基于遗传算法的不锈钢冷连轧机轧制规程优化  

Optimization of tandem cold rolling schedule for stainless steel strip based on genetic algorithm

作　　者：张欣 陈树宗 李旭[3] 孙杰[3] 梅瑞斌 王晓宇 ZHANG Xin;CHEN Shuzong;LI Xu;SUN Jie;MEI Ruibin;WANG Xiaoyu(School of Resources and Materials,Northeastern University at Qinhuangdao,Qinhuangdao 066004,Hebei,China;School of Electrical Engineering,Yanshan University,Qinhuangdao 066004,Hebei,China;Stage Key Laboratory of Rolling and Automation,Northeastern University,Shenyang 110819,Liaoning,China)

机构地区：[1]东北大学秦皇岛分校资源与材料学院,河北秦皇岛066004 [2]燕山大学电气工程学院,河北秦皇岛066004 [3]东北大学轧制技术及连轧自动化国家重点实验室,辽宁沈阳110819

出　　处：《钢铁》2023年第7期99-105,共7页Iron and Steel

基　　金：国家重点研发计划资助项目(2022YFB3304800);国家自然科学基金资助项目(U20A20187);“兴辽英才计划”资助项目(XLYC2007087)。

摘　　要：冷轧不锈钢属于高端精品钢材,其生产过程控制系统是最复杂的工业控制系统之一,系统控制精度要求高、动态响应时间短、轧制速度快且工艺参数多。在不锈钢冷轧过程中,合理的轧制规程是保证冷轧不锈钢产品质量和生产效率的基础,其模型精度不仅会影响产品尺寸精度,而且会影响冷轧企业高端产品的研发拓展。为了提高冷轧不锈钢生产过程中的产品质量精度,同时考虑尽可能降低轧制过程总能耗,在深入研究传统负荷分配法和轧制规程计算原理的基础上,针对世界首套7机架不锈钢冷连轧机组,根据冷连轧生产工艺,设计了一种包含目标项和惩罚项的目标函数结构形式,选取电机功率均衡、轧制力均衡、压下分配合理、轧制能耗小、板形良好为目标,构建由5个单目标函数加权和组成的总目标函数,通过采用遗传算法对某850 mm不锈钢冷连轧机典型产品规格进行轧制规程优化计算,并与比例负荷分配法和Nelder-Mead单纯形法求解的轧制规程进行对比,从而确定最优的不锈钢冷连轧轧制规程。结果表明,遗传算法实现了所建立目标函数的高效求解,通过调整目标函数的权重系数,可以有选择地优化轧制规程,采用遗传算法优化轧制规程的计算精度、收敛速度和求解稳定性方面均满足在线控制要求,为高质量不锈钢冷轧生产奠定了基础。Cold rolled stainless steel is part of the high-end fine steel,and its production process control system is one of the most complex industrial control systems,with high system control accuracy,short dynamic response time,fast rolling speed and many process parameters.In the stainless steel cold rolling process,the reasonable rolling schedule is the basis of guaranteeing the quality and production efficiency.The optimization of rolling schedule is an effective way to save energy and reduce carbon in China′s iron and steel enterprises,which has become a research hotspot in the iron and steel industry.And the model accuracy of the process control system will not only affect the dimensional control accuracy of the product,but also affect the R&D and expansion of high-end products and intelligent transformation of cold rolling enterprises.In order to improve the product quality accuracy and reduce the total energy consumption in the stainless steel cold rolling process,the traditional load distribution method and the calculation principle of rolling schedule were deeply studied.The first set of seven-stand stainless steel tandem cold rolling mill in the world was taken as the research object,and an objective function structure form containing the target term and the penalty term was designed.The rolling power balance,rolling force balance,reasonable reduction distribution,low rolling energy consumption and good strip shape were selected as the goals to construct a total objective function.The rolling schedule of typical product in an 850 mm stainless steel cold rolling mill is optimized by genetic algorithm,and compared with the rolling schedule solved by proportional load distribution method and Nelder-Mead simplex algorithm.Thus,the optimal cold rolling schedule for stainless steel is determined.The results show that,the genetic algorithm realizes the efficient solution of the established objective function.The rolling schedule can be optimized selectively by adjusting the weight coefficient of the objective funct

关 键 词：不锈钢 冷连轧 轧制规程 多目标优化 遗传算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TG335.12[自动化与计算机技术—控制科学与工程]