基于数据驱动的热连轧终轧温度预测  

Prediction of finishing rolling temperature of hot stripmill based on data driving

作　　者：张祥壮 张帅 李爱莲[1] 崔桂梅[1] 杨培宏[1] ZHANG Xiangzhuang;ZHANG Shuai;LI Ailian;CUI Guimei;YANG Peihong(Information Engineering Institute,Inner Mongolia University of Science and Technology,Baotou 014010,China)

机构地区：[1]内蒙古科技大学信息工程学院,内蒙古包头014010

出　　处：《中国测试》2024年第3期152-159,共8页China Measurement & Test

基　　金：国家自然科学基金资助项目(61763039);内蒙古自治区自然科学基金项目资助(2022MS06003);内蒙古自治区重大专项(2021ZD0029)。

摘　　要：终轧温度是热连轧生产过程中主要控制的工艺参数,是确保带钢质量的重要前提。带钢在精轧阶段经历复杂的换热过程,现场采用的半机理模型很难提高预测精度。针对此问题,从数据驱动角度出发,建立一种基于多策略改进鲸鱼优化算法(IWOA)与极限学习机(ELM)相结合的终轧温度预测模型。融入柯西变异提升鲸鱼算法跳出局部最优的能力;借助余弦控制因子平衡鲸鱼算法全局搜索与局部开发能力;引入翻身觅食策略降低鲸鱼算法陷入局部最优的概率和提升算法的收敛速度。实验结果表明:建立的IWOA-ELM终轧温度预测模型在预报精度方面优势明显,预测终轧温度在±6℃以内的命中率为94%,具有广阔的应用前景。The finishing rolling temperature is the main control process parameter in the hot continuous rolling process,and is an important prerequisite to ensure the quality of strip steel.The strip steel has experienced a complex heat exchange process in the finishing rolling stage,and it is difficult to improve the prediction accuracy of the semi mechanism model adopted in the field.In view of this problem,from the perspective of data-driven,a prediction model of finishing rolling temperature based on the combination of multi strategy improved whale optimization algorithm(IWOA)and extreme learning machine(ELM)is established.Incorporating cauchy variation improves the ability of whale algorithm to jump out of local optimization;Balance the global search and local development capabilities of whale algorithm with cosine control factors;The turning over for food is introduced to reduce the probability of the whale algorithm falling into the local optimum and improve the convergence speed of the algorithm.The experimental results show that the IWOAELM finishing rolling temperature prediction model has obvious advantages in prediction accuracy,and the hit rate of predicting the finishing rolling temperature within±6℃is 94%,which has broad application prospects.

关 键 词：终轧温度预测 鲸鱼优化算法 柯西变异 翻身觅食 余弦控制因子 

分 类 号：TG335.56[金属学及工艺—金属压力加工] TP183[自动化与计算机技术—控制理论与控制工程] TB9[自动化与计算机技术—控制科学与工程]