基于惩罚项的热连轧轧制规程多目标函数优化

Multi-objective function optimization of rolling schedule in tandem hot mill based on penalty item

为了解决传统热连轧负荷分配优化普遍采用约束方法进行求解时求解步骤繁琐、不容易获得最优解的问题,提出了一种惩罚函数算法.通过引入函数惩罚项,将多目标函数约束求解问题转化为无约束求解问题.建立了带有惩罚项的轧制力、板形、功率和温度的单目标函数,进而采用线性加权求和法建立了综合多目标函数.利用Nelder和Mead单纯形解法则进行求解,简化了求解的步骤,最终得到了兼顾轧制力平衡、板形最优及温度合理的负荷分配方案.实际应用效果表明,与传统负荷分配方式相比,基于多目标优化得到的轧制规程更符合实际生产要求,具有良好的应用前景.

In order to solve the problem that the solving steps are tedious and the optimum solution is hard to be obtained when the the traditional load distribution optimization in hot tandem mill is usually performed with the constraint condition method, a penalty function algorithm was proposed. Through introducing the penalty item of function, the constraint solution problem of multi-objective function was converted into the unconstraint solution problem. The single-objective functions for the rolling force, flatness, power and temperature with the penalty items were established, and thus the comprehensive multi-objective function was established with the linear weighted summation method. The solution was performed with Nelder and Mead simplex algorithms, which could simplify the solution steps. Finally, the load distribution scheme with the balanced rolling force, optimum flatness and reasonable temperature was attained. The actual application results show that compared with the traditional load distribution method, the rolling schedule based on the multi-objective optimization is more suitable for the actual production requirements, and has good application prospect.