Intelligent Diagnosis and Evaluation of MOA's Operating Status

Metal oxide arrester (MOA) is the critical power equipment to ensure the long-term stable operation of power system. In view of the low accuracy, high lag and poor sensitivity of traditional MOA's diagnosis and evaluation methods, this paper proposed an intelligent evaluation method of MOA's operating status by combining the fuzzy comprehensive evaluation method and neural network analysis. Firstly, from the perspective of MOA's characteristic parameters and the environmental factors, the evaluation parameters were determined as full current, resistive current, ambient temperature and ambient humidity. Secondly, based on the fuzzy theory and cloud theory, the cloud model of membership degree was established by fuzzy comprehensive evaluation method, and the operating status of MOA was evaluated preliminarily. Finally, based on the deep learning theory, the evaluation results of fuzzy comprehensive evaluation method were taken as known quantities, and they were used to train the neural network and evaluate the operating status of MOA. The result shows that this intelligent evaluation method has an accuracy of more than 90%, therefore, it can be used for the accurate diagnosis of MOA's operating status.