Application of the Combinatorial Model of Wavelet Analysis and Support Vector Machines in Deformation Analysis

Abstract

Engineering distortion analysis model is numerous, each has application characteristics and limitations. The combinatorial model used to deformation analysis can avoid the limitation of single model, and it can enhance the accuracy of prediction. Combining wavelet analysis with support vector machines (SVM) and introducing wavelet function to structure the kernel function of support vector machines, which establish a combinatorial model based on wavelet support vector machines. The model has the advantages of support vector machines, which can further understanding of the characteristics of the engineering distortion. And it also can eliminate the high frequency interference of data, so it has the good noise immunity. The results of the experiment show that the combinatorial model has good fit precision and prediction ability of support vector machines, which used for deformation analysis is feasible and effective.