基于变邻域遗传算法的离散制造物料配送调度优化

Distribution Scheduling Optimization of Discrete Manufacturing Materials Based on Variable Neighborhood Genetic Algorithms

针对含有AGV搬运的离散制造车间柔性作业调度问题(FJSP),以工件从立体库出库到入库的最大完工时间为优化目标,设计了基于混合算法的静态求解方法。首先,根据实际车间中AGV搬运与生产加工的特点建立了柔性车间调度的数学模型,采用了一种基于工件、机器、AGV序列的三层编码机制;然后,使用启发式与随机式结合的方式改进了初始编码的生成,采用自适应权重因子动态调整交叉和变异概率,在算法的执行过程中,引入了基于移动和交换工序策略的变邻域搜索(VNS)机制;最后使用MATLAB编程进行仿真测试,并给出了算例最优解的甘特图。通过与其他文献算法结果对比,验证了变邻域遗传算法能有效处理含AGV搬运的物料配送问题。

Aiming at the flexible job scheduling problem(FJSP)of discrete manufacturing workshop with AGV handling,a static solution method based on hybrid algorithm was designed to optimize the maximum completion time of the workpiece from the three-dimensional warehouse to the warehouse.Firstly,according to the characteristics of AGV handling and production and processing in the actual workshop,the mathematical model of flexible workshop scheduling is established,and a three-layer coding mechanism of workpiece,machine and AGV sequence is adopted.Then,the generation of initial coding is improved by combining heuristic with randomness.The crossover and mutation probability are dynamically adjusted by adaptive weight factors.In the process of algorithm execution,the variable neighborhood search(VNS)mechanism based on mobile and switching process strategy is introduced.Finally,MATLAB programming is used for simulation test,and the Gantt chart of the optimal solution is given.By comparing with the results of other algorithms in literature,it is verified that VGA can effectively deal with the problem of material distribution containing AGV.