改进声发射信号的桥梁焊缝裂纹识别仿真研究
Simulation Research on Bridge Weld Crack Identification Based on Improved Acoustic Emission Signal
李敏峰 1周小龙 2徐永峰2
作者信息
- 1. 河北建筑工程学院土木工程学院,河北 张家口 075000;兰州交通大学土木工程学院,甘肃 兰州 730070
- 2. 河北建筑工程学院土木工程学院,河北 张家口 075000
- 折叠
摘要
针对因环境中存在过多噪声,导致桥梁焊缝裂纹识别精准度低的问题,提出基于声发射信号的桥梁焊缝裂纹识别方法.利用传感器提取桥梁周围的实时信号,通过信号在周期序列上的幅值变化,判定噪声信号,采用小波变换算法对噪声信号实施重构变换,建立硬阈值和软阈值函数,约束噪声信号.采用神经元传递函数计算原始信号序列中隐含层神经元的具体特征表现参数,得到信号的特征类间平均值,通过类间参数求得特征量.以带有声发射信号提取技术的传感器作为识别载体,将特征参数输入到识别传感器中,针对不同的桥梁测试点,建立焊缝裂纹识别通道,完成有效识别.实验结果证明,所提方法的识别精准度较高,无论是以持续频率还是持续时间信号作为测试指标,均能实现高效识别.
Abstract
Aiming at the problem of low accuracy of bridge weld crack identification,this paper presented a meth-od of identifying the weld crack of a bridge based on acoustic emission signal.At first,a sensor was used to extract the real-time signal around the bridge,then the noise signal could be determined by the amplitude change of the signal in the periodic sequence.Moreover,the wavelet transform algorithm was adopted to reconstruct the noise signal.And then,hard threshold and soft threshold functions were established to constrain the noise signal.Furthermore,neuron transfer functions were used to calculate the specific characteristic performance parameters of the hidden-layer neu-rons in original signal sequences,and thus to obtain the mean value between feature classes of signals.Based on the inter-class parameters,the feature quantity was obtained.After that,the sensor with acoustic emission signal extraction technology was used as the identification carrier.Meanwhile,the feature parameters were input into the i-dentification sensor.Finally,the channel for identifying weld cracks was constructed for different bridge test points.Thus,the effective identification was achieved.Experimental results show that the proposed method has high recogni-tion accuracy,and can achieve efficient recognition regardless of whether the continuous frequency or the duration sig-nal is used as the test index.
关键词
声发射信号/桥梁焊缝裂纹/硬阈值/神经元传递函数/隐含层神经元Key words
Acoustic emission signal/Weld crack of bridge/Hard threshold/Neuron transfer function/Hidden layer neuron引用本文复制引用
基金项目
河北省省属高校基本科研业务费研究项目(2022QNJS07)
河北省省属高校基本科研业务费研究项目(2022QNJS06)
河北省教育厅青年基金(QN2020440)
出版年
2024
NETL
NSTL
万方数据