基于双层遗传编码的柔性作业车间自适应重调度研究

Self-adaptively Rescheduling Flexible Job Shop with Double Genetic Coding

针对不确定环境下的柔性作业车间调度,设计了能适应订单异动、操作延时和机器故障等3种常见扰动类型的重调度方法。通过设置可用机器集和操作时间等基本参数,制定各扰动单独或组合发生后基本参数的更新策略,建立了面向3种不确定扰动类型的自适应重调度框架;结合工序码和机器码形成了一种双层染色体编码,该编码能有效表征上述系统参数,实现自适应重调度;利用遗传算法,通过选择算子寻优及交叉变异算子的种群拓展实现全局优化。400个具有不同规模的实例证明了该重调度方法所得方案可信,计算时间可控,能有效应用于生产实际。

Aiming at three kinds of uncertain disturbances appearing frequently within flexible Job Shop,i. e. order changes,operation delays and machine breakdown,a novel rescheduling approach was designed. First,a self--adaptive rescheduling framework was constructed after setting up a set of sys- tem parameters including optional machines and operation times,and formulating strategies about up- dating systematical parameters under single or multiple disturbances. Second, codes for depicting job sequence and machine assignment were combined together in double genetic chromosomes to encode the above system parameters so as to execute self--adaptive rescheduling. Third, through genetic algo- rithm, better solutions were inherited by selection operators and the population of each generation was expanded gradually via crossover and mutation operators, and hence global optimization was achieved effectively. Experimental studies with 400 cases from small to large scale verify that near--optimal re- schedule can be obtained under controllable computational times and thus can be applicable in prac- tice.