摘要
针对多阶段非等同并行机模式下的准时化(Just-in-TimeJ,IT)调度问题,采用由遗传算法和禁忌搜索算法混合的禁忌遗传递阶算法进行求解。禁忌遗传递阶算法用禁忌搜索算法对工件最佳加工次序进行搜索,而相应评价值由遗传算法计算得出。遗传算法采用基于阶段机器号的二维矩阵编码,可有效地避免不可行解的产生,同时采用自适应改进提高遗传算法跳出局部最优的能力。实际算例说明禁忌遗传递阶算法计算结果稳定可靠,适合于解决多阶段非等同并行机的JIT调度问题。
Based on JIT scheduling problem in the mode of multi-stage non-identical parallel machines,a tabu-genetic algorithm combined Genetic Algorithm（GA）with Tabu Search Algorithm（TSA）was proposed.This algorithm adopted TSA to search best processing sequence and the fitness value was computed by GA.A two-dimensional matrix coding based on machine number in stages was adopted by GA,and it can avoid the occurrence of infeasible solutions.At the same time,self-adaptive was adopted to improve GA＇s ability of escaping from local optimization.It was revealed that this algorithm has reliable and stable solutions,and fit to solve multi-stage non-identical parallel machines JIT scheduling problem.
出处
《现代制造工程》
CSCD
北大核心
2011年第10期32-36,共5页
Modern Manufacturing Engineering
关键词
准时化
多阶段
禁忌搜索算法
遗传算法
自适应
非等同并行机
Just-in-Time（JIT）
multi-stage
Tabu Search Algorithm（TSA）
Genetic Algorithm（GA）
self-adaptive
non-identical parallel machines
