一种混合嵌套分区算法及其在车间调度中的应用

设计了一种嵌套分区算法框架下的局部搜索算法,即基于最优计算量分配技术的序遗传算法,该算法采用序优化思想保证在有限计算量条件下得到局部最优解,并用遗传算法的进化搜索能力和学习能力对解空间进行搜索.将设计的局部搜索算法与嵌套分区算法相结合提出一种新的混合优化算法,用该混合优化算法求解几个标准的随机车间调度问题,数字仿真的结果表明该混合算法的优化性能好于遗传算法及基于最优计算量分配技术的序优化方法.

Hybrid Nested Partitions Algorithm and Its Application to JSP

A class of order optimization (OO) with optimal computing budget allocation (OCBA) based genetic algorithm (GA) was designed to perform local search in the framework of nested partitions method (NP). The local searching algorithm borrows from the idea of OO to ensure the quality of the design found with a reduction in computation effort and applies the evolutionary searching mechanism and learning capability of GA to effectively perform exploration and exploitation. Proposed hybrid algorithm retains the property of global search and convergence of NP and the powerful local searching capability of OO with OCBA based GA algorithm. The effectiveness of hybrid algorithm was demonstrated by numerical simulation results based on stochastic job-shop scheduling benchmarks and its optimization quality is much better than pure GA and OO with OCBA algorithm.