论文部分内容阅读
铝热连轧生产中,合理的规程制定,能够改善产品质量,提高产量。利用最小二乘支持向量机对轧制力模型建模,并以降低能耗和预防打滑为目标,对河南某铝热连轧精轧机组进行规程优化。针对多目标分布估计算法(MOEDA)存在收敛速度慢和精度低的问题,采用改进的差分进化算法(DE)与之结合。改进了差分进化算法的差向量和最优粒子选取方法,并设计了合理的算法切换机制,该组合算法的收敛性和分布性相对原算法有明显改善。在河南某铝热连轧精轧机组的规程优化中,该方法能够获得收敛性和分布性较好的近似Pareto前沿,求解精度和算法可靠性优于传统方法。
Aluminum hot-rolling production, the development of a rational order, to improve product quality and increase production. The rolling force model was modeled by least square support vector machine, and the schedule optimization of Henan aluminum tandem rolling mill was carried out with the goal of reducing energy consumption and preventing slip. The multi-objective distribution estimation algorithm (MOEDA) has the problems of slow convergence speed and low accuracy, and it is combined with improved differential evolution algorithm (DE). The difference vector and the optimal particle selection method of the differential evolution algorithm are improved, and a reasonable algorithm switching mechanism is designed. The convergence and distribution of the combined algorithm are obviously improved compared with the original algorithm. In the process optimization of a hot strip mill in Henan, this method can obtain the approximate Pareto front with good convergence and distribution, and its accuracy and reliability are superior to those of traditional methods.