为保证热轧生产调度计划的可行性,提高排程的效率,根据热轧生产模式和轧制计划的结构特点,提出了一种车辆路径问题(VRP)模型来建模轧制调度问题,发展了一种混合调度方法(SAMPSO算法)来解决这个问题.试方法利用修正粒子群优化算法的局部和全局搜索能力来寻找全局最优解,利用模拟退火方法来避免陷于局部最优。对某钢厂实际生产数据的仿真结果表明,所提出的模型和算法具有良好的适应性和可行性。
To guarantee the feasibility of hot rolling scheduling plan and improve the efficiency of arranging the scheduling, according to the characteristics of the hot rolling production mode and rolling plan structure, a VRP model was proposed to model the scheduling problem and a hybrid method (SAMPSO) was developed to solve the problem. In the hybrid method, the modified particle swarm optimization (MPSO) algorithm combines local search with global search to search the optimal results and simulated annealing (SA) employs certain probability to avoid becoming trapped in a local optimum. Computational results on practical production data are presented and show that the proposed model and algorithm are feasible and effective.