多种群果蝇优化算法求解自动化仓库拣选作业调度问题

Multiple Population Fruit Fly Optimization Algorithm for Automatic Warehouse Order Picking Operation Scheduling Problem

针对自动化仓库的拣选作业调度问题,提出了一种多种群果蝇优化算法。采用随机键编码方式,利用味道浓度判定值的大小次序来映射调度解。通过同时学习子种群的局部最优和全局最优个体,实现对果蝇个体的更新计算。为了避免陷入局部最优,采用了一种果蝇个体变异机制。计算结果显示,多种群果蝇优化算法在计算精度和收敛效率方面要好于基本果蝇优化算法,并且搜索过程能够有效跳出局部最优。

A multiple population fruit fly optimization algorithm was proposed for automatic warehouse order picking operation scheduling problem. A coding method of random key was adopted, and the sequence of the smell concentration judgment value was mapped to the schedule solution. The fruit individuals were calculated and updated by simultaneously learn from both the local optimum of the offspring population and the whole optimum of overall populations in the itera- tion. A mutation method was employed to jump away from the local optimum for the fruit individual. The computational results show that the multiple-population fruit fly optimization algorithm has better calculation precision and convergence efficiency than the basic fruit fly optimization algorithm, and it can effectively avoid falling into the local optimum.