系统工程与电子技术

• 系统工程 • 上一篇    下一篇

考虑不确定因素影响的保障任务调度算法

曾斌, 姚路, 胡炜, 杨光   

  1. (海军工程大学管理工程系信息管理研究室, 湖北 武汉 430033)
  • 出版日期:2016-02-24 发布日期:2010-01-03

Scheduling algorithm for maintenance tasks under uncertainty

ZENG Bin, YAO Lu, HU Wei, YANG Guang   

  1. ZENG Bin, YAO Lu, HU Wei, YANG Guang
  • Online:2016-02-24 Published:2010-01-03

摘要:

针对装备保障任务的优化调度问题,首先进行静态建模,提出了一个新的数学规划模型,能够有效描述保障单元的力量配置及与保障对象的指派关系等复杂约束条件;随后实施动态建模,利用混合Petri网把数学模型转换为流程模型,不仅构建了变迁激发规则以表达静态数学模型的约束条件,而且设计了新的子网结构模拟突发事件及协同保障的动态过程。为了求解该规划模型,提出了一个基于退火进化的优化调度算法,该算法首先计算保障单元的分配问题,随后搜索资源分配的优先级列表生成保障任务的工作顺序,算法中利用Petri网过程模型计算不确定条件下的目标函数值。仿真实验表明算法能在较快的收敛速度下提高保障单元的利用率。

Abstract:

It is important to schedule the limited maintenance resources efficiently for equipment supportability. In the static modeling stage, a new mathematic description model is presented considering complex constraints such as power configuration and assignments relationships of maintenance units. In the dynamic stage, a Petri nets model is established to translate the static mathematic model into the dynamic workflow model. In the Petri nets model, a set of firing rules are proposed to implement the constraints of the mathematic model, and the new net structures are designed to simulate the workflow of uncertainties and cooperative process. In order to solve the planning model, an optimization algorithm based on the annealing evolution algorithm is presented in which a simulated annealing algorithm is adopted to solve the maintenance unit assignment problem and a search algorithm is used to generate the schedule results according to the Petri nets model. The simulation results indicates that the algorithm can improve the unit utilization with a high evolution speed.