodies, safety and investigation agencies and training establishments, as well as senior and middle non-operational airline management.This manual is an introduction to the latest information available to the international civil aviation community on the control of human error and the development of counter-measures to error in operational environments. Its target audience includes senior safety, training and operational personnel in industry and regulatory bodies.This manual is intended as a living document and will be kept up to date by periodic amendments. Subsequent editions will be published as new research results in increased knowledge on Human Factors strategies and more experience is gained regarding the control and management of human error in operational environments.ACRONYMS AND ABBREVIATIONS ADS Automatic Dependent SurveillanceA TC Air Traffic ControlCFIT Controlled Flight Into TerrainCNS/A TM Communications, Navigation and Surveillance/Air Traffic Management CPDLC Controller-Pilot Data Link CommunicationsCRM Crew Resource ManagementDFDR Digital Flight Data RecorderETOPS Extended Range Operations by Twin-engined AeroplanesFAA Federal Aviation AdministrationFDA Flight Data AnalysisFMS Flight Management SystemFOQA Flight Operations Quality AssuranceICAO International Civil Aviation OrganizationLOSA Line Operations Safety AuditMCP Mode Control PanelQAR Quick Access RecorderRTO Rejected Take-OffSCP Safety Change ProcessSOPs Standard Operating ProceduresTEM Threat and Error ManagementUTTEM University of Texas Threat and Error ManagementINTRODUCTION1.This manual describes a programme for the management of human error in aviation operations known as Line Operations Safety Audit (LOSA). LOSA is proposed as a critical organizational strategy aimed at developing countermeasures to operational errors. It is an organizational tool used to identify threats to aviation safety, minimize the risks such threats may generate and implement measures to manage human error in operational contexts. LOSA enables operators to assess their level of resilience to systemic threats, operational risks and front-line personnel errors, thus providing a principled, data-driven approach to prioritize and implement actions to enhance safety.2.LOSA uses expert and highly trained observers to collect data about flight crew behaviour and situational factors on “normal” flights. The audits are conducted under strict no-jeopardy conditions; therefore, flight crews are not held accountable for their actions and errors that are observed. During flights that are being audited, observers record and code potential threats to safety; how the threats are addressed; the errors such threats generate; how flight crews manage these errors; and specific behaviours that have been known to be associated with accidents and incidents.3.LOSA is closely linked with Crew Resource Management (CRM) training. Since CRM is essentially error management training for operational personnel, data from LOSA form the basis for contemporary CRM training refocus and/or design known as Threat and Error Man-agement (TEM) training. Data from LOSA also provide a real-time picture of system operations that can guide organizational strategies in regard to safety, training and operations. A particular strength of LOSA is that it identifies examples of superior performance that can be reinforced and used as models for training. In this way, training inter-ventions can be reshaped and reinforced based on successful performance, that is to say, positive feedback. This is indeed a first in aviation, since the industry has traditionally collected information on failed human performance, such as in accidents and incidents. Data collected through LOSA are proactive and can be immediately used to prevent adverse events.4.LOSA is a mature concept, yet a young one. LOSA was first operationally deployed following the First LOSA Week, which was hosted by Cathay Pacific Airways in Cathay City, Hong Kong, from 12 to 14 March 2001. Although initially developed for the flight deck sector, there is no reason why the methodology could not be applied to other aviation operational sectors, including air traffic control, maintenance, cabin crew and dispatch.5.The initial research and project definition was a joint endeavour between The University of Texas at Austin Human Factors Research Project and Continental Airlines, with funding provided by the Federal Aviation Admin-istration (FAA). In 1999, ICAO endorsed LOSA as the primary tool to develop countermeasures to human error in aviation operations, developed an operational partnership with The University of Texas at Austin and Continental Airlines, and made LOSA the central focus of its Flight Safety and Human Factors Programme for the period 2000 to 2004.6.As of February 2002, the LOSA archives contained observations from over 2 000 flights. These observations were conducted within the United States and internationally and involved four United States and four non-United States operators. The number of operators joining LOSA has constantly increased since March 2001 and includes major international operators from different parts of the world and diverse cultures.7.ICAO acts as an enabling partner in the LOSA programme. ICAO’s role includes promoting the importance of LOSA to the international civil aviation community; facilitating research in order to collect necessary data; acting as a cultural mediator in the unavoidably sensitive aspects of data collection; and contributing multicultural obser-vations to the LOSA archives. In line with these objectives, the publication of this manual is a first step at providing information and, therefore, at increasing awareness within the international civil aviation community about LOSA.8.This manual is an introduction to the concept, methodology and tools of LOSA and to the potential remedial actions to be undertaken based on the data collected under LOSA. A very important caveat must be introduced at this point: this manual is not intended to convert readers into instant expert observers and/or LOSA auditors. In fact, it is strongly recommended that LOSA not be attempted without a formal introduction to it for the(viii)Line Operations Safety Audit (LOSA)following reasons. First, the forms presented in Appendix A are for illustration purposes exclusively, since they are periodically amended on the basis of experience gained and feedback obtained from continuing audits. Second, formal training in the methodology, in the use of LOSA tools and, most important, in the handling of the highly sensitive data collected by the audits is absolutely essential. Third, the proper structuring of the data obtained from the audits is of paramount importance.9.Therefore, until extensive airline experience is accumulated, it is highly desirable that LOSA training be coordinated through ICAO or the founding partners of the LOSA project. As the methodology evolves and reaches full maturity and broader industry partnerships are developed, LOSA will be available without restrictions to the international civil aviation community.10.This manual is designed as follows:•Chapter 1 includes an overview on safety, and human error and its management in aviationoperations. It provides the necessary backgroundinformation to understand the rationale for LOSA.•Chapter 2 discusses the LOSA methodology and provides a guide to the implementation of LOSAwithin an airline. It also introduces a model of crewerror management and proposes the error classi-fication utilized by LOSA, which is essentiallyoperational and practical.•Chapter 3 discusses the safety change process that should take place following the implementation ofLOSA.•Chapter 4 introduces the example of one operator’s experience in starting a LOSA.•Appendix A provides examples of the various forms utilized by LOSA.•Appendix B provides an example of an introductory letter by an airline to its flight crews.•Appendix C provides a list of recommended reading and reference material.11.This manual is a companion document to the Human Factors Training Manual (Doc 9683). The cooperation of the following organizations in the production of this manual is acknowledged: The University of Texas at Austin Human Factors Research Project, Continental Airlines, US Airways and ALPA, International. Special recognition is given to Professor Robert L. Helmreich, James Klinect and John Wilhelm of The University of Texas at Austin Human Factors Research Project; Captains Bruce Tesmer and Donald Gunther of Continental Airlines; Captains Ron Thomas and Corkey Romeo of US Airways; and Captain Robert L. Sumwalt III of US Airways and of ALPA, International.Chapter 1BASIC ERROR MANAGEMENT CONCEPTS1.1INTRODUCTION1.1.1Historically, the way the aviation industry has investigated the impact of human performance on aviation safety has been through the retrospective analyses of those actions by operational personnel which led to rare and drastic failures. The conventional investigative approach is for investigators to trace back an event under consideration to a point where they discover particular actions or decisions by operational personnel that did not produce the intended results and, at such point, conclude human error as the cause. The weakness in this approach is that the conclusion is generally formulated with a focus on the outcome, with limited consideration of the processes that led up to it. When analysing accidents and incidents, investigators already know that the actions or decisions by operational personnel were “bad” or “inappropriate”, because the “bad” outcomes are a matter of record. In other words, investigators examining human performance in safety occurrences enjoy the benefit of hindsight. This is, however, a benefit that operational personnel involved in accidents and incidents did not have when they selected what they thought of as “good” or “appropriate” actions or decisions that would lead to “good” outcomes.1.1.2It is inherent to traditional approaches to safety to consider that, in aviation, safety comes first. In line with this, decision making in aviation operations is considered to be 100 per cent safety-oriented. While highly desirable, this is hardly realistic. Human decision making in operational contexts is a compromise between production and safety goals (see Figure 1-1). The optimum decisions to achieve the actual production demands of the operational task at hand may not always be fully compatible with the optimumFigure 1-1.Operational Behaviours — Accomplishing the system’s goals1-2Line Operations Safety Audit (LOSA)decisions to achieve theoretical safety demands. All production systems — and aviation is no exception —generate a migration of behaviours: due to the need for economy and efficiency, people are forced to operate at the limits of the system’s safety space. Human decision making in operational contexts lies at the intersection of production and safety and is therefore a compromise. In fact, it might be argued that the trademark of experts is not years of experience and exposure to aviation operations, but rather how effectively they have mastered the necessary skills to manage the compromise between production and safety. Operational errors are not inherent in a person, although this is what conventional safety knowledge would have the aviation industry believe. Operational errors occur as a result of mismanaging or incorrectly assessing task and/or situ-ational factors in a specific context and thus cause a failed compromise between production and safety goals.1.1.3The compromise between production and safety is a complex and delicate balance. Humans are generally very effective in applying the right mechanisms to successfully achieve this balance, hence the extraordinary safety record of aviation. Humans do, however, occasionally mismanage or incorrectly assess task and/or situational factors and fail in balancing the compromise, thus contributing to safety breakdowns. Successful compromises far outnumber failed ones; therefore, in order to understand human performance in context, the industry needs to systematically capture the mechanisms underlying suc-cessful compromises when operating at the limits of the system, rather than those that failed. It is suggested that understanding the human contribution to successes and failures in aviation can be better achieved by monitoring normal operations, rather than accidents and incidents. The Line Operations Safety Audit (LOSA) is the vehicle endorsed by ICAO to monitor normal operations.1.2BACKGROUNDReactive strategiesAccident investigation1.2.1The tool most often used in aviation to document and understand human performance and define remedial strategies is the investigation of accidents. However, in terms of human performance, accidents yield data that are mostly about actions and decisions that failed to achieve the successful compromise between production and safety discussed earlier in this chapter.1.2.2There are limitations to the lessons learned from accidents that might be applied to remedial strategies vis-à-vis human performance. For example, it might be possible to identify generic accident-inducing scenarios such as Controlled Flight Into Terrain (CFIT), Rejected Take-Off (RTO), runway incursions and approach-and-landing acci-dents. Also, it might be possible to identify the type and frequency of external manifestations of errors in these generic accident-inducing scenarios or discover specific training deficiencies that are particularly related to identified errors. This, however, provides only a tip-of-the-iceberg perspective. Accident investigation, by definition, concen-trates on failures, and in following the rationale advocated by LOSA, it is necessary to better understand the success stories to see if they can be incorporated as part of remedial strategies.1.2.3This is not to say that there is no clear role for accident investigation within the safety process. Accident investigation remains the vehicle to uncover unanticipated failures in technology or bizarre events, rare as they may be. Accident investigation also provides a framework: if only normal operations were monitored, defining unsafe behaviours would be a task without a frame of reference. Therefore, properly focused accident investigation can reveal how specific behaviours can combine with specific circumstances to generate unstable and likely catastrophic scenarios. This requires a contemporary approach to the investigation: should accident investigation be restricted to the retrospective analyses discussed earlier, its contribution in terms of human error would be to increase existing industry databases, but its usefulness in regard to safety would be dubious. In addition, the information could possibly provide the foundations for legal action and the allocation of blame and punishment.Combined reactive/proactive strategies Incident investigation1.2.4 A tool that the aviation industry has increasingly used to obtain information on operational human perform-ance is incident reporting. Incidents tell a more complete story about system safety than accidents do because they signal weaknesses within the overall system before the system breaks down. In addition, it is accepted that incidents are precursors of accidents and that N-number of incidents of one kind take place before an accident of the same kind eventually occurs. The basis for this can be traced back almost 30 years to research on accidents from different industries, and there is ample practical evidence that supports this research. There are, nevertheless, limitationsChapter 1.Basic error management concepts1-3on the value of the information on operational human performance obtained from incident reporting.1.2.5First, reports of incidents are submitted in the jargon of aviation and, therefore, capture only the external manifestations of errors (for example, “misunderstood a frequency”, “busted an altitude”, and “misinterpreted a clearance”). Furthermore, incidents are reported by the individuals involved, and because of biases, the reported processes or mechanisms underlying errors may or may not reflect reality. This means that incident-reporting systems take human error at face value, and, therefore, analysts are left with two tasks. First, they must examine the reported processes or mechanisms leading up to the errors and establish whether such processes or mechanisms did indeed underlie the manifested errors. Then, based on this relatively weak basis, they must evaluate whether the error manage-ment techniques reportedly used by operational personnel did indeed prevent the escalation of errors into a system breakdown.1.2.6Second, and most important, incident reporting is vulnerable to what has been called “normalization of deviance”. Over time, operational personnel develop infor-mal and spontaneous group practices and shortcuts to circumvent deficiencies in equipment design, clumsy pro-cedures or policies that are incompatible with the realities of daily operations, all of which complicate operational tasks. These informal practices are the product of the collective know-how and hands-on expertise of a group, and they eventually become normal practices. This does not, however, negate the fact that they are deviations from procedures that are established and sanctioned by the organization, hence the term “normalization of deviance”. In most cases normalized deviance is effective, at least temporarily. However, it runs counter to the practices upon which system operation is predicated. In this sense, like any shortcut to standard procedures, normalized deviance carries the potential for unanticipated “downsides” that might unexpectedly trigger unsafe situations. However, since they are “normal”, it stands to reason that neither these practices nor their downsides will be recorded in incident reports.1.2.7Normalized deviance is further compounded by the fact that even the most willing reporters may not be able to fully appreciate what are indeed reportable events. If operational personnel are continuously exposed to sub-standard managerial practices, poor working conditions and/or flawed equipment, how could they recognize such factors as reportable problems?1.2.8Thus, incident reporting cannot completely reveal the human contribution to successes or failures in aviation and how remedial strategies can be improved to enhance human performance. Incident reporting systems are certainly better than accident investigations in understanding system performance, but the real challenge lies in taking the next step — understanding the processes underlying human error rather than taking errors at face value. It is essential to move beyond the visible manifestations of error when designing remedial strategies. If the aviation industry is to be successful in modifying system and individual per-formance, errors must be considered as symptoms that suggest where to look further. In order to understand the mechanisms underlying errors in operational environments, flaws in system performance captured through incident reporting should be considered as symptoms of mismatches at deeper layers of the system. These mismatches might be deficiencies in training systems, flawed person/technology interfaces, poorly designed procedures, corporate pressures, poor safety culture, etc. The value of the data generated by incident reporting systems lies in the early warning about areas of concern, but such data do not capture the concerns themselves.Training1.2.9The observation of training behaviours (during flight crew simulator training, for example) is another tool that is highly valued by the aviation industry to understand operational human performance. However, the “production”component of operational decision making does not exist under training conditions. While operational behaviours during line operations are a compromise between production and safety objectives, training behaviours are absolutely biased towards safety. In simpler terms, the compromise between production and safety is not a factor in decision making during training (see Figure 1-2). Training behaviours are “by the book”.1.2.10Therefore, behaviours under monitored conditions, such as during training or line checks, may provide an approximation to the way operational personnel behave when unmonitored. These observations may contribute to flesh out major operational questions such as significant procedural problems. However, it would be incorrect and perhaps risky to assume that observing personnel during training would provide the key to understanding human error and decision making in unmonitored operational contexts.Surveys1.2.11Surveys completed by operational personnel can also provide important diagnostic information about daily operations and, therefore, human error. Surveys1-4Line Operations Safety Audit (LOSA)provide an inexpensive mechanism to obtain significant information regarding many aspects of the organization, including the perceptions and opinions of operational personnel; the relevance of training to line operations; the level of teamwork and cooperation among various employee groups; problem areas or bottlenecks in daily operations; and eventual areas of dissatisfaction. Surveys can also probe the safety culture; for example, do personnel know the proper channels for reporting safety concerns and are they confident that the organization will act on expressed concerns? Finally, surveys can identify areas of dissent or confusion, for example, diversity in beliefs among particular groups from the same organization regarding the appropriate use of procedures or tools. On the minus side, surveys largely reflect perceptions. Surveys can be likened to incident reporting and are therefore subject to the shortcomings inherent to reporting systems in terms of understanding operational human performance and error. Flight data recording1.2.12Digital Flight Data Recorder (DFDR) and Quick Access Recorder (QAR) information from normal flights is also a valuable diagnostic tool. There are, however, some limitations about the data acquired through these systems. DFDR/QAR readouts provide information on the frequency of exceedences and the locations where they occur, but the readouts do not provide information on the human behaviours that were precursors of the events. While DFDR/QAR data track potential systemic problems, pilot reports are still necessary to provide the context within which the problems can be fully diagnosed.1.2.13Nevertheless, DFDR/QAR data hold high cost/efficiency ratio potential. Although probably under-utilized because of cost considerations as well as cultural and legal reasons, DFDR/QAR data can assist in identifying operational contexts within which migration of behaviours towards the limits of the system takes place.Proactive strategiesNormal line operations monitoring1.2.14The approach proposed in this manual to identify the successful human performance mechanisms that contribute to aviation safety and, therefore, to the design of countermeasures against human error focuses on the monitoring of normal line operations.Figure 1-2.Training Behaviours — Accomplishing training goals。
ITIL术语中英文对照表
发表于：2009—11—29 14：26发表主题：ITIL术语中英文对照表,论坛里有看到，但是没有过验证，在这里贴ITIL术语中英文对照表Absorbed overhead 可分摊间接费用Absorption costing 吸收成本法，完全成本法Acceptance 验收Acceptance environment 验收环境Acceptance test 验收测试Access control 访问控制Accounting 会计核算Accuracy 准确度Action lists 行动列表Activity Based Costing （ABC) 作业成本法（ABC）Adaptive maintenance 适应性维护Additive maintenance 补充性维护Adjustability 可调整性Agreed Service Time (AST) 约定服务时段Alert 告警Alert phase 告警阶段Allocated cost 可直接分配成本Application 应用,应用系统Application maintenance 应用维护Application management 应用管理Application sizing 应用选型Application software 应用软件Apportioned cost 待分摊间接成本Architecture 架构Archive 存档Asset 资产Asset management 资产管理Assurance 保证Attributes 属性Audit 审计Auditability 可审计性Authentication 验证Authenticity 真实性Authorisation 授权Automatic Call Distribution （ACD）自动呼叫转发（系统）(ACD）Availability 可用性Availability management 可用性管理Availability Management Database (AMDB) 可用性管理数据库（AMDB）Backup 备份Balanced Scorecard （BSC) 平衡计分卡Baseline 基线Baseline security 安全基线Batch processing rate 批处理速度Benchmark 标杆Biometrics 生物测定学BS7799 BS7799Budgeting 预算编制Bug BUG(也可形象地译为“臭虫")Build 构建Building environment 构建环境Business 业务，商业Business capacity management 业务能力管理Business Continuity Management （BCM) 业务持续性管理（BCM）Business function 业务功能，业务职能部门Business Impact Analysis （BIA) 业务影响分析（BIA）Business process 业务流程Business recovery objective 业务恢复目标Business recovery plan framework 业务恢复计划框架Business recovery plans 业务恢复计划Business recovery team 业务恢复小组Business Relationship Management （BRM）业务关系管理（BRM）Business request 业务请求Business Unit (BU) 业务单元（BU）Bypass 临时措施Call 呼叫Call center 呼叫中心Capacity Database (CDB) 能力数据库(CBD）Capacity management 能力管理Capacity plan 能力计划Capacity planning 能力规划Capital investment appraisal 资本投资评估Capitalization 资本化Category 类别，分类Central point of contact 联络中心Certificate 证书Certification Authority (CA）认证机构(CA）Certify 认证Change 变更Change Advisory Board (CAB）变更顾问委员会(CAB)Change Advisory Board /Emergency Committee(CAB/EC)变更顾问委员会/应急委员会（CAB/EC）Change authority 变更授权Change builder 变更构建者Change control 变更控制Change document 变更文档Change history 变更历史Change log 变更日志Change management 变更管理Change manager 变更经理Change model 变更模式Change processing 变更处理Change Record 变更记录Change request 变更请求Chargeable unit 计费单元Charging 计费CI level 配置项级别Clarity 易理解性Classification 分类，分级Clean desk 桌面清理,桌面整理Client 客户Cold stand-by 冷支持Command，control and communications 命令、控制和协调Communication facility 通信设备，通信设施Compatibility 兼容性Completeness 完整性Complexity 复杂性Component Failure Impact Analysis （CFIA）组件故障影响分析（CFIA）Compromise 泄漏Computer 计算机Computer Aided Systems Engineering（CASE)计算机辅助系统工程(CASE）Computer center 计算机中心Computer operations 计算机操作Computer platform 计算机平台Computer system 计算机系统Computer Telephony Integration （CTI）计算机电话集成(系统）（CTI）Confidentiality 保密性Confidentiality，Integrity and Availability (CIA) 保密性、完整性和可用性(CIA) Configuration 配置Configuration baseline 配置基线Configuration control 配置控制Configuration documentation 配置文档Configuration identification 配置标识Configuration Item (CI) 配置项（CI）Configuration management 配置管理Configuration Management Database（CMDB）配置管理数据库（CMDB)Configuration management plan 配置管理计划Configuration manager 配置经理Configuration structure 配置结构Configure 配置Connectivity 连通性Contingency manager 应急经理Contingency plan 应急计划Contingency planning 应急规划Contingency planning and control 应急规划及控制Continuity 持续性Continuity manager 持续性经理Continuous availability 持续可用性Continuous operation 持续运作Contract 合同Control 控制Controllability 可控性Cookie Cookie（也可形象地译为“甜饼”,译者注）Correctability 可纠正性Corrective controls 纠正性控制Corrective maintenance 纠正性维护Corrective measures 纠正措施Cost 成本，费用Cost effectiveness 成本效益Cost management 成本管理Cost unit 成本单元Costing 成本核算Countermeasure 防范措施Cracker 骇客CRAMM CRAMM（英国中央计算机与电信局行风险分析和管理的方法。
美国赛博空间作战行动Cyberspace _Operations
CHAPTER II
CYBERSPACE OPERATIONS CORE ACTIVITIES
Introduction................................................................................................................II-1
3.应用
a、本出版物中确立的联合原则适用于联合参谋部、作战司令部指挥官、下属统一司令部、联合特遣部队、这些司令部的下属部门、各军种和作战支持机构。
b、本出版物中的指南具有权威性；因此，除非指挥官认为特殊情况另有规定，否则将遵循这一原则。如果本出版物的内容与出版物的内容发生冲突，则以本出版物为准，除非参谋长联席会议通常与其他参谋长联合会成员协调，提供了更为现行和具体的指导。作为多国（联盟或联盟）军事指挥部一部分的部队指挥官应遵循美国批准的多国原则和程序。对于未经美国批准的条令和程序，指挥官应评估并遵循多国司令部的条令与程序，如果适用并符合美国法律、法规和条令。
•联合职能部门和网络空间运作
第三章权限、角色和职责
•简介III-1
•当局III-2
•角色和职责
•法律考虑因素III-11
第四章规划、协调、执行和评估
•联合规划过程和网络空间运营
•网络空间运营规划考虑因素
•对网络空间的情报和操作分析支持
运营计划IV-6
•针对性IV-8
•网络空间部队的指挥与控制
系统集成项目经理该掌握专业英语词汇
系统集成项目经理该掌握专业英语词汇系统集成项目经理该掌握专业英语词汇Temporary————临时性Unique————独特的Project management————项目管理Project requirements————项目需求Initiating————启动Executing————实施Monitoring————监视Controlling————控制Closing————收尾Project manager————项目经理Project Charter————项目章程Project Management Plan————项目管理计划Phase————阶段Direct————指导Monitor————监视Reviewing————评审Change requests————变更请求Deliverables————可交付物、可交付成果Organizational process assets————组织过程资产Time Management————时间管理Activity Definition————活动定义Activity Sequencing————活动排序Activity Resource Estimating————活动资源估算Activity Duration Estimating————活动持续时间估算Schedule Development————进度计划Schedule Control————进度控制Cost Management————成本管理Cost Planning————成本计划Cost Estimating————成本估算Cost Budgeting————成本预算Cost Controlling————成本控制Quality Management————质量管理Quality Planning————质量规划Quality Assurance————质量保证Quality Control————质量控制Human Resource Management————人力资源管理Organize————组织Acquire————组建Develop————建设、发展Communications Management————沟通管理Collection————收集Dissemination————传输Storage————存储Disposition————处理Communications Planning————沟通规划Information Distribution————信息发布Performance Reporting————绩效报告Stakeholders————项目干系人、利害相关者Risk————风险Risk Management————风险管理Risk Management Planning————风险管理规划Risk Identification————风险识别Qualitative Risk Analysis————定性风险分析Quantitative Risk Analysis————定量风险分析Risk Response Planning————风险应对计划Risk Monitoring and Control————风险监控Procurement Management————采购管理系统集成项目经理该掌握专业英语词汇Plan Purchases and Acquisitions————采购规划Plan Contracting————合同计划Select Sellers————卖方选择Contract Administration————合同管理Contract Closure————合同收尾Projects——项目。
软考十大管理英文
3.制定预算(Determine Budget) The process of aggregating the estimated costs of individual activities or work packages to establish an authorized cost baseline.
6.控制进度(Control Schedule) The process of monitoring the status of project activities to update project progress and manage changes to the schedule baseline to achieve the plan.
4.建设团队(Develop Project Team) The process of improving competencies, team member interaction, and overall team environment to enhance project performance.
2.管理质量(management quality) The process of applying an organization's quality policy to projects and transforming quality management plans into executable quality activities.
信息系统项目管理专业术语-英文对照表
验收- (Acceptance)指客户检查接受项目交付物的过程。
活动- (Activity)指在项目中的任何消耗资源（人力、物理和设施）、产生相应成本并且生产出一项或多项产品的过程。
活动通常会在项目工作分解结构（WBS）中明确展示。
现货- (Actuals)指项目或其他活动中的实际成本和消耗。
这是一个测量值，通常会被用来与计划或预期值做比较。
假设- (Assumption)预先接受为正确但缺乏证据的事物或定义。
假设通常会发生在项目规划阶段并且被用来作为估算的基础。
应当将所有的假设都记录下来；如果某假设在后期被证明不正确，可以对计划和估算做出相应地调整。
基线- (Baseline)基线是一个快照：一个被记录在案的项目所处位置或状态。
尽管项目所处的位置以后可能会被更新，但基线始终保持不变，并可被作为原始状态的参照与当前项目位置的对照。
基线产品，例如软件系统，应当被永久地记录在案，以便于在将来的任何时间召回。
预算- (Budget)经分配的、用于执行项目工作的款额及其他资源。
商业论证- (Business Case)指启动和继续一个项目的原因和依据。
商业论证应当定义出项目可产生的商业及其他收益，以及项目的成本和时长。
商业论证同时还指出如何衡量项目的成功商业需求- (Business Requirements)指的是利益相关者对完成项目的商业要求。
在项目起始阶段，商业需求处于较高等级，但随着项目的发展，这些需求被不断地提炼并最终被细致地记录在正式的文件中。
变更控制- (Change Control)指按照计划的范围、工期和预算来管理控制变更。
在小型项目中变更控制可能采用非书面的、非正式的方式，但是在大型项目中，变更控制意味着一个牵涉很多项目利益相关者的、正式的过程（变更控制委员会）变更控制委员会- (Change Control Board (CCB))由利益相关者组成的团体，负责批准或否决对项目基线的变更。
工作规划的名词解释英文
Work planning is a systematic process that involves defining, organizing, and coordinating the tasks and activities required to achieve specific goals within a given timeframe. It is a critical component of project management, organizational strategy, and individual productivity. The term encompasses a range of activities aimed at ensuring that resources are utilized efficiently, timelines are met, and objectives are achieved. Below is a detailed explanation of work planning, including its key elements, importance, and benefits.Key Elements of Work Planning1. Goal Setting: The first step in work planning is to establish clear, measurable, achievable, relevant, and time-bound (SMART) goals. These goals serve as the foundation for the entire planning process and guide the allocation of resources and the prioritization of tasks.2. Task Identification: Once the goals are set, the next step is to identify all the tasks and activities that need to be completed to achieve those goals. This involves breaking down the project into smaller, manageable components.3. Resource Allocation: Work planning requires the identification and allocation of the necessary resources, including personnel, equipment, materials, and finances. This step ensures that the right resources are available at the right time to execute the tasks effectively.4. Time Estimation: Estimating the time required to complete each taskis crucial for effective work planning. This involves considering the complexity of the task, the skills of the personnel involved, and any potential delays or constraints.5. Task Sequencing: Once the tasks are identified and their durations estimated, they need to be sequenced in a logical order. This ensuresthat tasks are completed in the most efficient and effective manner,with dependencies and prerequisites being taken into account.6. Risk Management: Identifying potential risks and developingmitigation strategies is an essential part of work planning. This helpsin anticipating and addressing potential issues that could impact the project's timeline or outcome.7. Communication: Effective communication is key to successful work planning. This involves sharing information about the goals, tasks, timelines, and resource requirements with all stakeholders, including team members, clients, and suppliers.Importance of Work Planning1. Enhanced Efficiency: By systematically planning and organizing work, organizations can optimize the use of resources, reduce waste, and minimize inefficiencies.2. Improved Productivity: Work planning helps in prioritizing tasks and focusing on the most critical activities, leading to increased productivity and the timely completion of projects.3. Risk Mitigation: By identifying potential risks early on, work planning enables organizations to develop strategies to mitigate these risks, thereby reducing the likelihood of project failures or delays.4. Stakeholder Satisfaction: Clear and transparent work planning helpsin managing stakeholder expectations and ensuring that the project meets their requirements and deadlines.5. Learning and Adaptation: Work planning allows organizations toreflect on past projects and learn from their successes and failures, leading to continuous improvement and adaptation.Benefits of Work Planning1. Clarity: Work planning provides a clear roadmap for the project, making it easier for team members to understand their roles and responsibilities.2. Consistency: By following a structured planning process, organizations can ensure consistency in their approach to managing projects and tasks.3. Accountability: Work planning helps in assigning accountability for specific tasks, ensuring that everyone knows what is expected of them and can be held responsible for their performance.4. Scalability: Effective work planning can be scaled up or down depending on the size and complexity of the project, making it adaptable to various organizational needs.5. Cost Reduction: By minimizing inefficiencies and waste, work planning can lead to cost savings for organizations.In conclusion, work planning is a multifaceted process that is essential for achieving organizational and project success. It involves careful consideration of goals, tasks, resources, and timelines, and requires effective communication and risk management. By implementing a robust work planning process, organizations can enhance efficiency, productivity, and stakeholder satisfaction, while also fostering a culture of continuous improvement and adaptability.。
2012年百仕瑞公开课计划-华东版0927
3200
蔡岳
3200
蔡岳
3200
蔡岳
3200
蔡岳
3200
李海
2800
李海
3200
张仲豪
3200
张仲豪
精益工厂Lean 精益工厂Lean Factory
精益基础-精益生产体系和基础工具运用 Lean Foundation-Lean system and basic tools 精益进阶-精益生产进阶提升 Lean advanced-Lean Production Management Skills upgrading 精益沙盘-精益生产沙盘模拟 Lean Sand table-Lean Production Sand table simulation 精益改善-精益现场改善实战训练营 Lean improvement-Workshop for Lean－Site Improvements 精益实战-精益生产（工厂实战版） Lean combat-Lean Production Management（in factory） 5S与目视化实战 Workshop of 5S and the Visual-based 5S与目视化实战（工厂实战版） Workshop of 5S and the Visual-based（in factory） 2800 马千里
3200
马晓峰
3200
曹亮
供应商选择、评估与全面管理 Supplier Selection, Evaluation & Management 谁是谈判高手－情景、方法与技巧 Purchasing Negotiation Skills － Situation Ways & Skills 采购合同谈判中的风险控制与合同管理 Risk Control & Contract Management in Purchasing Contract Negotiations 采购系统设计与流程管理 Procurement system design and process management
集成电路设计专业名词解释汇总英文版
集成电路设计专业名词解释汇总英文版English:"Integrated Circuit (IC) Design: The process of creating a blueprint for the manufacturing of integrated circuits, such as microchips, using specialized software and tools. IC design involves several stages, including architectural design, logic design, circuit design, physical design, and verification. Architectural design establishes the high-level functionality and organization of the circuit, determining the overall structure and major components. Logic design involves the translation of the architectural design into a set of logic equations and functional blocks, specifying the logical operation of the circuit. Circuit design focuses on the actual implementation of the logic design, defining the electrical connections and components needed to achieve the desired functionality. Physical design, also known as layout design, involves the placement and routing of the components to ensure proper functioning and optimal performance, considering factors such as power consumption, signal integrity, and manufacturing constraints. Verification is the process of ensuring that the designed circuit meets the specified requirements and functions correctly under various conditions. Field-ProgrammableGate Array (FPGA): An integrated circuit that can be configured by the user after manufacturing. FPGAs contain an array of programmable logic blocks and interconnects, allowing for the implementation of various digital circuits. Hardware Description Language (HDL): A specialized programming language used to describe the behavior and structure of electronic circuits, facilitating the design and simulation of digital systems. Common HDLs include Verilog and VHDL. Electronic Design Automation (EDA) Tools: Software tools used in the design of electronic systems, including integrated circuits. EDA tools automate various stages of the design process, from schematic capture and simulation to layout and verification. Some popular EDA tools include Cadence Virtuoso, Synopsys Design Compiler, and Mentor Graphics Calibre. Very-Large-Scale Integration (VLSI): The process of integrating thousands or millions of transistors into a single chip. VLSI technology enables the creation of complex, high-performance integrated circuits, such as microprocessors and memory chips, by packing a large number of transistors into a small area. Application-Specific Integrated Circuit (ASIC): An integrated circuit customized for a particular application or purpose. Unlike FPGAs, ASICs are manufactured to perform a specific function, offering advantages in terms of performance,power consumption, and cost for mass production. ASIC design involves the development of custom circuitry optimized for a particular application, often using standard cell libraries and specialized design methodologies."中文翻译:"集成电路（IC）设计：是指利用专业软件和工具创建集成电路（如微芯片）制造的蓝图的过程。
CYME 9.1 电力工程与分析软件说明书
CYME power engineering and analysis softwareBrightlayer Utilities suiteImproving the capacity planning integration and the distributed resources planning processes CYME 9.1 new featuresCYME power engineering and analysis software, part of the the Brightlayer Utilities suite,is a next generation solution aimed at supporting utilities with efforts to modernize long-term grid planning framework. Built on core components such as chronological as-planned network modeling, time-series analysis, DER optimization,non-wires alternatives design, and project portfolio evaluation, CYME 9.1 enables the integration of the capacity planning and the distributed resources planning processes while paving the way to integrated system planning.Key new features include:• L ong-Term Planner module engineered to generatecrucial insights on long-termhourly demand forecasts anddistribution system’s risksand performance over time.• C able Thermal Rating module that brings forth the world-renowned CYMCAP Software calculation engine into therealm of the CYME Softwarefor the thermal analysisof underground duct bankstructures.• I mproved Steady-StateAnalysis with Profiles module that pushes the limit of time-series analysis through localor remote multiprocessingwith direct integration toexternal data sources via theDynamic Data Pull module.• C ontinuous improvement of the core components suchas Load Flow Analysis, LoadAllocation function, Advanced Project Manager, IntegrationCapacity Analysis, Load ReliefDER Optimization, Techno-Economic Analysis, and thePython® Scripting Tool.• O ptimized performances andseveral enhancements to thesoftware framework, userinterface, and the one-linediagram navigation.By working together withelectric utilities, we continue toadvance the capabilities of theCYME power system analysissoftware to meet the needsof any engineering study fromelementary to complex.Long Term Planner module:The CYME Long-Term Plannermodule combines a series oftools designed to optimally carrytasks involved with the annualplanning cycle. Leveragingthe power of the AdvancedProject Manager and the LoadFlow with profiles, a time-series analysis combining loadallocations and load flows,this module features powerfulforecast analytics capabilitiesand a novel report technologythat aggregates content fromdifferent contexts to providea holistic view of distributioncircuits and substationsperformance over time.The Brightlayer Utilities suite is a full complement ofsoftware applications that enable utilities to use datato optimize the performance and reliability of the grid,integrate renewables, comply with regulations andplan for the future. As a key solution in the BrightlayerUtilities suite, our CYME power engineering softwarecontinues to evolve its best-of-breed power systemanalysis software with the release of CYME 9.1, thesecond version of a new generation aimed atsupporting utilities in their efforts to align theirpractices with the climate and clean energy goalsof the 21st century.Follow us on social media to get the latest product and support information.Eaton1000 Eaton Boulevard Cleveland, OH 44122United States © 2021 EatonAll Rights Reserved Printed in USAPub No: BR917096EN / GG July 2021Eaton is a registered trademark. All other trademarks are property of their respective owners../cyme or contact us at ******************.Cable Thermal Rating module:•The CYME Cable Thermal Rating module brings forth the world-renowned CYMCAP Software calculation engine into the realm of the CYME Software, anabling users to performe ampacity and temperature rise calculations for power cable installations.•This module addresses steady-state and transient thermal cable rating as per the analytical techniques described by Neher-McGrath and the International Standards IEC 287© and IEC 853©.Engineering Analysis Enhancements: Load Flow — Load Allocation•Controls – Transformer tap operation with two new operation modes•Improved algorithm to better handle: Complex BESS controls, new voltage regulator control modes (bias controls), convergence for Y-g/D connected transformers.•Define Voltage Sensitivity Load Models and default power factors per load model (Customer Type dialog box) > seasonal parameters (winter, summer, etc.) •Time Parameters: a load model can now be assigned to each season.•Abnormal conditions: feeder sources fed by an upstream substation are now considered in the evaluation of abnormal loading conditions.Network Modeling Enhancements:•Voltage Regulator: The First House Protection option is now enabled for voltage regulators working in reverse mode.•It is now possible to Drag & Drop a network from the toolbox onto a node to create a network on this node.•Network Properties: three (3) more grouping attributes have been added to the Network Properties in order to capture additional non-electrichierarchical information for each circuit. The total number of grouping attributes is now six (6).•Network Properties: a new Limits tab has been added to define the circuit's capacity and protection trip limits to evaluate overload conditions at the circuit level or at the source.•Anew CYME Library of AC/DC converters is now available from the Equipment menu. This library features 680 documented converter models for a mix of single-phase and three-phase converters from different manufacturers.Improved User Experience:One-Line diagram•Anew option button has been added to prevent automatic “best display calculation” (switch from automatic to manual calculation).•The user experience has been improved when making multiple selections using the mouse and the Ctrl Key.•Now users can modify the Remote-controlled attribute for devices from the group properties dialog box.•“Low voltage cable/overhead" symbols concept extended to Secondary Network (CYME / SNA) type networks. Line styles can be now assigned by OH line or cable ID.Reports, Color Coding and Charts•Anew option is available to toggle a tabbed report into full-screen mode.•Create a report for upstream sections , including laterals.•Several new chart options have been added to enhance the user er interface and Framework•It is now possible to access the TCC settings of a protective device from the Properties control of the Explorer bar. •Anew icon for Batch Analysis has been defined.•Anew tool, “Clear simulation Circles”, is available in the toolbar to remove any circling of devices due to a simulation.•It is now possible to reset the position of the result’s box to be centered on the CYME software window.•Anew option, “Restore Layout”, was added to the CYME Configuration Setup dialog box to restore the display to default when restarting the CYME application.We are committed to advancing thecalculation engines and modeling capabilities of CYME power engineering software by continuously working to meet and exceed the evolving needs of the energy landscape. As part of the Brightlayer Utilities suite, CYME 9.1 is a fundamental tool enabling utilities to analyze data for optimizinggrid performance and reliability, integrating renewables today and planning forthe future.。
电子信息工程专业英语(第三版)词汇表
电子信息工程专业英语(第三版)词汇表Aa portion of一部分a variety of各种各样的a mass of 大量的AC abbr. Alternating Current交流电accidental adj.意外的accumulator n．累加器acquisition n.获取，采集acquisition time采集时间acquisition time采集时间activate vt．激活active adj．有源的actuator n 致动器，执行器add-on n.附件administration邮电管理局address vt.从事，忙于address generator地址产生器address pointer地址指针addressing mode寻址模式adjustment n 调整，调节ADSL abbr. Asymmetrical Digital Subscriber Loop非对称数字用户线adverse adj 不利的，相反的AFG Arbitrary Function Generator任意函数发生器aggregate v.聚集，合计AGP Accelerated Graphic Port 加速图形接口akin adj.同族的，类似的algorithm n.算法aliasing n.混叠现象alkaline adj.碱性的all in all 总而言之all of a sudden突然allocate vt.分配allocate vt.分配allow for 虑及，体谅allow for虑及，酌留alphanumeric adj.包括文字与数字的alter v.改变alternative n．选择ALU abbr Arithmetic Logic Unit算术逻辑单元aluminium n.铝ambient adj.周围的n.周围环境analogous adj.类似的analogy n.类似，类推ancillary adj.辅助的，副的anguish n 痛苦，苦恼angular frequency角频率annotation n.标注，注解antenna n.触角，天线anti-aliasing filter抗亍昆叠滤波器anti-aliasing filter抗混叠滤波器appliance n.用具，器具appliance n．用具，器县application interface 应用程序接口approach n. 方法appropriate adj.适当的approximation n.近似（值）approximation n.逼近，近似值archive vt.存档n.档案文件arena n.竞技场，舞台arena n.竞技场舞台arise from 由...引起；从...中产生arithmetic n 算数array n.阵列，数组array n.数组，阵列artificial adj.不自然的as a consequence 因此as always照常as opposed to .. 与...相反as yet到目前为止ASIC abbr. Application Specific Integrated Circuit专用集成电路ASIC Application Specific Integrated CircuitASIC Application-Specific Integrated Circuit专用集成电路assembler n 汇编器assembly language汇编语言assignment n.赋值ASSP abbr. Application Specific Standard Product专用标准器件ASSP Application-Specific Standard Parts 专用标准器件assume vt 假定asynchronous adj.异步的asynchronous adj.异步的attenuator n.衰减器audiophile n.高保真音响爱好者auditorium n.会堂，礼堂auditory system听觉系统automatic variable自动变量automotive adj.汽车的AWG Arbitrary Waveform Generator任意波形发生器B(be) known as…称作……(be) capable of…具备……的能力(be) equivalerit to相当于……，等价于……(be) proportional to与……成比例back bias 反向偏压backplane n.背叛backside n.背部，后方backward compatible向下兼容bar graph条形图bargain n.交易，协议，廉价品barrier n.隔板，势垒，阻挡层base station 基站base station基站baseband n.基带baud n 波特be concerned with…对……关心be encumbered with为……所累be mad e up of由……组成be referred to as.... 被称作...be thought of as…被认为……beam splitter 分光镜behavioral synthesis 行为综合beneficial adj.有益的，受益的Bessel filter贝塞耳滤波器biased adj.加偏压的，有偏向的bill of materials材料单BIOS abbr．Basic Input Output System基本输入输出系统bipolar adj.双极性的bit vector位向量bland adj．平淡的block diagram方框图blow up 爆炸，放大blur v 使……模糊BNC bayonet neill-concelman 同轴电缆卡环形接头boast v．夸耀Bode plot伯德图bond n. 接头Boolean variable 布尔变量boost n.升压，放大boot n.启动，引导，自举boot sector引导扇区bootstrap n. 引导程序bootstrap loader 引导装入程序brake n.刹车branch instruction分支指令brief adj.短暂的bring up 捉出，引出browse v.浏览budget n．预算budget n．预算budgetary adj.预算的buffer n 缓冲器buffer n.缓冲器，缓冲区building block 构件，模块built-in adj.内置的bulky adj.体积大的bulky adj 容量大的，体积大的bunching n.聚束bus interface总线接口bus interface总线接口by one’s （own）bootstraps 通过自己的努力by way of 经由；作为Ccable n.电缆cable modem 线缆调制解调器cable TV 有线电视cache n.高速缓存CAD Computer Aided Design 计算机辅助设计calculable adj.可计算的，能预测的calculation-intensive algorithm运算密集型算法camcorder n.便携式摄像机candid adj.非排演的，偷拍的capacitive adj.电容性的capacitor n.电容器capacity n.容量，电容capture v .记录，输入carrier wave 载波cascade n 级联cathode n.阴极cauldron n.大锅炉CB citizens'band 民用波段CCD Charge Coupled Device 电荷耦合器件CD Compact Disc 光盘cell n.细胞，蜂房，电池cellular adj.蜂窝状的characterization n.描述，表征charge pump电荷泵chat n．聊天Chebyshev Type l filter切比雪夫1型滤波器chip rate码片速率chrominance n.色度circular adj.圆形的，循环的circular adj．循环的，环形的circular buffer循环缓冲区class n.类clear-cut adj．界限分明的clever adj.精巧的，灵巧的，巧妙的cliché n 空话，套话，废话clock jitter 时钟抖动clump n.块，团CMOS abbr. Complementary Metal-Oxide-Semiconductor互补金属氧化物半导体coding theory 编码理论coexist vi.共存cold boot 冷启动collide vi.碰撞，抵触collision n.碰撞，冲突combat v.反对防止come down to归结为，涉及commute v 通勤comparable adj.可比较的，比得上的comparator n.比较器comparator n 比彰芝器compatibility n.兼容性compelling adj.强制的compiler n.编译器complex plane复平面complex-frequency variable复频率变量complicate vt使复杂，使难做，使恶化comply vi.遵守comply with同意，遵守component n 组件computing n.计算，处理concerned adj.有关的concisely adv.简明地concurrent adj.并发的concurrent process并发进程conditional adj.条件的conditioning n 调节，调整conduct v传导conductivity n. 传导性，传导率configure vt.配置，设定conflict n．冲突，抵触conformance n.顺应，一致conjugate adj．共轭的consequently adv.从而，因此consist of...由……组成consolidated adj。
调度策略解译-概述说明以及解释
调度策略解译-概述说明以及解释1.引言1.1 概述在调度系统中，调度策略是指通过一系列规则和算法来决定任务的执行顺序和资源的分配方式。
它是实现任务管理和资源利用的核心机制，对于提高系统的性能和效率至关重要。
调度策略可以影响任务执行的优先级、并发度、资源分配，以及任务之间的相互依赖关系等方面。
通过合理选择和设计调度策略，可以使系统能够更好地应对不同场景下的任务调度需求，最大程度地提高系统的整体性能和资源利用效率。
调度策略的目标是尽可能地减少任务执行的延迟和资源的浪费。
通过合理安排任务的执行顺序和资源的分配，可以使任务得到及时响应，并且能够充分利用系统中的各项资源，提高系统的吞吐量和并发能力。
调度策略的选择和设计需要考虑众多因素，包括任务的优先级、任务类型、资源的可用性、系统的负载状况等。
针对不同的应用场景和需求，可以采用不同的调度策略来实现最佳的任务调度效果。
本文旨在对调度策略进行深入解析和解释，包括调度策略的定义、分类以及其在实际应用中的作用和意义。
通过对各类调度策略的介绍和分析，希望能够帮助读者更好地理解和应用调度策略，从而为系统的任务调度和资源管理提供指导和参考。
1.2 文章结构文章结构:本文主要介绍调度策略的相关概念和分类，以及对未来调度策略的展望。
在引言部分，我们会概述本文的主题和内容，并介绍文章的结构安排。
在正文部分，我们将首先定义调度策略，并探讨其在实际应用中的重要性。
接着，我们会详细介绍调度策略的分类，包括静态调度策略和动态调度策略。
我们将会讨论不同的调度策略在不同场景下的应用与特点，以及其对系统性能和资源利用率的影响。
在结论部分，我们将对本文所介绍的内容进行总结，并展望未来调度策略的发展方向。
我们将探讨可能出现的新的调度策略，并讨论它们在面对不断变化的技术和环境条件时的应用前景。
通过对调度策略的深入研究和解读，我们将能够更好地理解和应用调度策略，为优化系统性能和资源利用提供参考和指导。
工作执行计划英语
工作执行计划英语Creating a work execution plan is an essential step in ensuring the successful completion of any project. 制定工作执行计划是确保项目顺利完成的一个重要步骤。
This plan serves as a roadmap for all team members, outlining their responsibilities, deadlines, and key milestones. 这个计划为所有团队成员提供了一份路线图，概述了他们的责任、截止日期和关键里程碑。
By carefully detailing each aspect of the project, a work execution plan helps to keep everyone on track and focused on the end goal. 通过仔细详细地描述项目的每个方面，工作执行计划有助于让所有人保持在正轨上，专注于最终目标。
In order to create an effective work execution plan, it is important to first clearly define the scope and objectives of the project. 确定项目的范围和目标是制定有效的工作执行计划的重要前提。
This involves identifying the specific deliverables that need to be completed, as well as any constraints or limitations that may impact the project timeline. 这涉及到确定需要完成的具体交付成果，以及可能影响项目时间表的任何约束或限制。
信息系统项目管理师重要专业英语词汇汇总
信息系统项目管理师重要专业英语词汇汇总1.项目管理基础和框架 - 关键术语项目（Project）运营（Operation）一般管理（General Management）项目管理（Project Management）大型项目（Program）子项目（Subproject）项目阶段（Project Phase）项目生命周期（Project Life Cycle）阶段出口或终止点（Phase exit or kill point）项目利益相关者/项目干系人（Stakeholder）过程（Process）控制（Control）可交付成果（Deliverable）项目经理（Project Manager）项目团队（Project Team）项目型组织（Projectized Organization ）2.项目整体管理挣值管理（Earned Value Management, EVM）变更控制委员会（Change Control Board）综合变更控制（Integrated Change Control）配置管理（Configuration Management）经验教训（Lessons Learned）3.项目范围管理项目章程（Project Charter）产品描述（Product Description）约束（Constraint）假设（Assumptions）项目范围（Project Scope）范围变更（Scope Change）范围定义（ Scope Definition）范围规划（Scope Planning）范围核实（Scope Verification）范围说明书（Scope Statement）工作分解结构（Work Breakdown Structure, WBS）工作包（Work Package）WBS字典（WBS Dictionary）工作责任分配矩阵（RAM Responsibility Assignment Matrix）4.项目时间管理活动（Activity）虚活动（ dummy Activity）工期（Duration, DU）项目网络图（Network Diagramming）顺序图法（Precedence Diagramming Method, PDM）箭线图法（Arrow Diagramming Method, ADM）计划评审技术（Program Evaluation and Review Technique, PERT）关键路径法（Critical Path Method, CPM）里程碑（Milestone）最早开始日期（Early Start Date, ES）最早完成日期（Early Finish Date, EF）最晚开始日期（Late Start Date, LS ）最晚完成日期（Late Finish Date, LF ）浮动时间（Float）资源平衡（Resource Leveling）5.项目成本管理资源计划（Resource Planning）成本估算（Cost Estimating）成本预算（Cost Budgets）类比估算（Analogous Estimating）应急储备（Contingency Reserve）S曲线（S-Curve）挣值（Earned Value, EV）6.项目人力资源管理组织规划（Organizational Planning）人员招募（Staff Acquisition）团队开发（Team Development）组织分解结构（Organizational Breakdown Structure, OBS）人员管理计划（Staffing Management Plan）权力（Power）7.项目采购管理合同（Contract）违约（Breach）终止（Termination）询价（Solicitation）8.项目质量管理项目质量管理（PQM）质量规划（Quality Planning）质量保障（Quality Assurance）质量控制（Quality Control）返工（Rework）9.项目沟通管理沟通规划（Communication Planning）信息发布（Information Distribution）绩效报告（Performance Reporting）管理收尾（Administrative Closure）绩效测量基准（Performance Measurement Baseline）沟通障碍（Barriers）10.项目风险管理风险（Risk）风险识别（Risk Identification）敏感性分析（Sensitivity Analysis）蒙特卡罗分析（Monte Carlo Analysis）应急规划（Contingency Planning）风险回避（Risk Avoidance）风险转移（Risk Transference）----------------------------------------------------------------------------1、立项管理NPV （Net Present Value）净现值：投资方案所产生的现金净流量以资金成本为贴现率折现之后与原始投资额现值的差额。
《工业工程专业英语》IE专业词汇句子摘抄2
缩略词：IE-industrial engineering-工业工程IT-information technology-信息技术BPR-business process redesign/reengineering-业务流程再设计/再造QR-operation research运筹学FMS-flexible manufacturing system-柔性制造系统DESS-discrete event stochastic system-离散事件随机系统AI-artificial intelligence-人工智能PMTS-pre-determined motion times system-预定动作时间系统MTM-methods time measurement-方法时间测量法MOST-Maynard operation sequence technique-梅纳德操作排序技术PTS-pre-determined time standards-预定时间标准法MSD-master standard data-主时间数据法MST-motion standard times-动作标准时间法SATO-speed-accuracy trade-off-速度和精度的平衡IDC- industrial-developing-country工业发展中国家WIP-work-in-process-在制品OEM-original equipment manufacturer-原始设备制造商IRR-internal rates of return-内部收益率NPV-net present value 净现值CAD-computer-aided design计算机辅助设计CAM- computer-aided manufacturing计算机辅助制造ABC-activity-based costing-基于活动的成本分析CE-concurrent engineering-并行工程DoD-Department of Defense-美国国防部IDA-the institute for defense analysis-防御分析研究所SPC-statistical process control-随机过程控制JIT-just-in-time--准时生产短语：持续改进-Continuous Improvement 人因学，工效学-Human Factors或者ergonomics 人机系统-Man-Machine System 车间活动-Shop-Floor Activities 仿真模型-Simulation Model 道德标准，执业准则-Code of Ethics 绩效测量-Performance Measure仿真-simulation 运筹学-operations research 质量改善工程-quality improvement engineering 管理服务-management services绩效改善工程-performance improvement engineering 物料搬运-material handling 物流-logistics 金融/财务管理-financial management 项目管理-project management 商业规划与开发-business planning and development质量运动-quality movement 数学规划-mathematical programming 预测-forecasting 专家系统-expert system 统计学-statistics 组织理论-organizational theory 单纯性（算）法-simplex algorithm 运输问题-transportation problem 网络问题-network problem 线性规划-linear programming组合优化问题-combinatorial optimization problem 多项式算法-polynomial algorithm 约束-constraint 界限-bound 网络排队模型-network queueing model 非凸的-nonconvex 仿真建模-simulation modeling 随机网络分析-stochastic network analysis 随机服务系统-stochastic service system 目标函数，目标方程-objective function 离散优化-discrete optimization 非线性优化-nonlinear optimization 多目标优化-multiobjective optimization 无约束优化-unconstrained optimization 整数优化-integer optimization 作业测量的劳动力标准-work-measured labor standards 动作分析-motion analysis 时间研究-time study 活动/工作抽样-activity/work sampling 历史数据-historical data 估算-estimate 预定动作时间系统-pre-determined motion times system 模块化安排法-modular arrangement 工作要素法-work factor 交互式专家系统-interactive expert system 评比因子-rating or leveling factor移动平均法-moving average approach 生产线平衡-manufacturing line balancing 职业危险-occupational hazards 面向人类的设计-human centered design 时间和动作的研究-time-and-motion study 工业心理学-industrial psychology 事故倾向性-accident proneness 工作生理学-work physiology 生物力学-biomechanics 人体测量学-anthropometry 人因工程-human factor engineering 工程心理学-engineering psychology 实验心理学-experimental psychology 系统工程-system engineering 人类感知-human perception 响应，反映-response 反馈回路-feedback loop 独立变量-independent variable视觉-visual sense 听觉-auditory sense 手动响应-manual response 语音响应-verbal response （人类的）特征变量-idiosyncratic variable 生理感应-physiological arousal 宏观工效学-macroergonomics 认知工效学-cognitive ergonomics 使用性研究-usability study 人类可靠性-human reliability 人机交互-human-computer interaction 骨骼失调，肌骨紊乱-musculoskeletal disorder 工厂布局-factory layout 产品式布局-product layout 工艺式布局-process layout 功能式布局-functional layout 单元式布局-cellular layout 模块式布局-modular layout 布局/设施设计-layout/facility design 物料搬运-material handling 制造单元-cell 生产量，生产率-throughput （布局的）可重组性，可重塑性-reconfigurability 工作中心-work center 契约制造-contract manufacturing 产品延迟差异化-delayed product differentiation 多通道制造-multichannel manufacturing 可扩展的机器-scalable machine工件-workpiece 分布式布局-distributed layout 敏捷布局-agile layout 废物处理设施-waste-disposal facility 能力分配-capacity assignment设备利用率-machine utilization 路径规划和调度-routing and dispatching 联合设施，公用设施-consolidated facility 星型布局-star layout工程经济学-engineering economics 信息系统-information system 制造系统-manufacturing system计算机集成制造系统-computer integrated manufacturing system 投资分析-investment analysis 收益率- rate of return 并行工程-concurrent engineering 风险分析-risk analysis企业一般管理费-overhead 间接成本-indirect cost 现金流-cash flow 轮廓评估-profile estimation 质量成本-quality cost 预防成本-prevention cost 估价成本-appraisal cost 失败成本-failure cost 技术成本-technological cost 系统成本-system cost 辅助成本-support cost 设备老化-equipment obsolescence 直接劳动力成本-direct labor cost 有形成本-tangible cost 既约成本-irreducible cost 无形成本-intangible cost 实际成本-real cost 机会成本-opportunity cost 灵敏度分析-sensitivity analysis 管理层支持-management support 强化沟通-enhanced communication 团队建设-team building 控制委员会，指导委员会-steering committee质量功能展开-quality function deployment 快速原型-rapid prototyping计算机辅助工艺规-computer-aided process planning 面向装配/制造的设计-design for assembly/manufacturing 面向可重复使用的设计-design for reusability 面向维护的设计-design for maintainability 面向可靠性的设计-design for reliability 技术创新，技术革新-technological innovation 产品生命周期-product life cycle 管理承诺-management commitment 持续改进-continuous improvement 以客户为中心-customer focus 员工参与-employee involvement 团队合作-teamwork 员工授权-employee empowerment 流程管理-process management质量控制圈-quality control circle 装配线-assembly line 大规模生产-mass production 与供应商的伙伴关系-supplier partnership 单元制造-cellular manufacturing 质量政策-quality policy 培训-training 产品/服务设计-product/service design 供应商质量管理-supplier quality management顾客参与-customer involvement 企业质量文化-corporate quality culture 战略质量管理-strategic quality management 服务的无形性-service intangibility 生产的同时性-simultaneity of production 易逝性-perishability1.将工业工程部重新命名，以明确描述其具体职能，期间所出现的问题与其说与实际完成的工作有关，倒不如说与问题的表象有关。
it系统项目管理英语表达
it系统项目管理英语表达IT Systems Project Management.IT systems project management is the process of planning, organizing, and executing IT systems projects. It involves a wide range of activities, from defining project scope and objectives to managing resources and deliverables. Effective IT systems project management is essential for ensuring that IT projects are completed on time, within budget, and to the required quality standards.Phases of IT Systems Project Management.IT systems project management typically involves the following phases:Project initiation: This phase involves defining the project scope, objectives, and deliverables. It alsoinvolves identifying the project team and stakeholders.Project planning: This phase involves developing a detailed project plan that outlines the project activities, timelines, and resource requirements.Project execution: This phase involves carrying outthe project activities according to the project plan. Italso involves managing risks and issues that may arise during the project.Project closure: This phase involves completing the project activities and delivering the project deliverables. It also involves evaluating the project and lessons learned.Key Considerations in IT Systems Project Management.There are a number of key considerations in IT systems project management, including:Scope management: Scope management involves definingthe project scope and objectives. It is important to ensure that the project scope is well-defined and agreed upon byall stakeholders.Time management: Time management involves planning and scheduling the project activities. It is important to develop a realistic project timeline and to track progress against the timeline.Cost management: Cost management involves planning and budgeting for the project. It is important to develop a realistic project budget and to track costs against the budget.Quality management: Quality management involves ensuring that the project deliverables meet the required quality standards. It is important to develop and implement a quality assurance plan.Risk management: Risk management involves identifying and managing risks that may impact the project. It is important to develop a risk management plan and to monitor risks throughout the project.Stakeholder management: Stakeholder managementinvolves identifying and managing stakeholders who are affected by the project. It is important to engage stakeholders throughout the project and to address their needs and concerns.Benefits of Effective IT Systems Project Management.Effective IT systems project management can provide a number of benefits, including:Improved project success rate: Effective project management can help to increase the likelihood of project success. By following a structured approach to project management, organizations can reduce the risk of project failure.Reduced project costs: Effective project management can help to reduce project costs. By planning and managing resources effectively, organizations can avoid unnecessary expenses.Shorter project timelines: Effective projectmanagement can help to shorten project timelines. By following a structured approach to project management, organizations can identify and eliminate inefficiencies.Improved project quality: Effective project management can help to improve project quality. By following quality management best practices, organizations can ensure that project deliverables meet the required quality standards.Increased customer satisfaction: Effective project management can help to increase customer satisfaction. By delivering high-quality project deliverables on time and within budget, organizations can meet and exceed customer expectations.Conclusion.IT systems project management is an essentialdiscipline for organizations that want to successfully implement IT systems projects. By following a structured approach to project management, organizations can increase the likelihood of project success, reduce project costs,shorten project timelines, improve project quality, and increase customer satisfaction.。
工作规划的名词解释英语
Work planning is a systematic process of organizing and structuring tasks, activities, and resources to achieve specific goals andobjectives within a defined timeframe. It involves the identification, prioritization, and allocation of resources, as well as theestablishment of timelines and milestones for the completion of tasks. Work planning is essential for ensuring that projects are completed on time, within budget, and with the desired quality.The process of work planning typically includes the following steps:1. Goal Setting: The first step in work planning is to define the goals and objectives of the project or task. This involves understanding the purpose of the work and identifying the desired outcomes. Goals shouldbe specific, measurable, achievable, relevant, and time-bound (SMART).2. Task Identification: Once the goals are established, the next step is to identify the tasks that need to be completed to achieve those goals. This involves breaking down the project into smaller, more manageable tasks. Each task should be clearly defined and have a specific objective.3. Resource Allocation: After identifying the tasks, the next step is to allocate the necessary resources to complete each task. Resources may include personnel, equipment, materials, and funding. It is essential to ensure that the resources are available and sufficient to complete the tasks.4. Time Estimation: Once the resources are allocated, the next step isto estimate the time required to complete each task. This involves considering the complexity of the task, the skills and experience of the personnel involved, and any dependencies between tasks. Time estimates should be realistic and allow for unexpected delays.5. Schedule Development: Based on the time estimates, a schedule is developed to outline the sequence of tasks and their respective timelines. The schedule should include milestones, which are key points that indicate the completion of specific phases or stages of the project. This helps in tracking progress and ensuring that the project stays on track.6. Risk Management: Work planning also involves identifying potential risks and developing strategies to mitigate them. Risks can arise from various sources, such as changes in project scope, resource constraints, and unforeseen events. By identifying and managing risks, the project team can minimize their impact on the project timeline and outcomes.7. Communication and Collaboration: Effective communication and collaboration are crucial in work planning. Team members should be informed about their roles, responsibilities, and the project's goals. Regular meetings and updates are essential to ensure that everyone is on the same page and working towards the same objectives.8. Monitoring and Control: Once the project is underway, it is important to monitor its progress and make adjustments as needed. This involves comparing the actual progress against the planned schedule, identifying any deviations, and taking corrective actions. Regular monitoring helps in ensuring that the project remains on track and that any issues are addressed promptly.In summary, work planning is a comprehensive process that involves setting goals, identifying tasks, allocating resources, estimating time, developing a schedule, managing risks, fostering communication and collaboration, and monitoring progress. By following a structured approach to work planning, organizations can improve their project management capabilities, enhance productivity, and achieve their objectives more efficiently.。
执行计划岗位英语
执行计划岗位英语Embarking on a journey to excel in the execution of plans, the role of a plan execution officer is pivotal in any organization. This position is not just about ticking boxes and following a set of instructions; it's about being the driving force behind the successful implementation of strategies and initiatives. It's about having a keen eye for detail, a strategic mindset, and the ability to adapt to the ever-changing landscape of business operations.As a plan execution officer, you are the linchpin that ensures all components of a project come together seamlessly. You're the one who translates high-level objectives into actionable steps, ensuring that every team member is aligned with the company's vision. Your role is to keep the momentum going, to facilitate communication between departments, andto troubleshoot any issues that may arise during theexecution phase.The art of execution lies in your ability to anticipate challenges and devise solutions before they become roadblocks. It's about being proactive, not reactive. You must be amaster of time management, juggling multiple tasks and deadlines with ease. Your communication skills are your greatest asset, as they enable you to convey complex ideas in a clear and concise manner.In this role, you are also a mentor and a leader. Youguide your team through the intricacies of project management, fostering a culture of accountability and excellence. You set the standard for performance and inspire others to meet it.Ultimately, the execution of plans is the heartbeat of an organization's success. As the plan execution officer, youare the conductor of this symphony, ensuring that every noteis played in harmony. The satisfaction of seeing a plan cometo fruition under your stewardship is unparalleled, and the impact you have on the growth and success of the company is immeasurable.。
Integrating motor schemas and reinforcement learning
Abstract
Motor schemas are an important example of behavior-based robot control. The motor schema paradigm is the central method in use at the Georgia Tech Mobile Robot Laboratory, and is the platform for this research. Motor schemas are the reactive component of Arkin's Autonomous Robot Architecture (AuRA) 2]. AuRA's design integrates deliberative planning at a top level with behavior-based motor control at the bottom. The lower levels, concerned with executing the reactive behaviors are incorporated in this research. 1
tucker@
GIT-CC-97-11
Clay is an evolutionary architecture for autonomous robots that integrates motor schema-based control and reinforcement learning. Robots utilizing Clay bene t from the real-time performance of motor schemas in continuous and dynamic environments while taking advantage of adaptive reinforcement learning. Clay coordinates assemblБайду номын сангаасges (groups of motor schemas) using embedded reinforcement learning modules. The coordination modules activate speci c assemblages based on the presently perceived situation. Learning occurs as the robot selects assemblages and samples a reinforcement signal over time. Experiments in a robot soccer simulation illustrate the performance and utility of the system.
Hyperion全面预算管理方案
Oracle Fusion editions of Oracle's Hyperion performance management products are currently available only on Microsoft Windows server platforms. The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, functionality or platforms, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features, functionality or platforms described for Oracle’s products remains at the sole discretion of Oracle. Oracle may not release the functionality or products discussed on any additional platforms in the future.ORACLE HYPERION PLANNINGKEY FEATURES:• Integrated, best-in-class analytical capabilities• Synchronized financial and operational planning processes• Microsoft Office integration • Simplified Web interface • Predefined functional solutions• Powerful workflow• Scalable architecture• Robust data integrationKEY BENEFITS:• Create accurate forecasts • Reduce budgeting and planning cycles by weeks or months• Integrate financial and operational planning in one system• Meet immediate finance needs while enabling operations-specific budgeting processes• Support advanced, power-user modeling capabilities with seamless Microsoft Excel integration• Appeal to a wider user community through a simplified Web interface• Add quick-to-implement, prepackaged, fully-supported functional solutions• Meet industry-specific needs by leveraging industry expertise Oracle Hyperion Planning is a centralized, Microsoft Excel- and Web-based planning, budgeting, and forecasting solution that integrates financial and operational planning processes. Oracle Hyperion Planning provides an in-depth look at business operations and its impact on financials by tightly integrating financial and operational planning models. With Oracle Hyperion Planning, you can meet your immediate financial planning needs while enabling a platform for future cross-functional expansion and automated process integration.Lay the Foundation for Integrated PlanningAccurately predicting revenue and operating performance is a daunting challenge facing many enterprises today. Despite their awareness of the significant adverse impact that missed forecasts and plans can have on their business, the most common solution for budgeting and planning is still the disconnected spreadsheet that makes the planning process unreliable and inefficient. The resulting long budget cycles and forecasting inaccuracies prevent responsiveness to change, causing companies to miss business opportunities while wasting money and resources on detoriating business segments.Oracle Hyperion Planning allows you to rationalize the existing spreadsheet environment in a single planning system while enabling a platform for future organizationwide deployment. As a result, you can meet your immediate planning needs, typically driven by finance, while providing an expanded budgeting process that integrates finance and business units.Ensure Accurate Forecasts with Integrated Analytical Capabilities Oracle provides native business intelligence and planning capabilities in a single environment. Oracle Hyperion Planning lets you perform sophisticated analysis of business operations and use that information to create accurate forecasts and plans without switching to different tools. Using a single interface, you can access user-friendly dashboards, interactive analytics, and richly formatted financial reports while interacting with the planning system. With the Enterprise Performance Management Architect feature, you can easily create, manage, and integrate your planning processes for faster deployment and lower cost of ownership.Figure 1: The intuitive Web interface of Oracle Hyperion Planning allows financial and business users to quickly update key business drivers and see their related impact on financials.Expand Beyond Finance to Support Operational Business Planning Oracle Hyperion Planning synchronizes your financial and operational planning processes in your organization by connecting planning activities—such as sales revenue planning, marketing campaign planning, demand planning, workforce planning, and operations project planning—throughout the supply chain. With driver-based planning capabilities, nonfinancial users can enter business and operational drivers, while sophisticated business rules will then calculate their financial impact. By expanding financial planning into operational planning, you can attain more-accurate forecasts, and consequently, better adapt to market changes.Leverage Existing User Competencies with Microsoft Office Integration Oracle Hyperion Planning allows you to update and report plans using familiar tools such as Microsoft Office Excel, PowerPoint, and Word. Through Excel, for example, you can easily perform sophisticated financial modeling and reporting using reliable and up-to-date information from the central planning database. Oracle Hyperion Planning lets you take your plans offline, and at your own convenience, change assumptions, perform calculations, analyze results, and connect back to the central database to synchronize your updates.Additionally, you can directly integrate data from the planning database into PowerPoint and Word. This allows you to quickly create highly customized reports with accurate information and automatically refresh the report when the underlying data changes. The end result: reduced manual intervention, improved data integrity, and increased reporting accuracy.Figure 2: The built-in Microsoft Office integration in Oracle Hyperion Planning allows users to update and report plans in Excel, PowerPoint, and Word.Connect a Wide User Community with Simplified Web InterfaceOracle Hyperion Planning enables your business users to easily view plans and reports regardless of their computer skills. It also guides your less-frequent users step by step through the planning process using a wizard called Task Lists, allowing you to reduce cycle times while ensuring that your plans are complete and reliable. With Oracle Hyperion Planning’s secure Web infrastructure and user-friendly interface, you can even include your external business partners, such as suppliers or distributors, in your planning process. As a result, you will be able to increase the timeliness and accuracy of your forecasts.Enable Quick Implementations Through Predefined SolutionsThe Oracle Hyperion Planning platform allows you to easily add predefined modules for workforce planning and capital asset planning thereby reducing implementation time and cost while providing best-practice functionality.• Oracle Hyperion Workforce Planning lets you quickly and efficiently plan for head count, salary, and compensation across the enterprise. By automatically connecting with your workforce data, it helps you assess the business impact of your workforce decisions in real time.• Oracle Hyperion Capital Asset Planning allows you to plan existing and new capital assets, maintenance, transfers, and depreciation while analyzing their impact on income, balance sheet, and cash flow.Ease Collaboration and Maintain Control with Powerful Workflow Oracle Hyperion Planning’s powerful workflow functionality—including e-mail notifications, alerts, and task lists—empowers users to track and communicate the progress of plans and budgets. In addition to creating, validating, and changing plans and task lists, you can also identify bottlenecks and conduct what-if analysis and scenario testing.Support Large Deployment with Scalable Architecture Oracle Hyperion Planning is built on a scalable architecture, which allows you to, if appropriate, start with an initially small deployment before rolling it out to hundreds or thousands of required users in the organization. The architecture allows for flexible data entry, analysis, and frequent real-time updates from anywhere, in a safe and secure environment using a standard Web browser. Moreover, Oracle Hyperion Planning provides collaborative planning processes and dependable, centralized data management, significantly reducing your maintenance and distribution costs. Oracle Hyperion Planning interoperates with existing security mechanisms to help ensure maintenance and security consistency. Ensure Data Quality with Robust Data Integration Oracle Hyperion Planning facilitates bidirectional information exchange with legacy applications, data warehouses, and enterprise resource planning and transaction systems. It is an open and extensible system that leverages data that was previously “locked” in silos and other systems. Contact Us For more information about how your organization can leverage the power of Oracle Hyperion Planning, please visit or call +1.800.ORACLE1 to speak to an Oracle representative.Copyright 2007, 2008, Oracle. All Rights Reserved.This document is provided for information purposes only, and the contents hereof are subject to change without notice. This document is not warranted to be error-free, nor is it subject to any other warranties or conditions, whether expressed orally or implied in law, including implied warranties and conditions of merchantability or fitness for a particular purpose. Wespecifically disclaim any liability with respect to this document, and no contractual obligations are formed either directly or indirectly by this document. This document may not be reproduced or transmitted in any form or by any means, electronic or mechanical, for any purpose, without our prior written permission.Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners.ORACLE’SPERFORMANCEMANAGEMENTAPPLICATIONSOracle’s performancemanagement applicationscomprise a modular suite ofintegrated applications thatsupport a broad range ofstrategic and financialperformance managementprocesses to enablemanagement excellence.Part of Oracle’s enterpriseperformance managementsystem, these appli-cations can be quicklydeployed out of the box,extended with Oracle’sbusiness intelligenceproduct family, or tailored tomeet your organization'sspecific needs.RELATED PRODUCTS:Oracle’s performancemanagement applicationsinclude the followingproducts:• Oracle Hyperion FinancialManagement• Oracle HyperionPerformance Scorecard• Oracle Hyperion StrategicFinance• Oracle HyperionWorkforce Planning• Oracle Hyperion CapitalAsset Planning• Oracle HyperionProfitability and CostManagement• Oracle IntegratedOperational Planning• Oracle Crystal Ball。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
CENTRO PER LA RICERCASCIENTIFICA E TECNOLOGICA38050 Povo (Trento), ItalyTel.: +39 0461 314312Fax: +39 0461 302040e−mail: prdoc@itc.it − url: http://www.itc.itInterleaving Execution and Planning for Nondeterministic, Partially Observable DomainsBertoli P., Cimatti A., Traverso P.May 2004Technical Report # T04−05−02©Istituto Trentino di Cultura, 2004LIMITED DISTRIBUTION NOTICEThis report has been submitted for publication outside of ITC and will probably be copyrighted if accepted for publication. It has beenissued as a Technical Report for early dissemination of its contents. In view of the transfert of copy right to the outside publisher, itsdistribution outside of ITC prior to publication should be limited to peer communications and specific requests. After outside publication, material will be available only in the form authorized by the copyright owner.Interleaving Execution and Planning for Nondeterministic,Partially Observable Domains Piergiorgio Bertoli and Alessandro Cimatti and Paolo TraversoAbstract.Methods that interleave planning and execution are apractical solution to deal with complex planning problems in non-deterministic domains under partial observability.However,most ofthe existing approaches do not tackle in a principled way the impor-tant issue of termination of the planning-execution loop,or only doso considering specific assumptions over the domains.In this paper,we tackle the problem of interleaving planning andexecution relying on a general framework,which is able to deal withnondeterministic,partially observable planning domains.We pro-pose a new,general planning algorithm that guarantees the termi-nation of the interleaving of planning and execution:either the goalis achieved,or the system detects that there is no longer a guaranteeto progress toward it.Our experimental analysis shows that our algorithm can efficientlysolve planning problems that cannot be tackled with a state of the artoff-line planner for nondeterministic domains under partial observ-ability,MBP.Moreover,we show that our algorithm can efficientlydetect situations where progress toward the goal can be no longerguaranteed.1IntroductionPlanning in nondeterministic domains under partial observability isone of the most significant and challenging planning problems.Sev-eral approaches have been proposed in the past[15,18,3,2,1].How-ever,the problem has been shown to be hard,both theoretically andexperimentally,and building plans purely off-line still remains un-feasible in most realistic applications.Methods that interleave plan-ning and execution,see,e.g.,[13,11,17]are the practical alternativeto the problem of planning off-line with large state spaces.In safelyexplorable domains[11],i.e.,domains where execution cannot gettrapped in situations where plans that lead to the goal no longer exist,it is possible to devise methods that are complete,i.e.,that guaran-tee to reach the goal if there exists a solution,and that guarantee thetermination of the planning/execution loop if no solution exists.In this paper,we tackle the problem of interleaving planning andexecution in the general case of nondeterministic domains and partialobservability.We define an architecture for interleaving plan genera-tion and plan execution,where a planner generates conditional plansthat branch over observations,and a controller executes actions inthe plan and monitors observations to decide which branch has to beexecuted.This extends an off-line approach to planning in nonde-terministic,partially observable domains,based on symbolic modelchecking[2,1];the framework exploits the same data structures usedin off-line planning to generate plans,to execute them and to moni-tor their execution.The plan generation component exploits a novel Notice that considering weak plans,that may not reach the goal,would remove the guarantee of termination.Nondeterministic behaviors of the domain might in fact cause endless planning/execution loops where weak plans always exist,and every time fail to reach the goal.Figure1.A simple robot navigation domainThe way actions evolve the domain status,and observations are related to the current status,can be described by means of nondeter-ministic functions;more formally:Definition1A nondeterministic planning domain with partial ob-servability is a tuple,where:is the set of states.is the set of actions.is the set of observations.is the set of initial states;we require.is the transition function;it associates to each current state and action the set of next states.is the observation function;it associates to each state the set of possible observations.Action is executable in state if;it is executable in a set of states iff it is executable in every state.We require that in each state there is some executable action.Also,some observation must be associated to each state,i.e.,. This model allows for uncertainty in the initial states and in the out-come of action execution.Also,since the observation associated to a given state is not unique,it is possible to model noisy sensing and lack of information.In the domain offigure1,the actions are GoNorth,GoSouth, GoWest and GoEast.We have four states,corresponding to the four positions of the robot in the room.It is in general possible to present the state space by means of state variables,where each state is presented by a truth assignment to the state variables.In the ex-ample,the state variables might be E and S.In state NW they would be both associated with a false value(),while in SE they would be associated with.We have16observations,each corresponding to one of the possible configurations of walls around the robot.The space of observation can also be presented by means of observations variables,(e.g.WallN,WallS,WallW and WallE).Each obser-vation associates a truth value to each observation variable.In the following,we will assume a variable-based presentation for and .We define to denote the set of states that are compatible with a assignment to the observation variable;is inter-preted similarly.In partially observable domains,we consider plans that branch on the value of observation variables.Definition2((Conditional)Plan)A plan for a domain is either the empty plan,an action,the concatenation of an action and a plan,or the conditional plan, with an observation variable of the domain.For instance,corresponds to the plan“if you see a wall north,then move south,otherwise move west”.A plan is executable on a set of states if is empty,if, or if one of the following holds:,is executable on,and is executable on,and the plans and are executable over and,respectively.Intuitively,given a domain,a set of initial states and a set of goal states in,a plan is a strong solution for the planning problem iff it is executable on,and every execution on the states of results in[2].During the execution of a plan,in general the executor has to consider a set of states which are equally plausible given the ini-tial knowledge,and given the informations acquired through current and past observations.We call this set a belief state;it represents the current knowledge about the domain status.The search space then can be seen as an and-or graph whose nodes are belief states,con-sidering that actions transform belief states into new belief states, and observations identify subsets of the current belief state.Search for a plan can be performed by visiting and-or graph representing the search space;given its size,a convenient“lazy”approach recursively constructs the graph from the initial belief state,expanding each en-countered belief state by every possible combination of applicable actions and observations.The graph is possibly cyclic;in order to rule out cyclic plan behaviors,however,its exploration–at planning time–can be limited to the acyclic prefix of the graph.The initial belief state in Figure1is;our goal is to reach the condition.In the example,the actions are determin-istic,with the exception of moving GoSouth from,which may cause the robot to slip in one of two states.Figure1depicts a portion of thefinite prefix of the search space for the described problem.The prefix is constructed by expanding each node in all possible ways,each represented by an outgoing arc. Single-outcome arcs correspond to simple actions(action execution is deterministic in belief space).For instance,N4expands into N7by the action GoSouth.Multiple outcome arcs correspond to observa-tions.For instance,node N2results in nodes N4and N5,correspond-ing to the observation of WallN.3Interleaving Planning and ExecutionRather than searching the and-or graph of belief states off-line,tak-ing into account all possible contingencies that can arise,we pro-pose a framework where a planner searches the graph partially,and a controller executes the partial plan and monitors the current state of the domain.The process is iterated until the goal is(hopefully) reached.The top level algorithm for interleaving planning and exe-cution,called P LAN E XEC M ONITOR,is the following:P LAN E XEC M ONITOR(,)1if()2return;3:=P ROGRESSIVE P LAN(,);4if(=)5return;6else7:=E XECUTE M ONITORING(,);8P LAN E XEC M ONITOR(,);P LAN E XEC M ONITOR is initially invoked by passing to it the set of possible initial states,and the set of goal states.We assume the domain representation to be globally available.P LAN E XEC M ON-ITOR recursively implements a loop,that alternatively calls P RO-GRESSIVE P LAN,which generates a plan,and the monitored executor E XECUTE M ONITORING,that executes the plan,and at the same time reports the new belief resulting from execution.P LAN E XEC M ONI-TOR stops either when given a belief such that the goal is known to be reached(that is,),or when the planner returns failure.E XECUTE M ONITORING(,)1M ARK E XECUTED();2if3return;4if5A CTUATE();6:=;7E XECUTE M ONITORING(,);8if if then else9if C URRENT V ALUE10return E XECUTE M ONITORING(,);11else12return E XECUTE M ONITORING(,);The executor E XECUTE M ONITORING recursively applies the plan actions to the domain,via A CTUATE.The plan execution is driven by the observations in the plan:it branches over the actual observation values,retrieved from the domain via C URRENT V ALUE.Parallel to this,E XECUTE M ONITORING uses a domain model,namely and ,to propagate the initial belief consistently with the execution. Each belief state traversed during the monitored execution is marked as traversed via M ARK E XECUTED.We call a sequence of beliefs traversed by the plan during its monitored execution a run of a plan; several runs are possible from a starting belief state,depending on the behavior of the domain:Definition3(Runs of a plan)Let be a plan for a domain.The set of runs of from an initial belief state is inductively defined as follows.If is,then is a run of from.If is,then the sequence is a run of from,where is a run of from.If is,then the sequences andare runs of from,where is a run of from,and is a run of from.P ROGRESSIVE P LAN(,)1:=M K I NITIAL G RAPH(,);2while(IS S UCCESS(GET R OOT())3IS E MPTY F RONTIER()4(IS P ROGRESS(GET R OOT())5T ERMINATION C RITERION()))6:=E XTRACT N ODE F ROM F RONTIER();7if(S UCCESS P OOL Y IELDS S UCCESS(,))8M ARK S UCCESS();9N ODE S ET P LAN(,R ETRIEVE P LAN(,)); 10P ROPAGATE S UCCESS(,);11else12:=E XPAND N ODE();13E XTEND G RAPH(,,);14if(IS E XECUTED())15M ARK P ROGRESS(,);16P ROPAGATE P ROGRESS(,);17end while18if(IS S UCCESS(GET R OOT()))19return E XTRACT S UCCESS P LAN();20if(REACHED T ERMINATION)21return E XTRACT P ARTIAL P LAN();20if(IS P ROGRESS(GET R OOT()))21return E XTRACT P ROGRESSING P LAN();22return;Figure2.The planning algorithm4Planning for InterleavingConsider the planning algorithm depicted in Figure2,disregard-ing the lines with boldfaced labels.This is a slight modification for strong planning under partial observability described in[1].The al-gorithm takes as input the initial belief state and the goal belief state, and proceeds by incrementally constructing afinite acyclic prefix of the search space,implemented as a.In the graph,each node is associated with a belief state;a directed connection be-tween a node and a node results either from an action such that,or from an observation such that,with or.We call the fa-ther of and the son of;we call“brothers”all the nodes that result from the same observation expansion of the same node.The graph is annotated with a frontier of the nodes that have not yet been expanded,and with a success pool,containing the nodes for which a strong plan has been found.The algorithm has its core in a search loop(lines2-21),iteratively selecting and expanding a node in the ly,at each itera-tion,a node is extracted from the frontier,and evaluated for success against the success pool(lines6-7).If the node successful,a strong plan is extracted and associated to it,the success pool is expanded, and success is propagated backward on the graph(lines8-10).Oth-erwise,the node is expanded by applying every executable action, and non-trivial observation to it,resulting into a graph expansion (lines12-13).The expansion routine avoids generating ancestors of the expanded node,inhibiting the presence of loops in the graph. The search loop terminates either when(a)the root of the graph is signaled as a success node,(b)the graph frontier is empty,or(c)a termination criterion is met.Condition(a)signals that a strong plan has been found;condition(b)indicates that no strong plan exists; condition(c)is responsible for the search being stopped while only partial plans have been expanded.Notice that the criterion defining condition(c)is the only distinc-tion with the original off-line approach,for the purpose of integrat-ing the planner within the interleaving framework,added to generate possibly partial plans rather than searching the whole search space. (Different termination criteria could be envisaged,e.g.partial suc-cess,number of nodes,run times.The specific details are not relevant here.)Unfortunately,this simple minded approach does not guaran-tee termination of the overall interleaving loop,even if a solution ex-ists.The problem is that the planner should guarantee that,for every possible run,at least a new belief state is reached during execution. If this is not so,plan-execution loops are possible that keep visiting the same beliefs over and over,never terminating.If instead the guar-antee is achieved,termination of plan-execution loops follows from the fact that belief states arefinitely many.(Notice that weakening the condition,and accepting plans which only might result in beliefs never been visited before,does not guarantee termination.)The no-tion that guarantees the termination of the top level is what we call progressiveness of the planner:each plan must guarantee that at least one belief state(not just a state)is traversed that has not been previ-ously encountered during execution.Definition4(Progressive Plan)Let be a run.Let be a plan for.The plan is progressive for the run iff,for any runof from,there is at least one belief state in that is not a belief state of.Let us consider again the statements in Figure2(with lines20and 21bold,replacing the non-bold counterparts).In order to obtain a progressive planning algorithm,we consider all the plans originating from a node,to make sure that at least for one of them,execution will visit a belief state,which has never been visited during previ-ous executions.The graph is therefore extended in order to maintain up-to-date information on progress of nodes.In order to guarantee progressiveness,at line14,we check if has already been visited at execution time.If not,we mark as a“progress”node (line15)(we remark that a successful node has surely not been vis-ited by a previous run,and as such it is marked as progress).In that case,the progress information is recursively propagated bottom-up on the tree(P ROPAGATE P ROGRESS O N T REE,line16):if the node is the result of the application of an action,then its father is marked as progress.If the node is the result of an observation,in order to propagate its progress backward it is necessary to check that all of its brothers are also marked as progress nodes.Finally,when the loop is exited,either a strong plan has been found,and is returned by E XTRACT S UCCESS P LAN;or,a progress-ing plan exists in the graph,and is extracted by E XTRACT P RO-GRESSING P LAN;or,failure is returned.While extracting the success plan is simple(it is associated with by the bottom-up propaga-tion),the progressing plan might not be unique:several such plans may exist.The selection operated by E XTRACT P ROGRESSING P LAN may affect the overall performance.Our implementation privileges, amongst progressing plans,the ones performing more observations.5ExperimentsWe implemented a planner called MBP P(Model Based Planner with Progressiveness)which extends the offline MBP[2]planner with the algorithm shown in Section3,and is equipped with a simulator to trace executions.MBP P,just as MBP,relies on symbolic data structures to represent the search;these are based on Binary Decision Diagrams(BDDs),also exploited within planners such as UMOP[9]. We compare the interleaved approach of MBP P with the state of the art(offline)MBP;comparison with UMOP is not possible since it does not handle partially observable domains.The tests were run on a Pentium III,700MHz with6GB RAM,running Linux.The mem-ory limit was set to512MB,and CPU timeout was set at180sec. For each problem instance,we collect the planning time for MBP. For MBP P,the information is statistical,since its performance de-pends on the actual domain behavior chosen by the simulator:there-fore,for each problem instance,100runs were generated,with initial states and nondeterministic outcomes selected randomly.We report average times for the total of MBP P planning and randomized plan execution.For our experiments,we considered three classes of experiments. Thefirst is a robot navigation problem in a maze,with nondetermin-istic action effects.The robot may start at any position in the maze, and has to reach the top left corner.The robot may move in the four directions,and is equipped with reliable wall-presence sensors in the four directions.The robot may nondeterministically slip on thefloor while trying to move,in which case it stays in the same position;this can occur at most20%of the times.We consider a set of randomly generated mazes of increasing sizes.The second set of test cases is taken from the Power Supply Restoration(PSR)domain,recently proposed as a significant bench-mark for planning under partial observability[16].In PSR,a network of electricity supply lines,which can be reconfigured by turning(pos-sibly unreliable)switches,are fed by a set of generators.Possible faults on the lines cause reopening of switches connected to genera-tors.Direct observation of line faults is not available;it is only possi-ble to detect,for each line,whether it has been de-connected due to a fault to lines below in the electricityflow.The PSR problem consists in feeding all possible lines within a given set,for a set of possible fault configurations,in spite of the limited sensing available.We con-sider a set of16PSR problems over a network featuring3generators, 5switches and6lines.The problems have0to3unreliable switches, and four different set of possible faults configurations.We present the problems sorted by growing complexity.Finally,we consider a robot navigation problem where the robot is in a cilindric tower offloors,and can only move around thefloor he’s in.Inside anyfloor,every tenth room has a writing on the wall which indicates thefloor number,and can be read by the robot.The robot has to reach a given room infloor1;it starts not knowing the floor he’s in,and is uncertain on the room also.We consider towers of increasing height,and100rooms perfloor.The average MBP P times and the MBP times for the three test cases are reported in Fig.4.For the maze and PSR problems consid-ered,a strong solution exist.However,due to the enormous number of contingencies introduced by nondeterminism(the robot slipping in the maze,faults and unreliable switches in the PSR),offline search becomes practically unfeasible for large/complex problem instances, while the on-line progressive search performed by MBP P scales up smoothly,dealing with contingencies as they arise.For the tower problem,the goal can never be reached:the robot might be in the wrongfloor,and in any case it cannot detect precisely in which room it is situated.Once more,MBP P is able to effectively exploit the knowledge gathered by observations performed during executions, restricting the search and promptly discovering after a few runs that no progressive plan exists.The time spent to achieve this is basically independent of the domain size,since anyway after a few short runs, the robot gets to know thefloor it is in,achieving the better knowl-edge possible given the situation.On the opposite,the larger the do-main,the more MBPfinds it complex to discover the absence of a solution.We experimentally verified that when the progressiveness check is disabled,MBP P fails discovering that no solution exists,and0.11101005101520253035C P U S e a r c h t i m e (s e c )size Slippery MazeMBPp: avg MBP 0.11101002468101214C P U S e a r c h t i m e (s e c )problem instance PSR network MBPp: avg MBP 0.1110100510152025C P U S e a r c h t i m e (s e c )height Tower MBPp: avg MBP does not terminate.In this setting,termination can only be achievedby adding a termination criteria that constraints the minimum lengthof the partial plans searched at each run to (at least)the number ofrooms in a floor.On top of being problem specific,this strongly de-grades the search,since it forces visiting off-line a vast portion of thesearch space at each planning run.The experimental evaluation shows that the approach is verypromising:despite being fully general,the interleaved approach byMBP P scales up much better,and is capable of efficiently dealingwith very complex problem instances,that cannot be dealt by thestate-of-the art planner MBP.The last experiment confirms the ca-pability of MBP P to terminate when there is no more chance to finda strong solution to the planning problem.Regarding plan quality,weremark that neither MBP nor MBP P grant optimal (average)lengthof execution.Since both planners rely on the same node extractionheuristics,and progressiveness check only prevents looping behav-iors,we do not expect relevant differences;however,a more accurateevaluation of this aspect is in our agenda.6Conclusions and Related workThe idea of interleaving planning and execution is certainly not newand has been around for a long time,see,e.g.,[6].In particular,some approaches exist that address the problem of interleaving plan-ning considering nondeterministic domains and partial observability,defined similarly to here.Amongst those works,the most notableare [12,13,11],which propose different techniques based on real-time heuristic search.These algorithms rely on the existence of dis-tance heuristics for the search space;in some case [11]they have thenice property that they can amortize learning over several planningepisodes.These algorithms only guarantee termination (and reachingthe goal)under the assumption that the domain is safely explorable,i.e.that the plan executor can never find himself in a position where itmay be impossible to reach the goal due to unlucky nondeterministicaction outcomes.By providing the notion of progressiveness,we areable to guarantee termination in a more unconstrained setting,whereassumptions on safe explorability and admissible heuristic evalua-tions are not involved.Other approaches,still based on real-time heuristic search,addressthe problem of planning in stochastic domains with probability distri-butions on action outcomes,like in POMDP (see,e.g.,[3,10,4,5]).Our technique is very different,and relies on symbolic model check-ing techniques and on an efficient,BDD-based representation.The maze domain presented in the experiments is inspired by thework by Koenig ([11]),where it has been tested extensively in theproblem of robot navigation and localization.However,the experi-mental domain of Section 5is much harder than the one used in [11],which assumes that there is no uncertainty in actuation and sensing.It would be interesting an in depth experimental comparison in dif-ferent domains of the two different approaches.Somehow related to our work,even if very different in scope and objective,are the works that propose architectures for interleaving planning and execution,reactive planning and continuous planning,see,e.g.,[14].Among them,CIRCA [8,7]is an architecture for real-time planning and execution where model checking with timed au-tomata is used to verify that generated plans meet timing constraints.REFERENCES [1]P.Bertoli,A.Cimatti,and M.Roveri,‘Conditional planning under par-tial observability as heuristic-symbolic search in belief space’,in Proc.of ECP’01,(2001).[2]P.Bertoli,A.Cimatti,M.Roveri,and P.Traverso,‘Planning in non-deterministic domains under partial observability via symbolic model checking’,in Proc.of IJCAI 2001,(2001).[3] B.Bonet and H.Geffner,‘Planning with incomplete information as heuristic search in belief space’,in Proc.of AIPS’00,(2000).[4] A.Cassandra,L.Kaelbling,and M.Littman,‘Acting optimally in par-tially observable stochastic domains’,in Proc.of AAAI94,(1994).[5]T.Dean,L.Kaelbling,J.Kirman,and A.Nicholson,‘Planning Under Time Constraints in Stochastic Domains’,Artificial Intelligence ,76(1-2),35–74,(1995).[6]M.Genereseth and I.Nourbakhsh,‘Time-saving tips for problem solv-ing with incomplete information’,in Proc.of AAAI93,(1993).[7]R.P.Goldman,D.J.Musliner,and M.J.Pelican,‘Using model check-ing to plan hard real-time controllers’,in Proc.of the AIPS2k Workshop on Model-Theoretic Approaches to Planning ,(2000).[8]R.P.Goldman,M.Pelican,and D.J.Musliner.Hard Real-time Mode Logic Synthesis for Hybrid Control:A CIRCA-based approach.Work-ing notes of the 1999AAAI Spring Symposium on Hybrid Control.[9]R.M.Jensen,M.M.Veloso,and M.H.Bowling,‘OBDD-based op-timistic and strong cyclic adversarial planning’,in Proc.of ECP’01,(2001).[10]L.Kaelbling,M.Littman,and A.Cassandra,‘Planning and acting in partially observale domains’,Artificial Intelligence ,1-2(101),99–134,(1998).[11]S.Koenig,‘Minimax real-time heuristic search’,Artificial Intelligence ,129(1),165–197,(2001).[12]S.Koenig and R.Simmons,‘Real-time search in non-deterministic do-mains’,in Proceedings of IJCAI-95,(1995).[13]S.Koenig and R.Simmons,‘Solving robot navigation problems with initial pose uncertainty using real-time heuristic search’,in Proc of AIPS-98,(1998).[14]K.L.Myers,‘Towards a framework for continuous planning and exe-cution’,in Proc.of the AAAI Fall Symposium on Distributed Continual Planning ,(1998).[15]L.Pryor and G.Collins,‘Planning for Contingency:a Decision Based Approach’,J.of Artificial Intelligence Research ,4,81–120,(1996).[16]S.Thiebaux and M.O.Cordier,‘Supply Restoration in Power Distribu-tion Systems:a Benchmark for Planning Under Uncertainty’,in Proc.of ECP-01,(2001).[17]S.Thiebaux and J.Hertzberg,‘A semi-reactive planner based on a pos-sible models action formalization’,in Proc.of AIPS-92,(1992).[18] D.S.Weld,C.R.Anderson,and D.E.Smith,‘Extending graphplan to handle uncertainty and sensing actions’,in Proc.of AAAI-98),(1998).。