Classification of power quality disturbances using Wavelet Transform and Optimized ANN

This paper presents a new approach to detect and classify the power quality disturbance using Wavelet Transform (WT) based Optimized Artificial Neural Network (ANN). The proposed algorithm extracts the energy based feature vector consisting of approximation and detail coefficients of WT. ANN based classifier is used to classify the power quality (PQ) disturbances. Six different types of PQ disturbances are considered to examine the versatility of the proposed approach. Furthermore, a novel and innovative approach is used to optimize the weights of ANN using Differential Evolution (DE). The optimized ANN results demonstrate the superiority, accuracy and robustness of the proposed approach compared to the reported techniques in literature. The comparisons demonstrated that the proposed approach is more superior in terms of classification error reduction and overall accuracy improvement.