面向柔性作业车间调度问题的改进博弈粒子群算法

An Improved Gaming Particle Swarm Optimization Algorithm for Flexible Job-shop Scheduling Problems

摘要:

对柔性作业车间调度问题的研究可以令实际生产加工过程更加贴合当今人们对商品个性化和定制化方面的需求。在对柔性作业车间调度问题中的多个性能评价指标进行研究后，巧妙利用它们间的矛盾点，在自创的问题编、解码方案的基础之上，建立了博弈解集，并对传统粒子群算法的寻优机制进行改进，提出了改进博弈粒子群算法。运用该算法对一组标准问题调度算例进行求解， 验证了该算法良好的求解性能。同时，通过与其他粒子群算法结果和耗时等的比对显示该算法可以更有效地求解以最小化最大完工时间作为唯一优化目标的柔性作业车间调度问题。

Abstract:

The research on flexible job-shop scheduling problems （FJSP） can help the production in practical to meet the ascending demands of personalization and customization from special customers. On the basis of a full study of scheduling criteria on FJSP， the paper proposes a gaming particle swarm optimization algorithm gaming PSO） with novel encoding and decoding schemes. In comparison with the traditional PSO， the communication mechanism of the proposed PSO is improved by a gaming solution set， which takes advantages of the contradictions among the scheduling criteria. Finally，based on a test of the standard benchmarks and a comparative study of the test results with those by other improved PSOs， the proposed gaming PSO proves to be effective in minimizing the maximum completion time of FJSP.