资源受限下维修任务网的调度问题研究

发布时间：2018-01-16 05:29

本文关键词：资源受限下维修任务网的调度问题研究　出处：《北京交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：装备维修保障系统是保证装备保持或恢复到规定状态的技术管理活动集合。合理地调度维修保障活动可以帮助企业快速解决问题或排除故障,避免事故的发生,同时还可以保证装备能够按时完成规定任务,对企业保持工作效率、提升效益有着重要的作用。优化调度维修保障活动中所需资源一直是维修保障系统中的关键性问题。维修任务网的调度问题属于资源受限项目调度问题。但是由于实际维修环境的复杂性,所以经典资源受限项目调度问题的求解算法并不完全适用。本文基于某企业的现实需求,在经典资源受限项目调度模型的基础上引入了工位、人员等一系列新的约束条件,设定最小化最大完工时间为求解目标,设计并实现了一个资源约束下维修任务网的调度模型,用于解决实际调度问题。由于精确算法对大规模问题无法在可接受时间内求解,而启发式算法可以在较短的时间内求得问题的一个较优解,所以本文采用启发式算法对维修资源受限条件下的调度优化问题进行求解。本文首先使用基于优先规则的构造性启发式算法,结合串行进度生成机制对问题模型进行求解,设计了四类优先规则用于选择工位、工序、资源和人员。为了进一步优化工位、维修人员等资源,本文运用最大最小蚂蚁系统,通过对信息素的更新加以限制从而实现对工位的选择的优化,其次我们研究了遗传算法和粒子群算法,并针对本文问题模型,提出了一种基于粒子群和遗传算法的混合优化算法,将遗传操作因子(选择、交叉和变异)应用到粒子更新规则上,实现对工位和维修人员两种资源同时进行优化。通过实验验证了两种优化算法的优化效果,并且通过对比及在仿真软件中的评估,发现基于粒子群和遗传算法的混合优化算法具有更优的优化效果。本文提出的资源约束下维修任务网的调度模型是合理的,设计的求解及优化算法能够得到正确且较优的结果,对改进维修保障作业有一定的指导作用。
[Abstract]:Equipment maintenance support system is a set of technical management activities to ensure that the equipment is maintained or restored to a specified state. Reasonable dispatch of maintenance support activities can help enterprises solve problems or troubleshoot quickly and avoid accidents. At the same time, it can also ensure that the equipment can complete the prescribed tasks on time, and maintain the efficiency of the enterprise. The resource requirement in optimal scheduling and maintenance support activities is always the key problem in maintenance support system. The scheduling problem of maintenance task network belongs to resource constrained project scheduling problem. The complexity of the actual maintenance environment. Therefore, the classical resource-constrained project scheduling algorithm is not fully applicable. Based on the actual needs of a certain enterprise, the classical resource-constrained project scheduling model based on the introduction of work station. A series of new constraints, such as personnel, set the minimum maximum completion time as the target, and designed and implemented a scheduling model of the maintenance task network under resource constraints. Because the exact algorithm can not solve the large-scale problem in acceptable time, the heuristic algorithm can find a better solution in a short time. Therefore, this paper uses heuristic algorithm to solve the scheduling optimization problem under the condition of limited maintenance resources. Firstly, this paper uses a constructive heuristic algorithm based on priority rules. Combined with the serial schedule generation mechanism to solve the problem model, four kinds of priority rules are designed to select the work, process, resources and personnel. In order to further optimize the work station, maintenance personnel and other resources. In this paper, the maximum and minimum ant system is used to limit the update of pheromone to optimize the work site selection. Secondly, we study the genetic algorithm and particle swarm optimization algorithm, and aim at the model of this paper. A hybrid optimization algorithm based on particle swarm optimization (PSO) and genetic algorithm (GA) is proposed, in which genetic operational factors (selection, crossover and mutation) are applied to particle update rules. The optimization results of the two optimization algorithms are verified by experiments, and the results are compared and evaluated in the simulation software. It is found that the hybrid optimization algorithm based on particle swarm optimization and genetic algorithm has better optimization effect. The scheduling model of maintenance task network under resource constraints proposed in this paper is reasonable. The solution and optimization algorithm can get the correct and better results, which can be used to improve the maintenance support operation.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】