钢板缺陷识别的Volterra-SVM模型研究

邓勇; 黄远伟; 赖治屹

doi:10.13433/j.cnki.1003-8728.20200590

钢板缺陷识别的Volterra-SVM模型研究

doi: 10.13433/j.cnki.1003-8728.20200590

1.
西南石油大学机电工程学院, 成都 610500
2.
西南油气田分公司输气管理处, 成都 610213

基金项目:

四川省科技支撑计划项目 2017FZ0033

详细信息

作者简介:
邓勇(1972-), 高级工程师, 硕士生导师, 研究方向为模式识别和超声无损检测, 201231010013@swpu.edu.cn

中图分类号: TB551;TP391.4
计量
- 文章访问数: 120
- HTML全文浏览量: 75
- PDF下载量: 21
- 被引次数: 0
出版历程
- 收稿日期: 2021-03-17
- 刊出日期: 2023-01-25

Study on Volterra-SVM Model for Defect Recognition of Steel Plate

1.
School of Mechatronic Engineering, Southwest Petroleum University, Chengdu 610500, China
2.
Gas Transmission Management Office of Southwest Oil and Gas Field Company, Chengdu 610213, China

摘要

摘要: 针对钢板缺陷识别问题, 结合超声波脉冲反射原理, 提出一种基于Volterra级数和支持向量机的钢板缺陷识别方法。首先, 利用Volterra级数模型建立起钢板缺陷的特征模型; 其次, 使用分数阶粒子群优化算法提取出原始信号中的特征参数, 即Volterra级数时域核; 最后, 将提取到的特征向量输入支持向量机模型进行训练与测试, 完成对钢板缺陷的分类识别。设计实验得到多组数据样本, 进行模型验证, 实验结果表明: 基于Volterra级数和支持向量机的识别模型能够较好的完成对钢板缺陷的分类识别, 识别准确率达93.3%。
- 缺陷识别 /
- Volterra级数 /
- 分数阶粒子群优化算法 /
- 支持向量机
Abstract: Aiming at the problem of steel plate defect recognition, combined with the principle of ultrasonic pulse reflection, a steel plate defect recognition method based on Volterra series and Support Vector Machine (SVM) is proposed. Firstly, Volterra series model is used to construct the characteristic model of steel plate defects. Then, the feature parameters in the original signal, namely Volterra series kernel, are extracted by using the Fractional Order Particle Swarm Optimization (FO-PSO). Finally, the extracted feature vectors are input into the SVM model for training and testing to complete the classification and recognition of steel plate defects. Experiments were designed to obtain multiple sets of data samples for model validation. The experimental results show that the recognition model based on Volterra series and SVM can better complete the classification and recognition of steel plate defects, and the recognition accuracy is 93.3%.
- defect recognition /
- volterra series /
- fractional order particle swarm optimization algorithm /
- support vector machine

HTML全文

图 1 脉冲反射原理示意图

下载: 全尺寸图片幻灯片

图 2 部分人工缺陷示意图

下载: 全尺寸图片幻灯片

图 3 超声检测实验装置图

下载: 全尺寸图片幻灯片

图 4 原始信号图

下载: 全尺寸图片幻灯片

图 5 FO-PSO适应度值曲线

下载: 全尺寸图片幻灯片

图 6 部分缺陷类型的三维曲面图

下载: 全尺寸图片幻灯片

图 7 径向基核函数SVM预测结果

下载: 全尺寸图片幻灯片

表 1 不同种类缺陷

序号	大小/mm	形状	深度/mm
1	6×6	圆形	2
2	6×6	圆形	3
3	6×6	矩形	2
4	12×12	圆形	2
5	12×12	圆形	3
6	12×12	矩形	2
7	6×12	不规则形	2
8	6×12	不规则形	3
9	无缺陷	-	-

下载: 导出CSV

表 2 不同模型分类性能对比

模型	耗时/s	训练准确率/%	测试准确率/%
SVM	67.3	100	93.3
BP-NN	54.6	97.8	84.4

下载: 导出CSV

参考文献(20)

[1]	GHOLIZADEH S. A review of non-destructive testing methods of composite materials[J]. Procedia Structural Integrity, 2016, 1: 50-57 doi: 10.1016/j.prostr.2016.02.008
[2]	JI B P, ZHANG Q D, CAO J S, et al. Non-contact detection of delamination in stainless steel/carbon steel composites with laser ultrasonic[J]. Optik, 2021, 226: 165893 doi: 10.1016/j.ijleo.2020.165893
[3]	SAMBATH S, NAGARAJ P, SELVAKUMAR N. Automatic defect classification in ultrasonic NDT using artificial intelligence[J]. Journal of Nondestructive Evaluation, 2011, 30(1): 20-28 doi: 10.1007/s10921-010-0086-0
[4]	YANG P, YANG Q T. Empirical Mode Decomposition and rough set attribute reduction for ultrasonic flaw signal classification[J]. International Journal of Computational Intelligence Systems, 2014, 7(3): 481-492 doi: 10.1080/18756891.2014.889877
[5]	VIRUPAKSHAPPA K, ORUKLU E. Multi-class classifica- tion of defect types in ultrasonic NDT signals with convolutional neural networks[C]//2019 IEEE International Ultrasonics Symposium (IUS). Glasgow: IEEE, 2019: 1647-1650
[6]	FEI C, HAN Z H, DONG J. An ultrasonic flaw-classification system with wavelet-packet decomposition, a mutative scale chaotic genetic algorithm, and a support vector machine and its application to petroleum-transporting pipelines[J]. Russian Journal of Nondestructive Testing, 2006, 42(3): 190-197 doi: 10.1134/S1061830906030077
[7]	潘峰, 唐东林, 陈印, 等. 管道腐蚀缺陷超声信号的PSO-SVM模式识别研究[J]. 机械科学与技术, 2020, 39(5): 751-757 doi: 10.13433/j.cnki.1003-8728.20190197 PAN F, TANG D L, CHEN Y, et al. Ultrasonic signal pattern recognition of pipeline corrosion defects with PSO-SVM[J]. Mechanical Science and Technology for Aerospace Engineering, 2020, 39(5): 751-757 (in Chinese) doi: 10.13433/j.cnki.1003-8728.20190197
[8]	车红昆, 吕福在, 项占琴. 超声检测信号多特征SVM-Bayes融合识别[J]. 振动与冲击, 2011, 30(12): 265-269 doi: 10.3969/j.issn.1000-3835.2011.12.053 CHE H K, LYU F Z, XIANG Z Q. Ultrasonic signal recogni- tion with multi-feature SVM-Bayes fusion method[J]. Journal of Vibration and Shock, 2011, 30(12): 265-269 (in Chinese) doi: 10.3969/j.issn.1000-3835.2011.12.053
[9]	吴刚, 关山月, 汪小凯, 等. 薄板点焊超声检测信号特征分析与缺陷识别[J]. 焊接学报, 2019, 40(4): 112-118 https://www.cnki.com.cn/Article/CJFDTOTAL-HJXB201904020.htm WU G, GUAN S Y, WANG X K, et al. Feature analysis and defect recognition of ultrasonic detection signal for spot welding of sheet[J]. Transactions of the China Welding Institution, 2019, 40(4): 112-118 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HJXB201904020.htm
[10]	卢超, 于润桥, 彭应秋. 焊接缺陷超声检测回波信号的双谱分析[J]. 数据采集与处理, 2003, 18(1): 57-61 doi: 10.3969/j.issn.1004-9037.2003.01.013 LU C, YU R Q, PENG Y Q. Bispectrum analysis on ultrasonic testing for echo signal of welding flaw[J]. Journal of Data Acquisition & Processing, 2003, 18(1): 57-61 (in Chinese) doi: 10.3969/j.issn.1004-9037.2003.01.013
[11]	杨琳瑜, 于润桥, 黄昌光, 等. 基于小波包变换的复合材料超声波检测信号特征提取[J]. 应用声学, 2007, 26(3): 176-180 doi: 10.3969/j.issn.1000-310X.2007.03.008 YANG L Y, YU R Q, HUANG C G, et al. Feature extraction from carbon fiber composites ultrasonic signals based on wavelet packet transform[J]. Journal of Applied Acoustics, 2007, 26(3): 176-180 (in Chinese) doi: 10.3969/j.issn.1000-310X.2007.03.008
[12]	王海涛, 张霄, 史丽晨, 等. 多脉冲激励法在轴承滚珠磨损中的状态研究及应用[J]. 机械科学与技术, 2018, 37(1): 30-35 doi: 10.13433/j.cnki.1003-8728.2018.0106 WANG H T, ZHANG X, SHI L C, et al. State research and application of multiple-pulse excitation method in diagnosis of bearing ball wear[J]. Mechanical Science and Technology for Aerospace Engineering, 2018, 37(1): 30-35 (in Chinese) doi: 10.13433/j.cnki.1003-8728.2018.0106
[13]	李志农, 蒋静, 冯辅周, 伟等. 基于量子粒子群优化Volterra时域核辨识的隐Markov模型识别方法[J]. 仪器仪表学报, 2011, 32(12): 2693-2698 LI Z N, JIANG J, FENG F Z, et al. Hidden Markov model recognition method based on Volterra kernel identified with Particle Swarm Optimization[J]. Chinese Journal of Scientific Instrument, 2011, 32(12): 2693-2698 (in Chinese)
[14]	魏瑞轩, 韩崇昭. 一种基于Volterra模型的非线性系统具有全解耦结构的递阶式自适应辨识方法[J]. 控制理论与应用, 2002, 19(2): 313-316 doi: 10.3969/j.issn.1000-8152.2002.02.038 WEI R X, HAN C Z. A staged adaptive identification method with fully decoupled structure for nonlinear system based on Volterra model[J]. Control Theory and Applications, 2002, 19(2): 313-316 (in Chinese) doi: 10.3969/j.issn.1000-8152.2002.02.038
[15]	乔立岩, 彭喜元, 马云彤. 基于遗传算法和支持向量机的特征子集选择方法[J]. 电子测量与仪器学报, 2006, 20(1): 1-5 https://www.cnki.com.cn/Article/CJFDTOTAL-DZIY200601001.htm QIAO L Y, PENG X Y, MA Y T. GA-SVM-based feature subset selection algorithm[J]. Journal of Electronic Measurement and Instrumentation, 2006, 20(1): 1-5 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DZIY200601001.htm
[16]	ZHAN Z H, ZHANG J, LI Y, et al. Adaptive particle swarm optimization[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 2009, 39(6): 1362-1381 doi: 10.1109/TSMCB.2009.2015956
[17]	PIRES E J S, MACHADO J A T, DE MOURA OLIVEIRA P B, et al. Particle swarm optimization with fractional-order velocity[J]. Nonlinear Dynamics, 2010, 61(1): 295-301
[18]	刘韬, 陈进, 董广明. KPCA和耦合隐马尔科夫模型在轴承故障诊断中的应用[J]. 振动与冲击, 2014, 33(21): 85-89 https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201421016.htm LIU T, CHEN J, DONG G M. Application of KPCA and coupled Hidden Markov Model in bearing fault diagnosis[J]. Journal of Vibration and Shock, 2014, 33(21): 85-89 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ201421016.htm
[19]	张学工. 关于统计学习理论与支持向量机[J]. 自动化学报, 2000, 26(1): 32-42 https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO200001005.htm ZHANG X G. Introduction to statistical learning theory and support vector machines[J]. Acta Automatica Sinica, 2000, 26(1): 32-42 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO200001005.htm
[20]	化春键, 周海英. 改进组合分类器的冷轧带钢表面缺陷识别研究[J]. 机械科学与技术, 2017, 36(11): 1785-1790 doi: 10.13433/j.cnki.1003-8728.2017.1124 HUA C J, ZHOU H Y. Study on surface defect recognition of cold rolled steel strip by improving combinationclassifier[J]. Mechanical Science and Technology for Aerospace Engineering, 2017, 36(11): 1785-1790 (in Chinese) doi: 10.13433/j.cnki.1003-8728.2017.1124