Automatic detection and classification of electrical disturbances by means of empirical mode decomposition method

Due to the negative impact on the equipment, monitoring of electrical disturbances has become a topic of interest for many researches around the world. In this work, a methodology for automatic classification of power quality disturbances (PQD) is proposed. It consists of three stages: first, the empirical mode decomposition for signal processing is applied; second, the entropy and energy are computed as features for pattern recognition and, finally, a neural network performs the automatic classification. The overall methodology is developed using Matlab software. Synthetic signals are used to train and validate de proposal. On the other hand, real measurements using a field programmable gate array (FPGA)-based data acquisition system (DAS) are carried out to test and show its effectiveness under real operating conditions. The obtained results show high accuracy and low computational burden.