Fault Detection in HVDC System with Gray Wolf Optimization Algorithm Based on Artificial Neural Network

Abstract

Various methods have been proposed to provide the protection necessitated by the high voltage direct current system. In this field, most of the research is confined to various types of DC and AC line faults and a maximum of two switching converter faults. The main contribution of this study is to use a new method for fault detection in HVDC systems, using the gray wolf optimization method along with artificial neural networks. Under this method, with the help of faulted and non-faulted signals, the features of the voltage and current signals are extracted in a much shorter period of the signal. Subsequently, differences are detected with the help of an artificial neural network. In the studied HVDC system, the behavior of the rectifier, along with its controllers and the required filters are completely modeled. In this study, other methods, such as artificial neural network, radial basis function, learning vector quantization, and self-organizing map, were tested and compared with the proposed method. To demonstrate the performance of the proposed method the accuracy, sensitivity, precision, Jaccard, and F1 score were calculated and obtained as 99.00%, 99.24%, 98.74%, 98.00%, and 98.99%, respectively. Finally, according to the simulation results, it became evident that this method could be a suitable method for fault detection in HVDC systems.