【作者】 何涛；

【导师】 罗荣桂；

【作者基本信息】 武汉理工大学 ， 管理科学与工程， 2007， 硕士

【摘要】 在复杂产品协同开发的不同阶段，开发活动需要不同的资源的相互协作。如何将资源尤其是信息技术、人力等柔性资源在复杂产品协同开发过程中进行有效的调度并做出快速的决策，是优化产品开发过程、缩短产品开发时间和降低产品开发成本进而构筑企业动态竞争力的一个至关重要的问题。经典的资源受限的项目调度问题(resource constrained project scheduling problem)没有考虑到资源的柔性，而柔性资源又客观存在与产品开发的各个环节。因此本文主要对具有柔性资源约束的项目调度问题(flexible resource constrained project scheduling problem)展开研究。具有柔性资源约束的调度问题比经典调度问题更为复杂，都是强NP-hard问题。解决问题的核心是模型和算法，有效的调度算法，可以大大提高资源的利用率和生产效益。因而，研究具有柔性资源约束的项目调度问题不仅具有较大的理论意义，而且具有相当高的实用价值。首先，本文讨论了经典的资源受限的项目调度问题及其研究现状，将柔性资源的概念引入该问题，给出了柔性的定义，并基于课题研究成果采用资源—能力矩阵对资源柔性分布进行表示，给出了资源柔性程度的度量方法，在此基础上建立了具有柔性资源约束项目调度问题(FRCPSP)的数学模型其次，介绍了Agent以及Multi-Agent System的相关理论，建立了基于Agent的资源动态调度系统的结构模型，并构造了调度Agent，活动管理Agent，活动Agent，资源管理Agent及资源Agent五类基本组件，以分布式结构作为系统的体系结构框架。同时，为满足Agent的自主决策与交互的需要，还详细分析了它们在项目调度系统中的角色分类，建立了各Agent类的内部模型及行为规则。再次，通过对目前比较流行的合同网协议的分析，将基于启发式规则的调度算法与Agent之间的合同网协议相集成，通过招投标技术，提出了基于多智能体柔性资源约束的项目调度算法模型。为实现Agent之间的协调与合作提供了合理的运行机制和运行策略，以充分发挥Agent的智能性、自治性和协调能力，使系统具有一定的全局优化能力和对动态环境的适应能力。更多还原

【Abstract】 At different stages for complex product coordinate development, the mutual cooperation of different resources are needed by develop activity. How to schedule the resource especially flexible resources such as information technology, manpower effective in the process of complex product coordinate development and make fast decision, which is a vital problem to optimize product development course, shorten product development period and reduce products development cost then construct enterprise dynamic competitiveness. The classical resource constrained project scheduling problem doesn’t consider the flexibility of resource, and flexible resources have each link with product development objectively. So this text mainly launches research to flexible resource constrained project scheduling problem. Flexible resource constrained project scheduling problem is more complex than classical scheduling problems, which is all NP-Hard problem. The core of solving the problem is model and algorithm, effective deployment algorithm, can improve the utilization ratio of resources and productivity effect greatly. Therefore, studying FRCPSP not only has greater theory meanings, but also have quite high practical value.Firstly, this text has discussed the classical resource constrained project scheduling problem and current research situation, and introduces this question of the concept of flexible resources, provide the definition of flexibility, and adopt the resource-skill matrix which is the research result of our subject to express the distribution situation of resources, provide the method of measuring of the flexible intensity of resource, have set up the Mathematical model of FRCPSP on this basis.Secondly, have introduce the relevant theories of Agent and Multi-Agent System, has set up the structural model of dynamic resource schedule system which is based on Agent, and structure schedule Agent, activity manage Agent, activity Agent, resource management Agent and resource Agent five kinds of basic modules, regard distributed structure as the systematic structural frame of system.Thirdly, through analyzing the popular contract net, integrate the contract net between agents with schedule algorithm based on heuristic rule, through bid technology, have proposed the algorithm model of flexibility resource constrained project scheduling problem based on intelligent agent. Provided rational operating mechanism and operation tactics to realize the coordination and cooperation between Agents, so that it can give full play to the intelligent, autonomy and coordinate ability of Agent, make system have the certain overall situation optimize ability and adaptive capacity to dynamic environment.更多还原