Detection And Classfication Of Power Quality Disturbances Based On S-Transform And Neural Network

With the rapid increasing of devices sensitive to power quality, more and more attentions are paid to power quality problems by power unit and users. So how to detect and classify the power quality disturbances automatically and accurately is the necessary process to construct integrated power quality monitoring system and the premise to improve power qualityThis paper reviews the methods used in the power quality disturbance analysis firstly, and secondly, that the characteristics and the causes of different power quality disturbances are analyzed in detail. Based on Matlab7.4, in this paper, an S-transform-based resilient back propagation neural network structure (RPROP) is adopted for automatic classification of power quality disturbances. Firstly, Through S-transform time-frequency analysis, the method detects and outputs kinds of PQ disturbances effectively. Then, feature components are extracted from the detecting outputs for classification. Secondarily, an optimum combination of the most useful features is identified for increasing the accuracy of classification. Features extracted by using the S-transform are applied as input to NN for automatic classification of the power quality (PQ) disturbances that solves a relatively simple problem. Sensitivities of the classifier under different noise conditions are also investigated.The results of simulation show that the proposed detection method could detect various power quality disturbances quickly and correctly; the classification method possesses higher recognition rate, more kinds of recognized disturbances, much stronger resistance to noises; the classification method of the sources of voltage sag is effective.