PLS的红外无损检测电磁激励的数学模型

Infrared nondestructive testing electromagnetic excitation model based on partial least-squares regression

主动式电磁激励红外无损检测中,电磁激励参数的选择是决定无损检测效果的重要因素之一。研究首先通过正交实验分析了激励功率、线圈总长、线圈等效直径、提离距离、激励时间等电磁激励参数对材料激励热效应的影响;基于偏最小二乘构建了各参数与材料激励热效应的数学模型,并分析了二者之间的关系;最后利用15组验证样本对模型进行了验证。研究结果表明;所构建数学模型的平均误差为9.79%,各激励参数中,提离距离对材料激励热效应的影响最大,其次分别为线圈总长、线圈等效直径、激励时间和激励功率。

In the method of active electromagnetic excitation infrared nondestructive detection, the selection of electromagnetic excitation parameter is one of the important factors that determine nondestructive testing effect. Firstly, the relation between electromagnetic excitation parameters （excitation power, coil length, coil equivalent diameter, lift- off distance, excitation time, etc） and excitation thermal effect was analyzed through orthogonal test. The mathematical model between parameters and excitation thermal effect was built based on the partial least- squares regression. Lastly, the model was verified by 15 experiment samples. The result shows that the average error of mathematical model is 9.79%; lift- off distance affects most among the excitation parameters, followed by coil length, coil equivalent diameter, excitation time and excitation power.