Please wait a minute...
工程设计学报  2023, Vol. 30 Issue (5): 562-570    DOI: 10.3785/j.issn.1006-754X.2023.00.066
机械设计理论与方法     
基于声信号识别的水下焊接质量检测方法研究
纪晓东(),程天宇(),华亮,张新松
南通大学 电气工程学院,江苏 南通 226000
Research on detection method of underwater welding quality based on acoustic signal recognition
Xiaodong JI(),Tianyu CHENG(),Liang HUA,Xinsong ZHANG
College of Electrical Engineering, Nantong University, Nantong 226000, China
 全文: PDF(4139 KB)   HTML
摘要:

水下焊接的应用领域广泛,但其焊接质量难以保障。针对水下焊接处理过程中存在的非线性程度高、参数耦合性强以及检测效率低等问题,提出了一种新的基于声信号识别的水下焊接质量检测方法。该方法通过在水下构建基于声信号采集的监测系统,实时采集焊件焊接过程中的声信息,并通过对声信号进行滤波降噪处理和特征提取,构建双权值神经网络(double-weight neural network, DWNN)模型。该模型具有优秀的高维数据非线性拟合能力,可实现水下焊接多参数与声信号多特征之间的非线性映射,且在小样本情况下仍能实现高精度的模式识别。以高强度低碳合金钢——HSLA-115钢作为焊接对象,开展水下焊接质量检测实验。结果表明,DWNN模型应用于水下焊接质量检测的识别精度可达100%。研究结果可为水下焊接工艺的优化和水下焊件专家知识库的构建提供参考依据。

关键词: 水下焊接声信号特征提取双权值神经网络    
Abstract:

Underwater welding is widely used in many fields, but its welding quality is difficult to guarantee. Aiming at the problems of high nonlinearity, strong parameter coupling and low detection efficiency in underwater welding process, a new method for underwater welding quality detection based on acoustic signal recognition was proposed. The method constructed a monitoring system based on acoustic signal acquisition underwater to collect the acoustic information during the welding process of weldments in real time, and built a double-weight neural network (DWNN) model through filtering and noise reduction processing and feature extraction for the acoustic signal. The model had excellent nonlinear fitting ability of high-dimensional data and could realize nonlinear mapping between multi-parameters of underwater welding and multi-features of acoustic signals, and it could still realize high-precision pattern recognition in the case of small samples. The underwater welding quality detection experiments were carried out with high strength and low carbon alloy steel—HSLA-115 steel as welding object. The results showed that the recognition accuracy of DWNN model applied to underwater welding quality detection could reach 100%. The research results can provide reference for the optimization of underwater welding process and the construction of underwater weldparts expert knowledge base.

Key words: underwater welding    acoustic signal    feature extraction    double-weight neural network
收稿日期: 2023-03-02 出版日期: 2023-11-03
CLC:  TG 456.5  
基金资助: 国家自然科学基金资助项目(51877112)
通讯作者: 程天宇     E-mail: 1731571044@qq.com;chengtianyu518@163.com
作者简介: 纪晓东(1998—),男,江苏盐城人,硕士生,从事模式识别研究,E-mail: 1731571044@qq.com
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  
纪晓东
程天宇
华亮
张新松

引用本文:

纪晓东,程天宇,华亮,张新松. 基于声信号识别的水下焊接质量检测方法研究[J]. 工程设计学报, 2023, 30(5): 562-570.

Xiaodong JI,Tianyu CHENG,Liang HUA,Xinsong ZHANG. Research on detection method of underwater welding quality based on acoustic signal recognition[J]. Chinese Journal of Engineering Design, 2023, 30(5): 562-570.

链接本文:

https://www.zjujournals.com/gcsjxb/CN/10.3785/j.issn.1006-754X.2023.00.066        https://www.zjujournals.com/gcsjxb/CN/Y2023/V30/I5/562

图1  水下焊接实验平台
实验参数(焊接速度、焊接电流、板材厚度)水下焊接处理结果
1.2 m/min、110 A、6 mm合格
0.6 m/min、110 A、6 mm合格
0.6 m/min、140 A、6 mm不合格
1.2 m/min、110 A、8 mm不合格
1.2 m/min、140 A、8 mm合格
1.2 m/min、110 A、10 mm不合格
1.2 m/min、160 A、10 mm合格
1.2 m/min、160 A、12 mm不合格
0.6 m/min、160 A、12 mm合格
0.6 m/min、180 A、12 mm合格
表1  部分水下焊接实验的测定结果
图2  水下焊接钢板实物
图3  Butterworth滤波器的幅值特性曲线
图4  滤波处理前后声信号的频谱图对比
图5  声信号的短时能量
图6  声信号的短时平均幅度
图7  声信号的短时平均过零率
图8  结构固定的DWNN
图9  引入PSO的DWNN算法流程
图10  水下焊接质量检测用DWNN的结构
图11  基于不同方法的水下焊接质量识别结果(100组训练样本)
图12  基于不同方法的水下焊接质量识别误差(100组训练样本)
识别模型平均误差最大误差均方误差
DWNN0.002 20.011 40.000 387
SVM0.075 20.988 30.011 039
BPNN0.062 50.876 10.008 856
表2  不同水下焊接质量识别模型的误差对比
图13  基于不同方法的水下焊接质量识别结果(20组训练样本)
图14  基于不同方法的水下焊接质量识别误差(20组训练样本)
1 郭俊良.局部干法水下镍基合金焊接工艺及性能研究[D].哈尔滨:哈尔滨工程大学,2020:11-12. doi:10.7498/aps.69.20191854
GUO J L. Research of underwater local dry welding process and properties of nickel-based alloy[D]. Harbin: Harbin Engineering University, 2020: 11-12.
doi: 10.7498/aps.69.20191854
2 岳国柱,王国安,徐武彬,等.焊接方向对H型构件焊后变形的影响研究[J].机械设计与制造,2022(11):122-125. doi:10.3969/j.issn.1001-3997.2022.11.026
YUE G Z, WANG G A, XU W B, et al. Research on the influence of welding direction on the deformation of H-shaped members[J]. Machinery Design & Manufacture, 2022(11): 122-125.
doi: 10.3969/j.issn.1001-3997.2022.11.026
3 Creaform.海上风电导管架的焊接检测和无损检测[J]. 现代制造,2022():56-57.
Creaform. Welding test and nondestructive test of offshore wind turbine jacket[J]. Maschinen Markt, 2022(): 56-57.
4 肖书浩,陈涵,吴蕾.一种陶质焊接衬垫产品质量检测方法[J].机械设计与制造,2021(11):139-142,146. doi:10.3969/j.issn.1001-3997.2021.11.032
XIAO S H, CHEN H, WU L. Method for detecting quality of ceramic welding gasket product[J]. Machinery Design & Manufacture, 2021(11): 139-142, 146.
doi: 10.3969/j.issn.1001-3997.2021.11.032
5 KIM S, HWANG I, KIM D Y, et al. Weld-quality prediction algorithm based on multiple models using process signals in resistance spot welding[J]. Metals, 2021, 11(9): 1459.
6 房海基,吕波,张艳喜,等.焊接过程声信号在线检测技术现状与展望[J].精密成形工程,2022,14(1):165-172. doi:10.3969/j.issn.1674-6457.2022.01.021
FANG H J, LÜ B, ZHANG Y X, et al. Status and prospect of on-line acoustic signal detection technology in welding[J]. Journal of Netshape Forming Engineering, 2022, 14(1): 165-172.
doi: 10.3969/j.issn.1674-6457.2022.01.021
7 刘亮.基于多信息融合的GTAW过程熔透状态识别及预测控制[D].上海:上海交通大学,2021:3-5.
LIU L. Penetration state recognition and predictive control of GTAW process based on multi-information fusion[D]. Shanghai: Shanghai Jiaotong University, 2021: 3-5.
8 兰兴川.基于神经网络优化案例推理的焊接质量检测方法研究[D].重庆:重庆邮电大学,2021:13-15.
LAN X C. Research on welding quality inspection method based on neural network optimization case-based reasoning[D]. Chongqing: Chongqing University of Posts and Telecommunications, 2021: 13-15.
9 王春香,郝林文,王耀,等.基于GA-BP神经网络的散乱点云孔洞自动修补[J].工程设计学报,2021,28(2):155-162. doi:10.3785/j.issn.1006-754X.2021.00.029
WANG C X, HAO L W, WANG Y, et al. Automatic repair of scattered point cloud hole based on GA-BP neural network[J]. Chinese Journal of Engineering Design, 2021, 28(2): 155-162.
doi: 10.3785/j.issn.1006-754X.2021.00.029
10 王守觉,李兆洲,陈向东,等.通用神经网络硬件中神经元基本数学模型的讨论[J].电子学报,2001,29(5):577-580. doi:10.3321/j.issn:0372-2112.2001.05.001
WANG S J, LI Z Z, CHEN X D, et al. Discussion on the basic mathematical models of neurons in general purpose neurocomputer[J]. Acta Electronica Sinica, 2001, 29(5): 577-580.
doi: 10.3321/j.issn:0372-2112.2001.05.001
11 庞轶环,胡志忠.一种分数阶巴特沃斯滤波器的有源电路设计[J].电子学报,2018,46(5):1160-1165. doi:10.3969/j.issn.0372-2112.2018.05.021
PANG Y H, HU Z Z. Active circuit design of a fractional order Butterworth filter[J]. Acta Electronica Sinica, 2018, 46(5): 1160-1165.
doi: 10.3969/j.issn.0372-2112.2018.05.021
12 HUA L, ZHENG C W, GU J P, et al. Laser arc sound signal processing and welding status recognition based on geometric learning[C]//International Conference on Manufacturing Science & Engineering, Paris: Atlantis Press, 2015: 1383-1393.
13 曹宇,赵星涛.一种新型双权值人工神经元网络的数据拟合研究[J].电子学报,2004,32(10):1671-1673. doi:10.3321/j.issn:0372-2112.2004.10.019
CAO Y, ZHAO X T. Data fitting based on a new double weights neural network[J]. Acta Electronica Sinica, 2004, 32(10): 1671-1673.
doi: 10.3321/j.issn:0372-2112.2004.10.019
14 李树松.粒子群算法在优化问题中的应用研究[J].科学技术创新,2020(32):103-104. doi:10.3969/j.issn.1673-1328.2020.32.047
LI S S. Application of particle swarm optimization algorithm[J]. Scientific and Technological Innovation, 2020(32): 103-104.
doi: 10.3969/j.issn.1673-1328.2020.32.047
15 郑煜.基于PSO和MSR的微弱信号检测方法研究[J].机电工程,2022,39(3):362-367. doi:10.3969/j.issn.1001-4551.2022.03.012
ZHENG Y. Weak signal detection method based on PSO and MSR[J]. Journal of Mechanical & Electrical Engineering, 2022, 39(3): 362-367.
doi: 10.3969/j.issn.1001-4551.2022.03.012
16 杨寒雨,赵晓永,王磊.数据归一化方法综述[J].计算机工程与应用,2023,59(3):13-22. doi:10.3778/j.issn.1002-8331.2207-0179
YANG H Y, ZHAO X Y, WANG L. A review of data normalization methods[J]. Computer Engineering and Applications, 2023, 59(3): 13-22.
doi: 10.3778/j.issn.1002-8331.2207-0179
17 张乐天,张莉,宋倩,等.改进粒子群算法的单杆柔性臂振动抑制方法研究[J/OL].机械科学与技术,2022:1-6(2022-10-14)[2023-03-02]..
ZHANG L T, ZHANG L, SONG Q, et al. Research on optimal control method of single flexible manipulator by improved particle swarm optimization algorithm[J]. Mechanical Science and Technology for Aerospace Engineering, 2022: 1-6 (2022-10-14)[2023-03-02]. .
18 魏海茹,李冬梅,褚红瑞,等.基于改进粒子群算法的暴雨天气实时预警方法[J].信息技术,2022,46(11):95-99,105.
WEI H R, LI D M, CHU H R, et al. Real-time rainstorm warning method based on improved particle swarm optimization[J]. Information Technology, 2022, 46(11): 95-99, 105.
19 颜志浩.基于分阶段改进组合策略的粒子群优化算法研究[D].武汉:华中科技大学,2020:27-29.
YAN Z H. Study on particle swarm optimization algorithm based on the combinatorial strategy of phased improvement[D]. Wuhan: Huazhong University of Science and Technology, 2020: 27-29.
20 武娇,洪彩凤,顾永春,等.基于类邻域字典的线性回归文本分类[J].计算机工程,2021,47(8):93-99,108.
WU J, HONG C F, GU Y C, et al. Linear regression text classification based on class-wise nearest neighbor dictionary[J]. Computer Engineering, 2021, 47(8): 93-99, 108.
[1] 李梦,尹宗军. 一种零件综合质量评定方法研究[J]. 工程设计学报, 2022, 29(4): 410-418.
[2] 陈一帆,吴倩,蒋凌,华亮. 基于声信号识别的焊后残余应力处理质量检测方法[J]. 工程设计学报, 2022, 29(3): 272-278.
[3] 王春香, 李乐, 王建国. 基于提取特征的挖掘机斗齿的几何反求和模型重建[J]. 工程设计学报, 2010, 17(3): 211-214.