改进粒子群算法求解生产计划与柔性作业车间调度集成问题

Improved Particle Swarm Optimization to Solve the Integration Problem of Production Planning and Flexible Job Shop Scheduling

为解决机加工企业制定的生产计划与车间调度方案不兼容的问题,建立以最小化最大完工时间、最小化加工成本为目标函数的生产计划与柔性作业车间调度集成模型。提出一种改进粒子群算法(IPSO)作为全局优化算法。在传统粒子群优化算法(PSO)的基础上,引入遗传算子交叉的方式改进群体进化,同时设计随机性的边界变异,提高种群多样性,避免局部最优,学习因子及惯性权重采用幂函数动态变化,增强其搜索能力,更快收敛。最后通过生产实例,验证了IPSO在解决生产计划与车间调度集成问题上的可行性。同时将PSO、灰狼优化算法(GWO)和遗传算法(GA)作为对比算法,在15个Brandimarte基本算例上开展实验,得到的结果均优于其他算法,证明了IPSO求解柔性作业车间调度问题时的有效性和优越性。

In order to solve the problem of incompatibility between the production plan formulated by the machining enterprise and the workshop scheduling scheme,the production planning and flexible operation workshop scheduling integration model with the objective function of minimizing the maximum completion time and minimizing the processing cost vas established.An improved particle swarm optimization(IPSO)algorithm was proposed as a global optimization algorithm.Based on the traditional particle swarm optimization(PSO),the genetic operator crossover method was introduced to improve the population evolution,and the random boundary variation was designed to improve the population diversity and avoid local optimization.The power function was use for the learning factor and inertia weight to dynamically change to enhance their search ability and converge faster.Finally,through production examples,the feasibility of IPSO in solving the integration problem of production planning and workshop scheduling was verified.At the same time,the PSO,grey wolf optimizer(GWO)and genetic algorithm(GA)were used as comparative algorithms,and experiments were carried out on 15 basic examples of Brandimarte.The results obtained are better than other algorithms,which proves the effectiveness and superiority of IPSO in solving flexible job shop scheduling problems.