Optimal Design of Hot Rolling Process for C-Mn Steel Based on Industrial Big Data

摘要：

　　The optimization design of hot rolling process is a time-consuming work in the production of hot strips.The traditional optimal design method requires a lot of destructive physical experiments,which consumes a lot of manpower,material resources and time.In addition,due to the complexity of the microstructure evolution and various constraints in the actual rolling process,a better but not the best rolling process is obtained by the traditional optimization design method.Therefore,researchers propose a method based on the industrial big data collected from production line using neural network model to investigate the relationships between the chemistry,rolling process and mechanical property of steels.The optimal design of hot rolling process for hot rolled strips is carried out by combining with the multi-objective optimization technology.A successful mechanical property prediction model is the core of the optimal design of hot rolling process.However,the original industrial big data,usually unbalanced,are inevitably mixed with fluctuant and abnormal values.Models established on the data without data processing can cause misleading results,which cannot be used for the optimal design of hot rolling process.In this paper,a method of industrial big data processing of C-Mn steels was proposed based on the data analysis.The Bayesian neural network was used to establish the reliable mechanical properties models for the optimal design of hot rolling process.By using the multi-objective optimization algorithm,considering the personal requirements and the constraints in the actual production line,the optimal design of hot rolling process was successfully applied to new rolling process design for Q345B with 0.017%Nb and 0.046%Ti content removed.The optimal process design results were in good agreement with the industry trials results,which verify the effectiveness of the optimal design of hot rolling process.

关键词： industrial big data data processing mechanical property optimal design of hot rolling process C-Mn steel

作者: WU Siwei ZHOU Xiaoguang CAO Guangming LIU Zhenyu SHI Naian

作者单位: State Key Laboratory of Rolling and Automation,Northeastern University,Shenyang 110819,Liaoning,Chin Continuous Rolling Plant of No.3 Steelmaking of Angang Co.,Ltd.,Anshan 114021,Liaoning,China

会议类型: 国际会议

会议名称: STEELSIM2017(第七届冶金过程及钢铁材料模拟及仿真国际会议)

会议地点: 青岛

会议语种:英文

页码: 346-349

在线出版日期: 2017-08-16（万方平台首次上网日期，不代表论文的发表时间）