No abstract

No abstract

Modern electric power facilities – stations and high-voltage substations – have become digital objects with the active use of high-speed local networks directly involved in the technological process. Management, analysis and control of information exchange in the digital substation of the power system require the development of new means and approaches. For these purposes, machine learning methods can be used, in particular the apparatus of artificial neural networks (ANN). The paper shows the possibilities of using direct propagation ANNs (multilayer perceptrons) for modeling and identifying anomalies in the operation modes of relay protection with a time delay. The results of training and testing of the ANN are presented on the example of analyzing the operation of the over current protection in the “sliding time window” mode in a three-phase electrical network. The proposed neuroalgorithm and configuration of the ANN can be used to control the modes and accuracy of relay and cybernetic defenses.

Modern electric power facilities-stations and high-voltage substations have become digital objects with the active use of high-speed local networks directly involved in the technological process. Management, analysis and control of information exchange in the digital substation of the power system require the development of new tools and approaches. For these purposes, machine learning methods can be used, in particular, the artificial neural networks. The paper presents the results of neural network modeling of the operation of the overcurrent protection – as a variant of the information exchange analysis. An elementary perceptron is used as a neural network with the simplest structure. The optimized structure of the neural network and estimates of the accuracy of the neural network algorithm are given, depending on the size of the training sample (from 1000 to 50000 records), the number of training epochs. It is shown that the analysis of the neural network algorithm errors encountered during testing of the neural network enables to estimate the threshold (the setting value) current protection depending on the size of the training sample. It is found that the recognition of the protection trigger threshold in neural network modeling is violated only when the all three phase currents in electrical mains are close to the threshold. The possibilities of improving the proposed approach and its use for detecting anomalies in the information exchange and operation of secondary equipment of digital substations of the power system are discussed.