Eddy current characterization using robust meta-heuristic algorithms for LS-SVM hyper-parameters optimization

Abstract

This paper presents the use of the Least Square Support Vector Machines (LS-SVM) technique, combined with the Finite Element Method (FEM), to characterize small cracks in order to get a fast non-destructive inspection. The LS-SVM is a statistical learning method that has good generalization capability and learning performance. LS-SVM trained model is proposed to predict crack sizing using experimental signals acquired from an Eddy Current (EC) sensor. The FEM is used to create the data set required to train this model. The performance of LS-SVM model depends on a careful setting of its associated hyper-parameters. Different tuning techniques for optimizing the LS-SVM hyper-parameters are studied: Electromagnetism-Like Mechanism (EM), Opposition Based Electromagnetism-Like Mechanism (OBEM) and Teaching Learning Based Optimization (TLBO). Results show that TLBO algorithm provides a good compromise between accuracy and computational cost.