热连轧精轧负荷分配优化的多目标反向迭代法
Multi-objective inversion iteration algorithm for load distribution optimization in finishing mill group of hot strip rolling

在求解热精轧负荷分配优化的过程中, 针对目标函数中加权系数难以确定的问题, 提出多目标反向迭代法(multi-objective inversion iteration algorithm, MOII), 该法通过数值迭代的方式实现了热精轧机组负荷分配的多目标优化. 首先, 在综合考虑板形、板厚精度和质量的基础上, 设计负荷分配优化的多目标函数; 然后, 基于轧制力能参数与轧机入口和出口板带厚度的单调性关系, 设计MOII算法, 并对MOII算法的合理性及异常情况进行分析, 给出其计算流程; 最后, 基于实际生产数据进行仿真, 结果表明, MOII算法不仅能够实现负荷分配的多目标优化, 而且求解速度快、精度高, 体现出良好的在线应用前景; 与其他方法得到的解相比, MOII算法求得的解有更好的性能.
For solving the problem of load distribution optimization for hot finishing mills, we present the multiobjective inversion iteration (MOII) algorithm to tackle the difficulty in determining the weighting coefficients for objective functions. This algorithm realizes the multi-objectives optimization of load distribution in hot finishing mills through numerical iteration operation. Firstly, multi-objective functions for load distribution optimization are formulated by synthetically considering the accuracy and quality of the strip shape and thickness. Afterwards, MOII algorithm is designed based on the monotonic relationship between the rolling force and energy parameters, as well as the entry and outlet strip thickness of the stand. Moreover, the rationality and the abnormal situation for MOII algorithm are further analyzed, and the calculation procedure is explained. Finally, simulation results based on practical production data show that MOII algorithm is prospective for online application in multi-objective optimization of load distribution with rapidity and precision. Besides, comparing solutions from other load distribution methods, we find the one obtained from the MOII algorithm gives better performances.