变温度条件下基于代理模型的Lamb波裂纹定量诊断方法研究

Study on Quantitative Diagnosis Method of Lamb Waves for Cracks under Varying Temperature Conditions Based on Surrogate Models

高速列车运行环境复杂多变,因循环载荷作用内部结构会产生裂纹等损伤。为实现不同温度环境中对列车结构裂纹的实时检测和诊断,本文研究了-40℃~85℃的环境中Lamb波裂纹定量诊断方法。将试验材料置于不同的温度梯度下,进行Lamb波信号的激励和收集,接着进行变温条件下Lamb波裂纹定量监测试验,获得了不同温度和裂纹长度条件下的Lamb波信号,通过归一化幅值提取了相位变化和相关系数3个特征值。将特征值作为Kriging代理模型的训练样本,生成基于特征值的响应面构建诊断参数的三维模型,实现了不同温度条件下裂纹的定量诊断。

High-speed trains operate in complex and variable environments,in which cyclic loading will result in cracking and other damages to the internal structure of trains.To enable real-time detection and diagnosis of structural cracks in trains under diverse temperature environments,this study investigated a quantitative diagnostic method of Lamb waves for cracks over the temperature range of-40℃to 85℃.The specimens were subjected to various temperature gradients,followed by the excitation and acquisition of Lamb wave signals under these conditions.Subsequently,a temperature-dependent quantitative monitoring experiment for Lamb wave-assisted crack detection was conducted,yielding Lamb wave signals corresponding to different temperatures and crack lengths.Normalized amplitude values were used to extract three characteristic values representing phase shifts and correlation coefficients.These characteristic values served as training samples for a Kriging surrogate model,which generates a response surface based on the feature values to construct a three-dimensional model of diagnostic parameters.This model enables quantitative diagnosis of cracks under varying temperature conditions.