基于改进差分进化算法的带钢冷连轧负荷分配  

Load Distribution Optimization on Tandem Cold Rolling Mill Based on Improved Differential Evolution Algorithm

作　　者：杨景明[1] 呼子宇 车海军[1] 孙浩[1] 赵新秋[1] 

机构地区：[1]燕山大学工业计算机控制工程河北省重点实验室,河北秦皇岛066004

出　　处：《钢铁》2014年第12期44-49,共6页Iron and Steel

基　　金：国家自然科学基金钢铁联合基金重点资助项目(U1260203);河北省高等学校创新团队领军人才培育计划资助项目(LJRC013);国家冷轧板带装备及工艺工程技术研究中心开放课题资助项目(2012005);河北省科技支撑计划资助项目(13211817)

摘　　要：为了更加合理地制定冷连轧轧制规程,选取打滑因子、等功率裕度和最小能耗作为目标函数,采用改进的差分进化算法求解Pareto前沿,减少了冗余的寻优代数、缩短了优化时间。以求解出的Pareto前沿为选择池,基于不同的偏好因子选择出不同的负荷分配方法,满足不同工况下的要求。仿真结果表明,该算法可以快速均匀地覆盖Pareto前沿,产生偏重于不同目标的负荷分配规则。优化规程与原规程相比,降低了打滑现象出现的概率,减小了轧制功率。To reasonably make schedules of the load distribution of tandem cold rolling mill, slip factor, power margin and minimum energy consumption were selected as the optimization functions. The improved differential evolution algo-rithm was used to search the Pareto front, which reduces redundant iterations and decreases computation time. Different load distribution strategies based on different objectives were generated from the Pareto front to satisfy the requirements under different operating conditions. The simulation shows the proposed algorithm covers the front fast and evenly and generates different load distribution strategies based on different objectives. Compared with the original schedule, the pro-posed method reduces the probability of slippage and the rolling power.

关 键 词：冷连轧 负荷分配 多目标优化 差分进化算法 终止条件 

分 类 号：TG335.56[金属学及工艺—金属压力加工]