BP网络算法在脉冲涡流缺陷检测中的应用

Application of BP algorithm in pulse eddy current defect detection

脉冲涡流法是五大常规无损检测技术之一,在脉冲涡流检测中采用BP网络算法,旨在减轻提离效应的影响。在matlab中搭建实验平台,通过数据分析,记录不同大小缺陷的检测数据,提取相关特征量,利用BP神经网络分析这些特征量与不同缺陷、不同提离之间的关系。最后通过一组缺陷测试,证实了此方案的可行性。

Pulse eddy current method is often used for nondestructive testing of sheet metal defects.Lifting-off effect has always been a difficult problem to be solved in eddy current testing.Therefore,an experimental platform is built to records the detection data of different size defects through the analysis of MATLAB data and extract the relevant characteristic quantities.And uses BP neural network is used to analyze the relationship between these characteristic quantities and different defects and different lifting-off.After that,the feasibility of this scheme is proved by a set of defects test.