Please wait a minute...
 
2222材料工程  2022, Vol. 50 Issue (6): 170-177    DOI: 10.11868/j.issn.1001-4381.2021.000624
  研究论文 本期目录 | 过刊浏览 | 高级检索 |
基于GA改进BP神经网络预测热变形流变应力模型的建立
汪雅婷1, 黎俊良2, 袁楷峰1, 陈广义1,*()
1 佛山科学技术学院 机电工程与自动化学院,广东 佛山 528200
2 西安热工研究院有限公司,西安 710054
Establishment of hot deformation flow stress prediction model based on GA improved BP neural network
Yating WANG1, Junliang LI2, Kaifeng YUAN1, Guangyi Chen1,*()
1 School of Mechatronic Engineering and Automation, Foshan University, Foshan 528200, Guangdong, China
2 Xi'an Thermal Power Research Institute Co., Ltd., Xi'an 710054, China
全文: PDF(7057 KB)   HTML ( 1 )  
输出: BibTeX | EndNote (RIS)      
摘要 

应力-应变曲线对研究金属热变形过程中的加工硬化、动态再结晶和动态回复的变化具有重要的意义,而预测不同热变形参数下的应力-应变曲线有助于研究热加工过程中金属的可加工性和不稳定性。在应变速率为0.01~3 s-1以及变形温度为1000~1200 ℃条件下,利用Gleeble-3500热模拟试验机对Nb-V-Ti微合金钢进行热压缩实验,研究了Nb-V-Ti微合金钢的热变形行为。建立BP神经网络模型和基于GA改进BP神经网络模型,分别预测在应变速率0.5 s-1、变形温度1050 ℃和应变速率1 s-1、变形温度1100 ℃条件下的流动应力行为并验证模型效果。研究结果表明:经GA改进后的BP神经网络模型对测试数据的应力-应变曲线与实验曲线具有很好的吻合,相关系数分别达0.99202和0.99734,误差仅为2.7816%和2.1703%,预测结果与实验结果相对误差在[-2, 2]范围内,证明了模型的预测可靠性,且适用于较广的应变范围,为工业生产轧制工艺提供理论指导。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
汪雅婷
黎俊良
袁楷峰
陈广义
关键词 热变形流变应力遗传算法BP神经网络预测模型    
Abstract

Stress-strain curves are of great significance in studying the changes of work hardening, dynamic recrystallization and dynamic recovery of metal during hot deformation, and predicting the stress-strain curves under different thermal deformation parameters is helpful to study the machinability and instability of metal in hot working process. The thermal deformation behavior of Nb-V-Ti microalloy steel was studied by hot compression experiments at strain rates of 0.01-3 s-1 and deformation temperatures of 1000-1200 ℃ on Gleeble-3500 thermal simulation testing machine. The BP neural network model and GA improved BP neural network model were established to predict the stress-strain curves at the strain rate of 0.5 s-1 and deformation temperature of 1050 ℃, and the strain rate of 1 s-1 and deformation temperature of 1100 ℃. The results show that the BP neural network model improved by GA is in good agreement with the stress-strain curves of the test data and the experimental curves. The correlation coefficients are 0.99202 and 0.99734 respectively, and the errors are only 2.7816% and 2.1703%. The relative errors between the predicted results and the experimental results are within the range of [-2, 2]. It is proved that the model is reliable and applicable to a wide range of strain, which provides theoretical guidance for rolling process in industrial production.

Key wordsthermal deformation    flow stress    genetic algorithm    BP neural network    prediction model
收稿日期: 2021-07-08      出版日期: 2022-06-20
中图分类号:  TG142.1  
通讯作者: 陈广义     E-mail: 943903188@qq.com
作者简介: 陈广义(1963—),男,教授,硕士,主要研究方向为智能检测与智能控制,联系地址:广东省佛山市南海区狮山镇广云路佛山科学技术学院仙溪校区机电工程与自动化学院(528200),E-mail:943903188@qq.com
引用本文:   
汪雅婷, 黎俊良, 袁楷峰, 陈广义. 基于GA改进BP神经网络预测热变形流变应力模型的建立[J]. 材料工程, 2022, 50(6): 170-177.
Yating WANG, Junliang LI, Kaifeng YUAN, Guangyi Chen. Establishment of hot deformation flow stress prediction model based on GA improved BP neural network. Journal of Materials Engineering, 2022, 50(6): 170-177.
链接本文:  
http://jme.biam.ac.cn/CN/10.11868/j.issn.1001-4381.2021.000624      或      http://jme.biam.ac.cn/CN/Y2022/V50/I6/170
Fig.1  不同应变速率下不同温度的应力-应变曲线
(a)=0.01 s-1;(b) =0.1 s-1;(c) =0.5 s-1;(d) =1 s-1;(e) =3s-1
Fig.2  不同变形温度下不同变形速率的应力-应变曲线
(a)1000 ℃;(b)1050 ℃;(c)1100 ℃;(d)1150 ℃;(e)1200 ℃
Fig.3  GA改进型BP神经网络模型结构
Strain rate/s-1 T/℃
1000 1050 1100 1150 1200
0.01 C C C C C
0.10 C C C C C
0.50 C M C C C
1 C C M C C
3 C C C C C
Table 1  测试数据和训练数据的划分
Fig.4  确定过程中隐含层节点数与训练集均方根误差的关系曲线
Fig.5  不同样本中BP神经网络模型和GA改进型BP神经网络模型输出值的训练误差曲线
(a)=0.5 s-1, T=1050 ℃; (b) =1 s-1, T=1100 ℃
Fig.6  BP神经网络(a)及GA改进型BP神经网络(b)预测应力值与实验数据曲线对比曲线
Fig.7  不同条件时,BP神经网络(1)和GA改进BP神经网络(2)预测的应力值与实验值线性关系图、相对误差比值与相对误差值分布图
(a) =0.5 s-1, T=1050 ℃;(b) =1 s-1, T=1100 ℃
1 ABARGHOOEI H , ARABI H , SEYEDEIN S H , et al. Modeling of steady state hot flow behavior of API-X70 microalloyed steel using genetic algorithm and design of experiments[J]. Applied Soft Computing, 2016, 52, 471- 477.
2 PHANIRAJ M P , LAHIRI A K . The applicability of neural network model to predict flow stress for carbon steels[J]. Journal of Materials Processing Technology, 2003, 141 (2): 219- 227.
doi: 10.1016/S0924-0136(02)01123-8
3 BARAGAR D L . The high temperature and high strain-rate behaviour of a plain carbon and an HSLA steel[J]. Journal of Mechanical Working Technology, 1987, 14 (3): 295- 307.
doi: 10.1016/0378-3804(87)90015-5
4 CHUN M S , BIGLOU J , LENARD J G , et al. Using neural networks to predict parameters in the hot working of aluminum alloys[J]. Journal of Materials Processing Technology, 1999, 86 (1/3): 245- 251.
5 MANDAL S , SIVAPRASAD P V , VENUGOPAL S , et al. Artificial neural network modeling to evaluate and predict the deformation behavior of stainless steel type AISI 304L during hot torsion[J]. Applied Soft Computing, 2015, 9 (1): 237- 244.
6 SUBBARAO S M . Hindcasting of storm waves using neural networks[J]. Ocean Engineering, 2005, 32 (5/6): 667- 684.
7 沈昌武, 杨合, 孙志超, 等. 基于BP神经网络的TA15钛合金本构关系建立[J]. 塑性工程学报, 2007, 14 (4): 101- 104.
doi: 10.3969/j.issn.1007-2012.2007.04.021
7 SHENG C W , YANG H , SUN Z C , et al. Establishment of constitutive relation of TA15 titanium alloy based on BP neural network[J]. Journal of Plasticity Engineering, 2007, 14 (4): 101- 104.
doi: 10.3969/j.issn.1007-2012.2007.04.021
8 孙宇, 曾卫东, 王邵丽, 等. 应用人工神经网络建立Ti-22Al-25Nb合金高温本构关系模型[J]. 塑性工程学报, 2009, 16 (3): 132- 135.
8 SUN Y , ZENG W D , WANG S L , et al. Modeling the constitutive relationship of Ti-22AI-25Nb alloy using artificial neural network[J]. Journal of Plasticity Engineering, 2009, 16 (3): 132- 135.
9 马少伟, 张艳, 杨明, 等. Zener-Hollomon参数对Cr4Mo4Ni4V高合金钢热变形行为的影响[J]. 中南大学学报(自然科学版), 2021, 52 (2): 376- 388.
9 MA S W , ZHANG Y , YANG M , et al. Effect of Zener-Hollomon parameters on hot deformation behavior of Cr4Mo4Ni4V high alloy steel[J]. Journal of Central South University (Science and Technology), 2021, 52 (2): 376- 388.
10 张世伟, 杨明, 梁益龙, 等. 20CrNi2Mo钢高温变形的本构方程与动态再结晶行为[J]. 钢铁, 2017, 52 (8): 97- 106.
10 ZHANG S W , YANG M , LIANG Y L , et al. Constitutive equations and dynamic recrystallization behaviors of 20CrNi2Mo steel at stage of high temperature deformation[J]. Iron & Steel, 2017, 52 (8): 97- 106.
11 程晓农, 桂香, 罗锐, 等. 核电装备用奥氏体不锈钢的高温本构模型及动态再结晶[J]. 材料导报, 2019, 33 (11): 1775- 1781.
doi: 10.11896/cldb.18060090
11 CHENG X N , GUI X , LUO R , et al. Constitutive equation and dynamic recrystallization behavior of 316L austenitic stainless steel for nuclear power equipment[J]. Materials Reports, 2019, 33 (11): 1775- 1781.
doi: 10.11896/cldb.18060090
12 何龙, 张冉阳, 赵刚要, 等. 基于BP神经网络的GH5188高温合金本构模型[J]. 特种铸造及有色合金, 2021, 41 (2): 223- 226.
12 HE L , ZHANG R Y , ZHAO G Y , et al. Constitutive model of GH5188 superalloy based on BP neural network[J]. Special Casting & Nonferrous Alloys, 2021, 41 (2): 223- 226.
13 钟明君, 王克鲁, 鲁世强, 等. MoNb合金高温变形行为及BP神经网络本构模型研究[J]. 塑性工程学报, 2020, 27 (12): 177- 182.
doi: 10.3969/j.issn.1007-2012.2020.12.025
13 ZHONG M J , WANG K L , LU S Q , et al. Study on high temperature deformation behavior and BP neural network constitutive model of MoNb alloy[J]. Journal of Plasticity Engineering, 2020, 27 (12): 177- 182.
doi: 10.3969/j.issn.1007-2012.2020.12.025
14 冯瑞, 王克鲁, 鲁世强, 等. 基于应变补偿和BP神经网络的BT25钛合金本构关系研究[J]. 塑性工程学报, 2020, 27 (12): 183- 190.
doi: 10.3969/j.issn.1007-2012.2020.12.026
14 FENG R , WANG K L , LU S Q , et al. Constitutive relationship research on BT25 titanium alloy based on strain compensation and BP neural network[J]. Journal of Plasticity Engineering, 2020, 27 (12): 183- 190.
doi: 10.3969/j.issn.1007-2012.2020.12.026
15 张又文, 冯斌, 陈页, 等. 基于遗传算法优化XGBoost的油浸式变压器故障诊断方法[J]. 电力自动化设备, 2021, 41 (2): 200- 206.
15 ZHANG Y W , FENG B , CHEN Y , et al. Fault diagnosis method for oil-immersed transformer based on XGBoost optimized by genetic algorithm[J]. Electric Power Automation Equipment, 2021, 41 (2): 200- 206.
16 刘建军, 王克鲁, 鲁世强, 等. Ti-Nb合金热变形行为及应变耦合本构模型[J]. 材料热处理学报, 2019, 40 (5): 156- 161.
16 LIU J J , WANG K L , LU S Q , et al. Hot deformation behavior and strain compensation constitutive model of Ti-Nb alloy[J]. Journal of Materials and Heat Treatment, 2019, 40 (5): 156- 161.
17 李冬勤, 徐磊, 黄兴民, 等. 7A04铝合金动态再结晶的临界应变研究[J]. 材料工程, 2013, (4): 23- 27.
doi: 10.3969/j.issn.1001-4381.2013.04.005
17 LI D Q , XU L , HUANG X M , et al. Investigation on critical strain of dynamic recrystallization for 7A04 aluminum alloy[J]. Journal of Materials Engineering, 2013, (4): 23- 27.
doi: 10.3969/j.issn.1001-4381.2013.04.005
18 ARUN BABU K , PRITHIV T S , GUPTA A , et al. Modeling and simulation of dynamic recrystallization in super austenitic stainless steel employing combined cellular automaton, artificial neural network and finite element method[J]. Computational Materials Science, 2021, 195, 110482.
doi: 10.1016/j.commatsci.2021.110482
19 吴晋彬, 刘国权, 许磊, 等. 基于人工神经网络的中碳含钒微合金钢热变形流变应力预报[J]. 热加工工艺, 2010, 39 (10): 51- 54.
doi: 10.3969/j.issn.1001-3814.2010.10.015
19 WU J B , LIU G Q , XU L , et al. Prediction on hot deformation flow stress of medium-carbon micro-alloyed steels with vanadium based on artficial neural network[J]. Hot Working Technology, 2010, 39 (10): 51- 54.
doi: 10.3969/j.issn.1001-3814.2010.10.015
20 常若寒, 蔡中义, 程丽任, 等. 基于遗传BP网络的Mg-Sm-Zn-Zr合金应力预测模型及加工图[J]. 材料导报, 2017, 31 (3): 136- 149.
20 CHANG R H , CAI Z Y , CHENG L R , et al. Flow stress prediction model and processing map of Mg-Sm-Zn-Zr alloy based on GA-BP neural network[J]. Materials Reports, 2017, 31 (3): 136- 149.
21 BABU K A , MANDAL S . Regression based novel constitutive analyses to predict high temperature flow behavior in super austenitic stainless steel[J]. Materials Science and Engineering: A, 2017, 703 (4): 187- 195.
22 HAGHDADI N , ZAREI-HANZAKI A , KHALESIAN A R , et al. Artificial neural network modeling to predict the hot deformation behavior of an A356 aluminum alloy[J]. Materials & Design, 2013, 49, 386- 391.
23 ZHOU X , WANG K L , LU S Q , et al. Hot deformation behavior of Ti2041 alloy based on BP neural network and 3D processing map[J]. Rare Metal Materials and Engineering, 2021, 50 (4): 1233- 1240.
[1] 汤中英, 邢清源, 杨守杰, 丁宁. 新型Al-Zn-Mg-Sc-Er-Zr合金的热变形行为[J]. 材料工程, 2022, 50(3): 131-137.
[2] 宋广胜, 牛嘉维, 宋鸿武, 张士宏, 邓思瀛. Zirlo锆合金高温变形行为及本构关系[J]. 材料工程, 2022, 50(3): 138-147.
[3] 乔士宾, 何西扣, 刘敬杰, 赵德利, 刘正东. SA508Gr.4N钢热变形过程微观组织演变及流变应力模型[J]. 材料工程, 2021, 49(3): 67-77.
[4] 张连腾, 陈乐平, 徐勇, 袁源平. Mg-9Al-3Si-0.375Sr-0.78Y合金的热变形行为及本构模型[J]. 材料工程, 2021, 49(2): 88-96.
[5] 支盛兴, 李兴刚, 袁家伟, 李永军, 马鸣龙, 石国梁, 张奎. 挤压态AZ40镁合金热变形行为及热加工图分析[J]. 材料工程, 2021, 49(11): 136-146.
[6] 周亚利, 杨秋月, 张文玮, 刘田文, 何威, 吴珍, 伍铭, 谭元标. 具有层片状α相组织的TB8钛合金热变形行为及本构方程[J]. 材料工程, 2021, 49(1): 75-81.
[7] 朱鸿昌, 罗军明, 朱知寿. TB17钛合金β相区动态再结晶行为及转变机理[J]. 材料工程, 2020, 48(2): 108-113.
[8] 戚云超, 方国东, 梁军, 谢军波. 三维针刺C/C-SiC复合材料预制体工艺参数优化[J]. 材料工程, 2020, 48(1): 27-33.
[9] 周峰, 王克鲁, 鲁世强, 万鹏, 陈虚怀. Ti-22Al-24Nb-0.5Y合金流变行为及BP神经网络高温本构模型[J]. 材料工程, 2019, 47(8): 141-146.
[10] 周强, 程军, 于振涛, 崔文芳. 一种新型近β型Ti-5.5Mo-6V-7Cr-4Al-2Sn-1Fe合金热变形行为[J]. 材料工程, 2019, 47(6): 121-128.
[11] 任书杰, 罗飞, 田野, 刘大博, 王克鲁, 鲁世强. A100超高强度钢的流变应力曲线修正与唯象本构关系[J]. 材料工程, 2019, 47(6): 144-151.
[12] 万鹏, 王克鲁, 鲁世强, 陈虚怀, 周峰. 基于应变补偿和PSO-BP神经网络的Ti-2.7Cu合金本构关系[J]. 材料工程, 2019, 47(4): 113-119.
[13] 王宇, 熊柏青, 李志辉, 温凯, 黄树晖, 李锡武, 张永安. 新型超高强Al-Zn-Mg-Cu合金热压缩变形行为及微观组织特征[J]. 材料工程, 2019, 47(2): 99-106.
[14] 付平, 刘栩, 戴青松, 张佳琪, 邓运来. 5083铝合金热压缩流变应力曲线修正与本构方程[J]. 材料工程, 2017, 45(8): 76-82.
[15] 杨志强, 刘正东, 何西扣, 刘宁. 反应堆压力容器用SA508Gr.4N钢的热变形行为[J]. 材料工程, 2017, 45(8): 88-95.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
地址:北京81信箱44分箱 邮政编码: 100095
电话:010-62496276 E-mail:matereng@biam.ac.cn
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn