基于遗传算法的多模式资源受限项目调度问题

Multi-mode resource-constrained project scheduling based on genetic algorithm

针对多模式资源受限项目调度问题,采用问题任务链表和模式链表的双链表结构的编码方式、串行调度生成方案的解码方式对传统遗传算法进行改进,并通过交叉、变异、选择等算子进行全局性概率搜索确定较优方案。针对PSPLIB中的基准问题对该算法进行验证,证明该算法的有效性。

In view of problems in multi-mode resource-constrained project scheduling,encoding method of a double-task list structure（problem task list and pattern list） as well as the decoding method of serial schedule＇s generating scheme were adopted to improve the traditional genetic algorithm.Global probability search was conducted based on operators of crossover,mutation and selection to determine the optimal scheme.Finally,this algorithm was tested based on benchmark problems in PSPLIB for its effectiveness.