锻压技术

基于改进型支持向量机算法的轧机轧制力预测

英文标题：Rolling force prediction of rolling mill based on improved support vector machine algorithm

作者：王前锋

单位：河南经贸职业学院

分类号：TP273

出版年,卷(期):页码：2019,44(4):131-137

摘要：

考虑到基于神经网络算法建立的预测模型虽然具有较好的预测精度，但是神经网络模型需要大量的训练样本，另外会增加模型的复杂程度，研究了一种基于改进型支持向量机的轧机轧制力预测模型，建立基于RBF核函数和多项式核函数的最小二乘支持向量机，并使用协同量子粒子群算法对混合函数的参数进行寻优，以提高预测模型的预测性能。由协同量子粒子群算法优化得到了基于改进型支持向量机的轧机轧制力预测模型中的RBF核函数参数γ值、惩罚系数c值、多项式核函数参数q值和两个核函数的权重a值。通过实例研究表明：使用本文研究的改进型支持向量机的轧制力预测模型预测相对误差在4%~6%之间，多组数据的平均值误差为4.83%。验证了本文研究的基于改进型支持向量机的轧机轧制力预测模型的可行性。本文研究的预测模型相比其他3种对比模型耗时更长，但是相比之下提高了预测准确率，更具有实际意义。

Considering that the prediction model based on neural network algorithm has better prediction accuracy, but the neural network model needs a large number of training samples which increases the complexity of the model, a rolling force prediction model of rolling mill based on improved support vector machine was studied. Then, the least squares support vector machine based on RBF kernel function and polynomial kernel function was established, and the parameters of hybrid function were optimized by the cooperative quantum particle swarm optimization algorithm to improve the prediction performance of the prediction model. Furthermore, the RBF kernel function parameter γ, penalty coefficient c, polynomial kernel function parameter q and weight a of two kernel functions in the rolling force prediction model of rolling mill based on the improved support vector machine were obtained by the cooperative quantum particle swarm optimization algorithm. The example study show that the relative error of rolling force prediction model based on the improved support vector machine is between 4%-6%, and the average error of multi-group data is 4.83%. Thus the feasibility of the rolling force prediction model based on the improved support vector machine is verified. The prediction model studied in this paper takes longer time than the other three kinds of comparison models, but it is more practical to improve the prediction accuracy.

基金项目：

河南省自然科学基金资助项目（163400510331）

王前锋（1981-），男，硕士，讲师 E-mail：wangqianfengpaper@126.com

参考文献：

[1]顾波,王娇,白晶.基于热连轧机厚度精度的最优控制研究[J].机床与液压, 2018,46(8): 126-128.

Gu B, Wang J, Bai J. Optimal control of thickness accuracy of hot rolling mill [J]. Machine Tool & Hydraulic, 2018,46 (8): 126-128.

[2]樊浩森，胡建华，白雪，等. 汽车覆盖件拉深过程中的压边力预测[J]. 锻压技术，2017，42（7）：43-47.

Fan H S，Hu J H，Bai X，et al. Prediction on blank holder force of automobile covering parts in deep drawing process[J]. Forging & Stamping Technology，2017，42（7）：43-47.

[3]张生,姜万录,张佳慧.基于支持向量机预测的冷连轧机轧制力精确设定方法研究[J].液压与气动,2017,(7):50-55.

Zhang S, Jiang W L, Zhang J H. SVM prediction-based rolling force setting calculation method of tandem cold rolling mill [J]. Chinese Hydraulics & Pneumatics, 2017, (7): 50-55.

[4]何飞,石露露,黎敏,等.基于多模态和加权支持向量机的热轧轧制力智能预报[J].工程科学学报,2015,37(4):517-521.

He F, Shi L L, Li M, et al. Intelligent prediction of rolling force in hot rolling based on a multi-model and weighted support vector machine [J]. Chinese Journal of Engineering, 2015, 37 (4): 517-521.

[5]林伟路，丁小凤，双远华. BP神经网络对斜轧穿孔轧制力的预测[J].锻压技术，2018,43(10):175-178.

Lin W L，Ding X F，Shuang Y H. Prediction on rolling force of oblique rolling piercing based on BP neural network [J]. Forging & Stamping Technology，2018, 43(10):175-178.

[6]庄野,张辉,姜永芳. 基于神经网络的带钢热连轧机轧制力预报[J].控制工程,2013,20(S1):122-124.

Zhuang Y, Zhang H, Jiang Y F. Based on neural network strip steel strip machine rolling force prediction[J].Control Engineering of China, 2013,20(S1): 122-124.

[7]张俊明,刘军,俞晓峰,等.轧制力预测中RBF神经网络的组合应用[J].钢铁研究学报,2008, 20(2):33-36.

Zhang J M, Liu J, Yu X F, et al. Application of combination of RBF neural network to prediction of roiling force [J].Journal of Iron and Steel Research, 2008, 20 (2): 33-36.

[8]王桂霞. 热轧带钢平整机轧制力与板形控制的研究[D]. 邯郸：河北工程大学,2015.

Wang G X. Research on Rolling Force and Strip Shape Control of Hot Strip Mill[D]. Handan: Hebei University of Engineering, 2015.

[9]吴东升,王大志,杨青,等.基于ACPSO优化SVR的棒材连轧轧制力预测研究[J].仪器仪表学报,2012,33(11):2579-2585.

Wu D S, Wang D Z, Yang Q, et al. Study on bar rolling force prediction based on support vector regression optimized by accelerate convergence particle swarm optimization [J]. Chinese Journal of Scientific Instrument, 2012, 33 (11): 2579-2585.

[10]Vapnik V N. The Nature of Statistical Learning Theory [M].New York:Spring,1995.

[11]Suykens J A K, Vandewalle J. Least squares support vector machine classifiers[J]. Neural Processing Letters, 1999, 3(9): 293-300.

[12]郑志成,徐卫亚,徐飞,等.基于混合核函数PSO-LSSVM的边坡变形预测[J].岩土力学,2012,33(5):1421-1426.

Zheng Z C, Xu W Y, Xu F, et al. Forecasting of slope displacement based on PSO-LSSVM with mixed kernel [J]. Rock and Soil Mechanics, 2012, 33 (5): 1421-1426.

[13]Sun J, Feng B, Xu W B. Particle swarm optimization with particles having quantum behavior[A]. Proceedings of the 2004 Congress on Evolutionary Computation[C]. Portland, USA: 2004.

[14]张拓,王建平.基于CQPSO-LSSVM的网络入侵检测模型[J].计算机工程与应用,2015,51(2):113-116.

Zhang T, Wang J P. Network intrusion detection based on cooperative quantum-behaved particle swarm algorithm and least square support vector machine [J]. Computer Engineering and Applications, 2015,51 (2): 113-116.

服务与反馈：

【文章下载】【加入收藏】