基于ISSA-BP神经网络的混凝土声发射定位算法  

Research on Acoustic Emission Localization in Concrete Based on ISSA-BP Neural Network

作　　者：黄振宇 郭涛[1] 时英元 HUANG Zhenyu;GUO Tao;SHI Yingyuan(Key Laboratory of Electronic Testing Technology for National Defense Science and Technology,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学电子测试技术国防科技重点实验室,太原030051

出　　处：《自动化与仪器仪表》2025年第1期19-23,共5页Automation & Instrumentation

基　　金：国家自然科学基金面上项目:极端环境下零维多相防护结构连续多脉冲动力学响应机理及测试技术研究(52375553)。

摘　　要：针对现阶段声发射技术在混凝土结构健康检测中由于衰减严重而导致的损伤源精度低问题,提出了一种基于改进麻雀搜索算法来优化BP神经网络(ISSA-BP)用于混凝土声发射定位的方法。首先采用BP人工神经网络作为混凝土声发射定位的基础框架,通过对声发射信号进行小波降噪,并将信号的峰值时间,时间差,能量和幅值作为网络的输入参数,然后利用ISSA对BP神经网络的权重和偏置进行优化。为了验证该方法的有效性采用0.5 mm硬度为2H的铅芯进行断铅实验,实验结果表明所提方法算法比传统BP神经网络和麻雀搜索算法(SSA)优化神经网络在混凝土损伤源声发射定位上平均误差分别减少了43.47%、14.02%,最大误差减少了56.52%、4.41%,最小误差减少了42.82%、72.42%,具有一定的工程应用价值。Due to the significant attenuation in acoustic emission technology for concrete structure health monitoring,which leads to low accuracy in damage source localization,this paper proposes a method based on an improved sparrow search algorithm to optimize the BP neural network(ISSA-BP)for acoustic emission localization in concrete.Firstly,the BP artificial neural network is used as the basic framework for concrete acoustic emission localization.The acoustic emission signals are denoised using wavelet transform,and the peak time,time difference,energy,and amplitude of the signals are taken as the input parameters of the network.Then,the improved sparrow search algorithm is employed to optimize the weights and biases of the BP neural network.To verify the effectiveness of this method,a lead break experiment using a 0.5mm 2H hardness lead core was conducted.The experimental results show that the proposed algorithm reduces the average error in acoustic emission localization of concrete damage sources by 43.47%and 14.02%compared to the traditional BP neural network and SSA-BP neural network algorithms,respectively.The maximum error is reduced by 56.52%and 4.41%,and the minimum error is reduced by 42.82%and 72.42%,demonstrating significant engineering application value.

关 键 词：ISSA-BP 声发射定位 混凝土 BP神经网络 

分 类 号：TP212.1[自动化与计算机技术—检测技术与自动化装置]