Application of Least Squares Support Vector Machine in the Damage Identification of Plate Structure

In order to study the application validity and veracity of regression technique of LS-SVM (Least Squares Support Vector Machine)in damage identification for plate structure, a numerical simulation of plate structure enduring great impact loading is built. Through the numerical simulation, the training samples are get for building the regression model of LS-SVM. Then the simulation experiments are did with the model which is optimized based on Bayesian framework, the results of simulation experiments shown that, the method can identify the damage location exactly, and can accurately identify the damage extent in a given range. For using structural natural frequency as the input parameter, the cost of health monitoring system and complexity of installation in project are lowered, so it has good worth to extend and employ.