摘要
根据某1+4铝热连轧厂现场采集的大量轧制数据对几种铝合金变形抗力利用最小二乘支持向量机进行了反向建模回归分析,用细菌觅食优化算法对支持向量机的参数进行了优化。将回归后的变形抗力模型用于二级设定计算中的轧制力预报,结果表明回归后的模型适用于轧制现场,精度优于传统模型。
According to lots of actual measured rolling data of aluminum alloy from one factory of aluminum hot tandem rolling, the deformation resistance model by least squares support vector machine reversed modeling method is got. Meanwhile, Bacteria Foraging Optimization algorithm was applied to optimize the parameters of LSSVM model to improve the accuracy. Compared with the traditional deformation resistance model, the results show that the deformation resistance model by reverse modeling is more accurate.
出处
《计量学报》
CSCD
北大核心
2013年第6期532-536,共5页
Acta Metrologica Sinica
基金
国家冷轧板带装备及工艺工程技术研究中心开放课题(2012005)
河北省工业计算机控制工程重点实验室开放课题(201112006)
关键词
计量学
铝热连轧
变形抗力
最小二乘支持向量机
细菌觅食优化算法
反向建模
Metrology
Aluminum hot tandem rolling
Deformation resistance
Least squares support vector machine
Bacteria foraging optimization algorithm
Reverse modeling