
This paper proposes a new approach for the classification of the power system disturbances using Wavelet-Multi resolution Decomposition and multi-level Support Vector Machine (SVM).The proposed approach is carried out at different serial stages. First, the original signal is decomposed in to the different level using the wavelet multi-resolution analysis (MRA). This decomposed signal is used to find the energy distribution of the wave. Next, the energy feature is used as input vector for training the SVM classifier. The appropriate input feature is used for classify the PQ event using SVM classifier. Multi-Level SVM technique is used to classify Power disturbances. Hence, Power disturbances are detected and classified with higher efficiency.