Blinov I. Fault location in high voltage transmission lines using artificial neural networks

The object of research: local objects of electro power systems (EPS) in particular transmission lines a high voltage (HVTL). The purpose of research: development of new methods of fault classification and location on HVTL using artificial neural networks (ANN’s) at use one-terminal line data. Methods of research: methods of ANN’s theory, methods of mathematical modelling, the theory of relay protection, methods of planning and statistical tests. Theoretical and practical results and innovations: for the first time the method of fault classification on HVTL with use of probabilistic neural networks that has allowed to guarantee correctness of fault classification on HVTL is offered; the method of the fault location on HVTL with use ANN’s that has provided necessary accuracy of fault location on HVTL is improved; types of ANN’s are defined, the choice of ANN’s is proved and requirements to preparation of ANN’s for effective fault classification and location on HVTL are established. A subject and degreeof the application: results of work are introduced in educational process of DNTU and recommended to use in society "Donetskoblenergo". Effectiveness of implantation: increase of accuracy of fault location on HVTL. The sphere of application – operation personnel of EPS.