Повышение надежности сельских распределительных электрических сетей с применением нейронных сетей

The article discusses the issues of improving the reliability of rural electrical distribution networks using neural networks. The main problem of the operation of rural distribution electric networks is the increased wear of the main elements, which are characterized by frequent failures, which entail economic losses from the shortage of electricity and damage to consumers. The authors investigated the problem areas of rural distribution electric networks on the basis of real reports of electric grid organizations, took into account the normalized indicators of the reliability of electric networks and determined the most promising method of increasing reliability using neural networks. Changes to the current system of scheduled preventive maintenance with the determination of calculated indicators based on diagnostic and complex reliability parameters are proposed. On their basis, it is planned to implement an algorithm for the operation of the neural network, which will predict the generated indicators and determine the list of necessary maintenance and repair work for the next calendar year.