交货期惩罚下柔性车间调度多目标Pareto优化研究

Multi-objective Pareto Optimization on Flexible Job-shop Scheduling Problem about Due Punishment

针对传统作业车间调度问题的局限性,结合实际生产过程的特点和约束条件,建立路径柔性的作业车间调度仿真模型。采用连续空间蚁群算法,对柔性车间作业进行多变量、多约束下的调度布局优化设计,在考虑各个机器提前/拖期完工的惩罚值,所有机器上的总负荷、成品合格率和最大设备利用率等性能指标更加合理情况下,为每次迭代产生的邻域解集作为Pareto非支配排序,防止算法操作过程中劣解的产生,提高求解效率。并与自适应免疫算法和交换序列混合粒子群法的优化结果进行对比,该算法可有效改善基本蚁群算法的停滞现象和全局寻优能力差的缺点。目前,该方法已在某机械公司进行示范,在提高加工效率、降低生产成本、减少协作费等方面效果显著。

For the purpose of solving the deficiency of traditional job-shop scheduling problem（JSP）, summarize some production process characteristics and constraints and establish the simulation model of flexible JSP（FJSP）. To minimize the weighed sum of E/T penalties of jobs among the machines, improve the quality of the initial population and accelerate the speed of the algorithm＇s convergence, the continuous ant optimization algorithm is adopted to solve the multi-variable and multi-constraint combinatorial optimization problem. In iterative, the solution set can be used as the Pareto neighborhood non-dominated sorting. Compared with the adaptive immune genetic algorithm and hybrid particle swarm optimization algorithm, the proposed algorithm can rise above efficiently such difficulties of the basic ant colony algorithm as stagnation and poor global search ability. Now the proposed method has been applied to a workshop scheduling on trial. It is shown that its effects on the improvement of processing efficiency, decreasing of production cost, and lowering the cooperation expenses are remarkable.