随机环境下多目标资源受限项目调度均衡优化研究

发布时间：2018-04-16 16:31

本文选题：资源受限项目调度 + 随机环境　；参考：《南京理工大学》2017年硕士论文

【摘要】：资源受限的项目调度问题(RCPSP)及均衡问题在项目管理中较为常见,本文在传统RCPSP的工期和成本目标基础上,增加鲁棒性和资源均衡目标,结合RCPSP的随机性,进行RCPSP均衡优化。本文均衡研究包括两个方面:一是时间成本均衡问题(TCTP),二是时间资源均衡问题(TRTP),并通过数例描述和验证了改进的均衡方案可以进一步优化随机RCPSP。本文首先引入随机环境下的RCPSP,假设活动时间及资源服从不同分布,对资源失效和故障造成的活动启动时间的偏离进行修正,并建立成本及鲁棒性的双目标模型,通过时间缓冲策略(STC)增加调度计划的鲁棒性。而上述鲁棒优化会延长工期,因此本文进一步研究了 TCTP。该部分研究基于传统TCTP,构建了一个非线性组合优化模型,以项目工期,成本和鲁棒性为目标,并在解的过程中加入赶工判定,以六项改进赶工原则和三个均衡点为判定方法,通过赶工判定过程和STC来获得具有鲁棒性的调度方案,并用数例证明考虑赶工的TCTP解决方法更具有效性。TCTP研究通常假设单位资源使用确定,而资源使用不确定时,本文基于传统TRTP进一步考虑柔性资源配置。该部分在TCTP模型上增加资源均衡目标,并比较了一系列鲁棒性替代措施,从而获得普适性的鲁棒性指标,运用基于优先级的启发式和资源分配启发式结合上述算法,证明柔性资源配置下的TRTP能够更有效的解决RCPSP中的权衡问题并提高调度方案的鲁棒性。
[Abstract]:Resource constrained project scheduling problem (RCPSP) and equilibrium problem are common in project management. Based on the traditional RCPSP target of time limit and cost, this paper increases the robustness and resource balance goal, and combines the randomness of RCPSP to optimize the RCPSP equilibrium.The study of equilibrium in this paper includes two aspects: one is the time cost equilibrium problem, the other is the time resource equilibrium problem. Several examples are given to illustrate and verify that the improved equilibrium scheme can further optimize the stochastic RCPSPs.In this paper, the RCPSPs in random environment are introduced firstly. Assuming that the activity time and service are distributed differently, the deviation of activity startup time caused by resource failure and fault is corrected, and the cost and robustness model is established.STC (time buffer Policy) is used to increase the robustness of scheduling plan.However, the above robust optimization can prolong the duration of the project, so this paper further studies the TCTP.In this part, based on the traditional TCTP, a nonlinear combinatorial optimization model is constructed, which aims at the project duration, cost and robustness, and adds the rush decision in the process of solution, taking six improved rush principles and three equilibrium points as the judgment methods.A robust scheduling scheme is obtained by means of rush decision process and STC. Several examples are used to prove that the TCTP solution considering rush work is more effective. TCTP studies usually assume that unit resource usage is determined, while resource use is uncertain.In this paper, flexible resource allocation is further considered based on traditional TRTP.In this part, the goal of resource equilibrium is added to the TCTP model, and a series of robust alternatives are compared to obtain the robustness index of universality. The priority-based heuristics and resource allocation heuristics are used to combine the above algorithms.It is proved that TRTP in flexible resource allocation can solve the tradeoff problem in RCPSP more effectively and improve the robustness of scheduling scheme.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F273

【参考文献】