Application of Reproducing Kernel Support Vector Machinein Fault Diagnosis of Tool Wea

A new reproducing kernel function of least square support vector machines based on Walsh series is presented in this paper. First, the reproducing kernel is constructed in reproducing kernel Hilbert space (RKHS). According to the wavelet multi-resolution analysis, Walsh consequence can be seen as a set of orthogonal basis to construct the reproducing kernel.Second, the features are extracted from different frequency segment with the technology of wavelet packet frequency segment power decomposition, which reflects different working state of the cutting tool, and th ese features are taken as inputs fault of support vector machine (SVM) classifier.The simulation result shows that reproducing kernel SVM can accurately divide the working conditions of a tool.The trained classifier, as fault intelligent classification, has very strong identification capability.