改进灰狼算法求解柔性作业车间插单问题

Solving Flexible Job-Shop Insertion Problem by Improved Grey Wolf Optimization

为解决柔性作业车间订单插入问题,课题组建立以最大完工时间、总能耗、总延迟时间和设备负载为目标函数的重调度模型,并对4个目标函数进行线性加权和法归一化,提出一种改进的灰狼算法(improved grey wolfoptimization,IGWO)作为全局优化算法。首先,采用离散整数编码方式以及混合初始化规则生成高质量初始种群;其次,引入非线性收敛因子,以平衡算法的全局搜索性和局部开发性,同时引入内部狼群个体进化和外来狼群入侵机制来增加算法的搜索广度,避免算法陷入“早熟”;针对订单插入点后未加工的工序,采用事件驱动策略重新调度;最后,通过生产实例验证。结果表明IGWO在求解柔性作业车间订单插入重调度问题上具有有效性和稳定性。

In order to solve the order insertion problem in flexible job-shop,a rescheduling model with maximum completion time,total energy consumption,total delay time and equipment load as objective functions was established,and the four objective functions were normalized using the linear weighted sum method,and an improved grey wolf optimization(IGWO)was proposed as a global optimization algorithm.Firstly,high-quality initial population was generate using discrete integer encoding and mixed initialization rules.The nonlinear convergence factor was introduced to balance the global search and local exploitation of the algorithm.Simultaneously internal wolf pack individual evolution and external wolf pack invasion mechanisms was introduced to increase the search breadth of the algorithm and avoid the algorithm falling into"precocity".An event driven strategy was adopted to reschedule the unprocessed processes after the order insertion point.Finally,verified through production examples.The results indicate that IGWO is effective and stable in solving the flexible job-shop order insertion rescheduling problem.