锻压技术

基于神经网络的模锻压机载荷在线预测模型

英文标题：Online predictive model for the loading of die forging equipment based on neural network

作者：蔺永诚梁英杰谌东东

分类号：TG316

出版年,卷(期):页码：2016,41(10):98-102

摘要：

由于模锻过程具有强时变性和非线性，因此精确控制模锻压机载荷预测至关重要。以7075铝合金模锻过程为例，提出了一种基于神经网络的模锻压机载荷在线建模方法。基于商业软件Deform-3D模拟了恒温恒速度工况下的载荷变化规律，根据获取的数据建立了初始神经网络模型。在实际模锻实验过程中，通过反向传播算法不断修正初始神经网络权值矩阵，以实现模型的在线更新。在50 t模锻实验台上进行实验，以验证所提方法的有效性。实验结果表明：所提出的在线建模方法可以准确预测复杂模锻工况下载荷的变化，与传统离线神经网络建模方法相比，其预测值更加准确，更能满足实际工程需求。

For the issue of time-variation and nonlinearity of the die forging process, the accurate prediction of die forging loading is very important. For the die forging process of Al alloy 7075, an online modeling method based on neural network was proposed to predict loading. Firstly, the change of loading at a constant temperature and velocity was simulated by Deform-3D software, and then the initial neural network model was developed based on the simulated data. Therefore, during the actual die forging, the weight matrix of neural networks was updated by the back-propagation algorithm continually. Finally, the confirmatory experiment was conducted by a 50 t die forging equipment to verify the validity of the proposed method. The results show that the proposed method can accurately predict the loading in the die forging process. Compared with the traditional offline neural network modeling, the predicted results by the proposed method are more accurate, which can satisfy the requirements of practical engineering.

基金项目：

国家重点基础研究发展计划（“973”计划）（2013CB035801）; 国家自然科学基金资助项目 (51375502)

作者简介：蔺永诚（1976-），男，博士，教授 E-mail：yclin@csu.edu.cn;linyongcheng@163.com

参考文献：

[1]蔺永诚，谌东东，陈明松. 基于BP神经网络的大型模锻压机上横梁速度预测控制方法[P]. 中国：CN105652666A，2016-06-08.

Lin Y C, Chen D D, Chen M S. BP neural network-based model predictive control method for the velocity of upper die of large die forging press[P]. China:CN105652666A,2016-06-08.

[2]蔺永诚，谌东东，陈明松. 基于泰勒展开的大型模锻压机上横梁速度在线预测方法[P]. 中国:CN105808949A,2016-07-27.

Lin Y C, Chen D D, Chen M S. Tayler expansion-based online predictive method for the velocity of upper die of large die forging press[P]. China:CN105808949A,2016-07-27.

[3]Lin Y C, Li K K, Li H B, et al. New constitutive model for high-temperature deformation behavior of inconel 718 superalloy[J]. Materials & Design, 2015, 74: 108-118.

[4]高斌, 李强. 身管径向锻造材料流动的仿真分析[J]. 锻压技术, 2015, 40(11): 155-158.

Gao B, Li Q. Study on simulation of material flow for radial forging barrel[J]. Forging & Stamping Technology, 2015, 40(11): 155-158.

[5]谭海林, 赖春明. 非均匀温度场对高强钢热冲压成形性的影响[J]. 锻压技术, 2015, 40(9): 32-36.

Tan H L, Lai C M. Influence of non-uniform temperature field on hot stamping formability of high strength steel[J]. Forging & Stamping Technology, 2015, 40(9): 32-36.

[6]吴瑶，许晓静，张振强，等. 多向锻造2099铝锂合金挤压材的组织和性能[J]. 稀有金属, 2014, 38(6): 961-966.

Wu Y, Xu X J, Zhang Z Q, et al. Microstructure and property of multi-directional forged 2099 Al-Li alloy extrusions[J]. Chinese Journal of Rare Metals, 2014, 38(6): 961-966.

[7]Lin Y C, Zhang J, Zhong J. Application of neural networks to predict the elevated temperature flow behavior of a low alloy steel[J]. Computational Materials Science, 2008, 43(4): 752-758.

[8]íma J. Neural expert systems[J]. Neural Networks, 1995, 8(2): 261-271.

[9]Zhang G P. Time series forecasting using a hybrid ARIMA and neural network model[J]. Neurocomputing, 2003, 50: 159-175.

[10]Venayagamoorthy G K, Rohrig K, Erlich I. One step ahead: short-term wind power forecasting and intelligent predictive control based on data analytics[J]. Power and Energy Magazine, IEEE, 2012, 10(5): 70-78.

[11]Rahman M A, Hoque M A. On-line adaptive artificial neural network based vector control of permanent magnet synchronous motors[J]. IEEE Transactions on Energy Conversion, 1999, 13(4): 311-318.

[12]Mller M F. A scaled conjugate gradient algorithm for fast supervised learning[J]. Neural Networks, 1993, 6(4): 525-533.

服务与反馈：

【本网站尚未开通全文下载服务】【加入收藏】