针对柔性作业车间调度复杂的问题,为缩短完工时间,课题组提出了一种基于模拟退火算法(simulated annealing,SA)的改进混合遗传算法.该算法进行搜索时,利用全局、局部和随机(global local random,GLR)种群初始化策略来优化初始解,再基于工序顺序和机器分配2种交叉变异策略扩展种群内优质解的数量,最后在模拟退火操作中加入邻域结构进行优化,在避免传统遗传算法易陷入局部最优的同时,提高了寻优速度.应用课题组提出的改进混合遗传算法对某齿轮轴加工车间加工数据进行仿真对比实验,与传统遗传算法相比,生产效率提高了 8.43％,证明了该改进混合遗传算法的有效性.
Abstract
Aiming at the high complexity of the flexible job shop scheduling problem,the research group proposed an improved hybrid genetic algorithm based on simulated annealing(SA)in order to shorten the completion time.When searching by this algorithm,optimizes the quality of initial solution by GLR population initialization strategy,and the number of high-quality solutions within the population is expanded based on two crossover and mutation strategies focusing on operation sequence and machine allocation.Finally,a neighborhood structure is added to the simulation annealing operation for searching and optimizing,so as to overcome the problem that traditional genetic algorithm is easy to fall into local optimal solution and slow iteration.The improved hybrid genetic algorithm proposed by the research group is used to conduct simulation and comparison experiment for the processing data of a gear shaft processing workshop.Compared with the traditional genetic algorithm,the production efficiency increased by 8.62％,which proves the effectiveness of the improved algorithm.
关键词
生产调度/柔性作业车间/GLR种群初始化策略/遗传算法/模拟退火算法
Key words
production scheduling/flexible work shop/GLR(Global Local Random)population initialization strategy/GA(Genetic Algorithm)/SA(Simulated Annealing)
维普
万方数据