多层感知机和支持向量机回归算法预测T型接头固有应变对比研究

doi:10.3901/JME.2025.18.181

摘要/Abstract

摘要： 传统的固有应变法为了保证获取的固有应变准确性，大多采用热弹塑性有限元法求取。然而，采用热弹塑性有限元法求取T型接头的固有应变会耗费大量的时间。为了提高T型接头固有应变的获取效率，并保证其准确度，获取了44组T型接头焊接实验的固有应变值，并基于实验的数据，分别采用多层感知机和支持向量机回归算法对T型接头的横向和纵向固有应变值进行了预测，对比分析了两种算法预测结果的准确度。对于T型接头的横向和纵向固有应变预测，支持向量机回归算法的预测结果在数据测试集上的决定系数R²分别为0.998和0.974，多层感知机算法的预测结果在数据测试集上的决定系数R²分别为0.985和0.945。结果表明，支持向量机回归算法相较于多层感知机算法预测T型接头的固有应变值更为准确。

关键词: 焊接变形, 固有应变法, 多层感知机, 支持向量机回归, T型接头

Abstract: Conventional inherent strain methods are mostly derived by the thermoelastic-plastic finite element method in order to ensure the accuracy of the acquired inherent strain. However, using the thermoelastic-plastic finite element method to obtain the inherent strain of T-joints will consume a lot of time. In order to improve the efficiency and accuracy of obtaining the inherent strains of T-joints, the inherent strain values of 44 groups of T-joint welding experiments were obtained, and based on the data of the experiments, the transverse and longitudinal inherent strain values of T-joints were predicted by using the multilayer perceptual machine and the support vector machine regression algorithms, respectively, and the accuracies of the prediction results of the two algorithms were compared. For the prediction of transverse and longitudinal inherent strains of T-joints, the coefficients of determination (R²) of the support vector machine regression algorithm are 0.998 and 0.974 for the data test set, and the coefficients of determination (R²) of the multilayer perceptron algorithm are 0.985 and 0.945 for the data test set, respectively, The results show that the support vector machine regression algorithm is more accurate in predicting the inherent strains of T-joints compared with the multilayer perceptron algorithm. joint inherent strain values more accurately than the multilayer perceptron algorithm.

Key words: welding deformation, inherent strain method, MLP, SVR, T-joint

中图分类号:

TG404

康俊涛, 韦朝校. 多层感知机和支持向量机回归算法预测T型接头固有应变对比研究[J]. 机械工程学报, 2025, 61(18): 181-189.

KANG Juntao, WEI Chaoxiao. Comparative Study of Multilayer Perceptron and Support Vector Machine Regression Algorithms for Predicting Inherent Strain in T-joints[J]. Journal of Mechanical Engineering, 2025, 61(18): 181-189.

参考文献

[1] Busari Y O，Manurung Y H P，Shuaib-Babata Y L，et al. Experimental validation on multi-pass weld distortion behavior of structural offshore steel HSLA S460 using FE-based inherent strain and thermo- mechanical method[J]. MRS Communications，2022，12(1)：104-111.
[2] 李元泰，吴华鑫，董斌，等. Q345/SUS304异种钢焊接接头固有应变的变化规律研究[J]. 中国机械工程，2021，32(15)：1868-1873. LI Yuantai，WU Huaxin，DONG Bin，et al. Study on the variation law of inherent strain of Q345/SUS304 dissimilar steel welded joints[J]. China Mechanical Engineering，2021，32(15)：1868-1873.
[3] LU Y H，ZHU S C，ZHAO Z T，et al. Numerical simulation of residual stresses in aluminum alloy welded joints[J]. Journal of Manufacturing Processes，2020，50(2)：380-393.
[4] 梁伟，郑颖，邓德安. 外拘束对铝合金薄板结构焊接变形的影响[J]. 机械工程学报，2021，57(6)：70-77. LIANG Wei，ZHENG Ying，DENG Dean. The influence of external constraints on welding deformation of aluminum alloy thin plate structures[J]. Journal of Mechanical Engineering，2021，57(6)：70-77.
[5] YI B，WANG J C. Mechanism clarification of mitigating welding induced buckling by transient thermal tensioning based on inherent strain theory[J]. Journal of Manufacturing Processes，2021，68(8)：1280-1294.
[6] LEE J. Multi-pass welding distortion analysis using layered shell elements based on inherent strain[J]. Journal of Marine Science and Engineering，2021，9(6)：632-647.
[7] 王超，吴春彪，邵龙潭. 焊接变形预测中基于假定热应变法的弹性有限元建模[J]. 机械工程学报，2022，58(4)：111-119. WANG Chao，WU Chunbiao，SHAO Longtan. Elastic finite element modeling based on assumed thermal strain method in welding deformation prediction[J]. Journal of Mechanical Engineering，2022，58(4)：111-119.
[8] TIAN L，LUO Y. A comparison study of BPN and SVM prediction models for inherent deformations of T-welded joints[J]. Mechanics of Advanced Materials and Structures. 2021，28(1)：365-383.
[9] TAO Y L，MIAO Y S，HAN J Q，et al. Prediction of welding shrinkage deformation of bridge steel box girder based on wavelet neural network [C]//Curran Associates, Inc..3rd International Conference on Automation. Mechanical Control and Computational Engineering：AMCCE 2018，May 12-13，2018，China：Dalian，2018，440-444.
[10] 苟中林，罗宇，袁昱超. 基于计算机视觉的焊接固有应变实测方法研究[J]. 热加工工艺，2024，53(3)：29-32. GOU Zhonglin，LUO Yu，YUAN Yuchao. Research on the measurement method of welding intrinsic strain based on computer vision [J]. Thermal Processing Technology，2024，53(3)：29-32.
[11] LI L，LIU D，REN S，et al. Prediction of welding deformation and residual stress of a thin plate by improved support vector regression[J]. Scanning，2021，2021(1)：1-10.
[12] LI Ran，DONG Manshu，GAO Hongming. Prediction of bead geometry with changing welding speed using artificial neural network[J]. Materials，2021，14(6)：1494-1503.
[13] LIU F F，TAO C C，DONG Z B，et al. Prediction of welding residual stress and deformation in electro-gas welding using artificial neural network[J]. Materials Today Communications，2021，29(1)：1-11.
[14] 刘紫阳. 基于APDL语言T型接头双侧同步焊有限元建模及在钢结构变形预测中应用研究[D]. 重庆：重庆交通大学，2015. LIU Ziyang. Finite element modeling of double-sided synchronous welding of T-joints based on APDL language and its application in deformation prediction of steel structures[D]. Chongqing：Chongqing Jiaotong University，2015.
[15] 高辰. 基于固有应变法的舰船标准平台焊接失稳变形预测研究[D]. 镇江：江苏科技大学，2017. GAO Chen. Research on the prediction of welding instability and deformation of ship standard platform based on intrinsic strain method[D]. Zhenjiang：Jiangsu University of Science and Technology，2017.
[16] 王秋实. 转向架构架侧梁的焊接变形数值模拟研究[D].成都：西南交通大学，2017. WANG Qiushi. Numerical simulation study on welding deformation of side beams of bogie frame[D]. Chengdu：Southwest Jiaotong University，2017.
[17] 李飞. 钢箱梁焊接固有应变的研究[D]. 天津：河北工业大学，2011. LI Fei. Study on the inherent strain of welded steel box girder[D]. Tianjin：Hebei University of Technology，2011.
[18] 魏天冬. T型构件焊接固有应变的研究及应用[D]. 秦皇岛：燕山大学，2011. WEI Tiandong. Research and application of natural strain in welding of T-joints[D]. Qinhuangdao：Yanshan University，2011.
[19] 丁雪松，黄立群，张步忠，等. 基于多层感知机的蛋白质变性温度预测[J]. 计算机应用研究，2019，36(8)：2421-2423. DING Xuesong，HUANG Liqun，ZHANG Buzhong，et al. Protein denaturation temperature prediction based on multi-layer perceptron[J]. Computer Application Research，2019，36(8)：2421-2423.
[20] 卢纯义，于津，余忠东，等. 基于改进灰狼算法优化SVR的混凝土中钢筋直径检测方法[J]. 计算机科学，2022，49(11)：228-233. LU Chunyi，YU Jin，YU Zhongdong，et al. Optimization of steel reinforcement diameter detection method in concrete based on improved grey wolf algorithm and SVR[J]. Computer Science，2022，49(11)：228-233.