基于免疫遗传算法的多模式资源受限养护工程调度问题研究

Multi-Mode Resource-Constrained Maintenance Project Scheduling Based on Immune Genetic Algorithm

公路养护工程项目调度通常会受现场资源与施工工期的限制,针对基本遗传算法不能有效解决多模式资源受限养护工程调度问题(MRCPSP),将生物免疫系统中抗体多样性的维持机制应用到养护工程调度管理,构建一种采用双链表结构编码方式(随机优先权值链表和模式链表)、串行调度生成机制解码方式的免疫遗传算法,通过选择、交叉、变异和记忆库更新算子进行全局搜索确定较优项目调度方案.运用PSPLIB问题库中的基准问题对该算法进行验证,以证明该算法的可行性和有效性.

Highway maintenance project scheduling is usually limited by site resources and construction period.Aiming at the fact that the basic genetic algorithm can’t effectively solve the multi-mode resource-constrained maintenance project scheduling problem(MRCPSP),the maintenance mechanism of antibody diversity in biological immune system is applied to maintenance project scheduling management,and an immune genetic algorithm with double linked list structure coding mode(random priority value linked list and mode linked list)and decoding mode of serial scheduling generation mechanism is constructed.Global search based on operators of selection,crossover,mutation and update memory is used to determine the optimal project scheduling program.Finally,the feasibility and effectiveness of this algorithm are tested based on benchmark problems in PSPLIB problem library.