【作者】 张敏；

【导师】 陈荣秋；

【作者基本信息】 华中科技大学 ， 管理科学与工程， 2011， 博士

【摘要】 在经济全球化、竞争白热化的市场环境下,项目管理理论和方法在各个领域的应用日益广泛。不确定因素的存在给传统的项目管理方法带来极大冲击。在众多的不确定因素中,行为不确定性正日益引起学术界的重视,行为运作管理已成为一门新兴学科。Goldratt首次考虑了项目计划和执行过程中的行为不确定性,他提出的关键链项目管理对传统的关键路径法有实质上的改进。但需要进一步挖掘行为不确定因素产生的根源,以便正确预测和积极引导项目决策者的行为。由于拖延行为直接影响项目绩效,而拖延发生的频度和程度都存在不确定性,因此对行为不确定性的分析主要集中于“拖延”这一典型行为展开。本项研究在对行为运作这一新兴研究领域的文献进行梳理的基础上,通过实验测试和案例研究验证心理因素带来的行为偏差,采用解释型研究分析产生行为偏差的根源；通过规范型研究建立模型、提出解决问题的对策并运用实证调查和情景实验来验证理论模型的有效性。不确定性无法完全消除,只能在一定范围内规避并通过管理机制的设计加以利用和内化,按照这一思路,本文在关键链项目管理背景下,以个体决策者为分析对象,对其在项目计划和执行过程中呈现出的不确定行为进行研究,主要取得了以下研究成果：(1)运用启发式原则分析项目进度管理中的延迟现象,论证了“前景理论”应用于项目进度管理过程的可行性；在此基础上分析项目决策者的时间决策过程并对单任务项目工期决策行为进行建模；通过实例分析、数值仿真及敏感性分析验证模型的有效性,从一定程度上解决了行为变量难以量化的问题,验证了通过适当的情景设计正确引导项目决策者行为的可行性。(2)以时间偏好不一致和自我约束理论为基础,通过情景实验观察项目进度计划和执行过程中的拖延行为,研究不同类型决策者在不同情景下的行为差异,统计分析验证了项目决策者在进度计划制定和执行过程中都会存在不同程度的拖延行为。通过激励强度、频度的设计以及进度计划的弹性设计能够有效引导项目决策者的行为；通过学习和不断积累经验加强职业素养,可以提高自我约束能力,有效减少拖延行为。(3)以“理性人”假设为基础,通过实物期权工具对行为不确定现象背后的隐含价值进行识别。建立延迟期权和转换期权价值模型,通过数值分析与实证检验验证了期权分析工具的可行性,并得到理性项目决策者的行为控制策略,为项目管理激励机制的设计提供依据。(4)在分析关键链缓冲机制局限性与期权分析工具可行性的基础上提出了将行为偏差内化的稳健进度计划优化思路,建立了不确定情景下的项目缓冲前置分配模型并设计了求解算法,并考虑了时间偏好不一致与心理账户对决策模型的影响。该模型以动态的多阶段决策环境为背景,从柔性管理的视角研究关键链项目缓冲的分配和控制问题,将缓冲放置于关键任务之前,以充分挖掘行为不确定性可能带来的机遇。(5)综合考虑成本目标和期权价值目标建立了基于扩展净现值的项目缓冲多目标优化模型,对基于实物期权的多条关键链识别策略和项目群管理中该策略的运用方法进行了说明。在借鉴关键链项目管理理念的基础上,形成了项目进度管理中针对行为不确定性的较为完整的柔性管理框架。本文主要针对项目计划与执行过程中存在的“学生综合症”和“帕金森定律”两种行为展开分析,由于行为不确定性的表象呈现出多元化和异质性,本文暂时屏蔽了文化价值这一社会维度,对项目进度管理中的行为不确定性研究只是做了初始的尝试。更多还原

【Abstract】 Uncertainty features are today’s market environment with economic globalization and fierce competition. How to manage the uncertainty in the corporate world dominated and organized by projects becomes huge challenge to traditional project management. Among a variety of uncertainty factors, behavior uncertainty caused by cognitive bias and risk aversion has come to draw academic attention and behavior OM gradually frames its research areas. Critical Chain Project Management firstly proposed by Goldratt takes behavior uncertainties in consideration in project scheduling problems has improved the traditional Critical Path method and pointed out that cause of behavior uncertainty need to further be explored. This dissertation analyzed and researched management and social service of behavior uncertainties in critical chain project management. It has important theoretical and practical significance on enriching research content of related subjects and solving practical problems in project management.Behavior of procrastination has direct impact on project operation performance, so varied frequency and extent of procrastination become one focused area of research in uncertainty management in projects. This dissertation addresses the problem of the behavior operation by reviewing literatures in this emerging area, analyzes the cognitive bias through experiment description and behavior interpretation, and describes the real phenomenon by normative research models to test efficiency of solutions to the problem. As uncertainties can’t be completely eliminated, we can only use and internalize them within a certain range through certain mechanism design. The dissertation treats individual decision maker as the analysis object and gets results as follows:Firstly, the dissertation confirms the feasibility of Prospect Theory in project management through heuristic analysis of procrastination phenomenon. Within the framework of Prospect Theory, a theoretical model was formulated for planner’s time choice behavior within a critical activity as follows:Firstly, the planner’s subjective perception function of cost was proposed, then we defined the reference point for decision and give the calculation method of prospect value. This dissertation demonstrates that the Prospect Theory could provide a more accurate description of planner’s behavior of time decision under the conditions of uncertainty than the Expect Utility Theory. To a certain extent, this dissertation solves the problem that the behavior variables are difficult to quantify. Secondly, this dissertation analyzes the impact of behavior variables on real procrastination scenarios. The findings indicate that the degrees of procrastination are various with different control variables in different scenarios. This dissertation demonstrates that appropriate strategies will decrease disadvantage by proper design of incentive intensity and frequency and schedule flexibility. Moreover, accumulated experience through learning will enhance professionalism and help practitioners reduce delays and improve their self-restraint ability effectively.Thirdly, this dissertation uses real options methodology to analyze the value of uncertainty of "student’s syndrome" in project management. As to delay option and switching option, the optimal super-trigger and the optimal lower-trigger of switching option are applied respectively. A practitioner with bounded rationality should choose whether or not to execute the option and maximize utility. The model provides the basis for the design of incentive mechanisms and control strategies.Fourthly, this dissertation analyzes internalization behavior uncertainties based on the limitations of critical chain buffering mechanisms and the feasibility of real options tools. This dissertation models a project buffer allocation in critical chain under uncertainties, incorporates real options concepts to construct an option-based 0-1 stochastic integer programming considering the probability distribution of uncertainty to allocate the provocative project buffer among critical activities. We can execute such activities with time flexibility and optimize the value of the whole critical chain in a proactive approach.Last but not least, this dissertation integrates objective functions of the cost and option value, and proposes a multi-objective optimization model of project buffer allocation method. It takes into account time preference inconsistent and mental account and offers a solution to identify the most valuable critical chain in multiple critical chains and project portfolios. Thus, this dissertation puts forward a framework for flexible scheduling strategy in critical chain project management.This dissertation mainly explored procrastination phenomena in scheduling and controlling period of critical chain project. Coordinating various behavioral dimensions of uncertainties for scheduling and controlling is an enormous engineering and needs more efforts. The purpose of this dissertation is throwing out a minnow to catch a whale and further research will be imperative.更多还原