有竞争的多目标项目协作计划与调度模型研究

发布时间：2018-04-27 13:58

本文选题：项目计划与调度 + 协作计划　；参考：《扬州大学》2017年硕士论文

【摘要】：网络计划技术是一套科学有效的项目管理方法,被广泛应用于工程项目计划管理中。经典项目计划模型中,部分研究模型考虑了任务可拆分的协作计划,而未引入任务可任意拆分的协作优化机制;当有多个协作伙伴同时竞争外协任务时,如何进行协作伙伴选择是一个复杂的组合优化问题,需用合适的数学模型以及优化方法,为协作伙伴选择提供量化计算方法,提高决策效率;研究网络计划时需同步考虑资源约束,构建计划与调度的协同优化机制;资源受限项目协作计划与调度问题具备天然的多目标特性,但部分研究仍局限于单目标领域,或本质仍将多目标转化为单目标进行优化。因此,研究有竞争的多目标资源受限项目协作计划与调度具有重要的现实意义和理论价值。本文以项目计划与调度为研究对象。首先建立有竞争的多目标资源受限项目协作计划与调度模型;其次建立有竞争的多目标资源受限项目柔性协作计划与调度模型;最后设计了 NSGA-Ⅱ算法以及多目标差分进化算法分别对以上两种模型进行求解。1.建立有竞争的多目标资源受限项目协作计划与调度模型。考虑有协作的计划与调度,针对任务有条件拆分外协的情形,以关键资源为约束,构建了以项目成本和完工周期为优化目标的有竞争的多目标资源受限项目协作计划与调度模型。该模型引入有竞争的协作计划,即有多个协作伙伴同时竞争外协任务,建立选择协作伙伴的新型决策机制,以及基于任务可拆分的协作优化机制,为关键资源调度和协作伙伴选择提供理论支撑,提高资源利用率和选择协作伙伴决策效率。2.建立有竞争的多目标资源受限项目柔性协作计划与调度模型。针对任务可任意拆分外协的情形,构建了以项目成本和完工周期为优化目标的有竞争的多目标资源受限项目柔性协作计划与调度模型,设计了满足任务可任意拆分要求的协作染色体编码方案,针对模型的特殊性,运用多目标差分进化算法对其进行求解。最后,用工程实例进行仿真对比。结果表明,制定基于任务可任意拆分的柔性协作计划,可使协作比例更为合理,求解出的项目成本和完工周期更优。3.设计了基于Pareto最优的多目标求解算法。为实现多目标同步优化,引入Pareto最优机制,提出了基于Pareto的以NSGA-II和多目标差分进化算法为基础的多目标进化算法。针对有竞争的协作计划中任务可拆分外协以及关键资源约束等问题,设计了包含优先级染色体、协作染色体和协作伙伴染色体的三维染色体编码方案。在此基础上,针对柔性协作计划,改进了协作染色体编码方案。同时根据各模型的特殊性,设计了相应的选择、交叉、变异等进化操作,并利用两种算法各自不同的特性求解本文中两个模型。最后,通过模拟一个工程实例,证明了模型及算法的有效性。
[Abstract]:Network planning technology is a scientific and effective project management method, which is widely used in project planning management. In the classic project planning model, some of the research models take into account the collaboration plan that can be disassembled by tasks without introducing a cooperative optimization mechanism that can be arbitrarily split into the task. How to select cooperative partners is a complex combinatorial optimization problem. It is necessary to use appropriate mathematical models and optimization methods to provide a quantitative calculation method for partner selection and improve decision-making efficiency. In the study of network planning, resource constraints should be considered synchronously and cooperative optimization mechanism for planning and scheduling is constructed; resource limited project collaboration plan is used. The problem of scheduling and scheduling has natural multi-objective characteristics, but some research is still limited to the single target field, or the multi target is still transformed into a single target in essence. Therefore, it is of great practical significance and theoretical value to study the cooperative planning and scheduling of a competitive multi target resource limited project. In the first place, a competitive multi target resource constrained project collaboration plan and scheduling model are established. Secondly, a competitive multi-objective resource constrained project flexible collaboration plan and scheduling model is established. Finally, the NSGA- II algorithm and multi-objective differential evolution algorithm are designed to solve the above two models to establish a competitive.1.. A multi objective resource constrained project collaboration plan and scheduling model. Considering the cooperative planning and scheduling, the situation of the conditional separation of the tasks and the constraints of the key resources, a competitive multi-objective resource constrained project collaboration plan and scheduling model is constructed with the project cost and completion cycle as the optimization goal. The competitive collaboration plan, that is, a number of partners compete with the collaborative task, establish a new decision mechanism to select partners, and a cooperative optimization mechanism based on task resolution, to provide theoretical support for key resource scheduling and partner selection, to improve the efficiency of resource utilization and to select the efficiency of cooperative partner decision making.2.. The flexible collaboration plan and scheduling model of multi target resource constrained project. Based on the case that the task can be arbitrarily disassembled, a competitive multi-objective resource constrained project flexible collaboration plan and scheduling model with the objective of the project cost and completion period is constructed, and a cooperative chromosome is designed for the arbitrary task to be disassembled at any level. In the light of the particularity of the model, the multi-objective differential evolution algorithm is used to solve it. Finally, the simulation comparison is made with an engineering example. The results show that a flexible collaboration plan based on the arbitrary resolution of the task can make the cooperative proportion more reasonable, the project cost and the completion period are better.3. design based on the Pareto In order to achieve multi-objective synchronization optimization and introduce the Pareto optimal mechanism, a multi-objective evolutionary algorithm based on the NSGA-II and multi-objective differential evolution algorithm based on Pareto is proposed to achieve multi objective synchronization optimization. The priority is designed to include the problems of task disassembly and key resource constraints in a competitive cooperative plan. A three-dimensional chromosome coding scheme for chromosomes, cooperative chromosomes and partner chromosomes. On this basis, the cooperative chromosome coding scheme is improved for the flexible collaboration plan. At the same time, according to the particularity of each model, the corresponding selection, cross, variation and other evolutionary exercises are designed, and the different characteristics of the two algorithms are used to solve this problem. Two models are presented in the paper. Finally, an engineering example is given to demonstrate the effectiveness of the model and algorithm.

【学位授予单位】：扬州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU72

【相似文献】