免疫模拟退火算法及其在柔性动态Job Shop中的应用

Immune Simulated Annealing Hybrid Algorithm and Its Application for Flexible Dynamical Job Shop Scheduling

针对车间作业调度问题,在深入分析免疫算法和模拟退火算法的基础上,将两种算法巧妙结合,提出免疫模拟退火算法。该算法引入了免疫记忆、抽取疫苗和接种疫苗等免疫机制,有助于优良个体和基因的保留和利用,提高了算法收敛性,而且其基于概率突跳特性的爬山性能可以避免早熟现象。针对西安航空发动机(集团)有限公司的柔性动态Job Shop,分别用模拟退火算法、免疫算法和免疫模拟退火算法进行了仿真和比较,研究结果表明,免疫模拟退火算法比单一算法性能更优,是求解柔性动态Job Shop问题的有效实用算法。

Immune simulated annealing hybrid algorithm was firstly put forward based on the artful combination of immune algorithm and simulated annealing algorithm after they were thorough analyzed. The immune mechanism including immunity memory, vaccine extraction and inoculation helps to hold the excellent individual and gene, and can enhance algorithm constringency. The mountain climbing based on probability jump of simulated annealing algorithm can avoid prematurity. The flexible dynamic Job Shop simulation experiment of Xi＇ an Aaviation Engine Manufacture Corporation was separately clone by immune algorithm, simulated annealing algorithm and immune simulated annealing hybrid algorithm. At last it is proved that the hybrid algorithm exceeds the single algorithm and is the available effective method for solving flexible dynamic Job Shop problem.