节点文献
基于神经网络的板凸度影响系数优化
OPTIMIZATION OF THE INFLUENCE COEFFICIENT OF STRIP CROWN BY NEURAL NETWORK
【作者】 朱洪涛; 王哲; 王国栋; 周勇; 劳兆利; 顾豪杰;
【Author】 ZHU Hongtao WANG Zhe WANG Guodong(Northeastern University)ZHOU Yong LAO Zhaoli GU Haojie(Shanghai Baosteel Group Corp. )
【摘要】 以实测数据为基础,采用参数神经网络结构对板凸度模型影响系数进行了优化,提出利用动态数据库滚动优化算法自我完善提高运行精度。预报结果表明,基于神经网络的板凸度影响系数优化能有效提高设定精度,适应不断变化的工艺过程和设备条件。
【Abstract】 In this paper,parameter network is used to optimize the influence coefficient of strip crown on the base of measured data. Dynamic database has been introduced in order to consummate itself and improve running precision. The comparison between measured and predicted values shows that profile setting-up precision is increased and the optimization model can be adapted to varied techniques and equipment.
- 【会议录名称】 1999中国钢铁年会论文集(下)
- 【会议名称】1999中国钢铁年会
- 【会议时间】1999
- 【分类号】TG334.9
- 【主办单位】中国金属学会