Faults detection, location, and classification of the elements in the power system using intelligent algorithm

Abstract

This study proposes an intelligent protection relay design that uses artificial neural networks to secure electrical parts in power infrastructure from different faults. Electrical transformer and transmission lines are protected using intelligent differential and distance relay, respectively. Faults are categorized, and their locations are pinpointed using three-phase current values and zero-current characteristics to differentiate between non-earth and ground faults. The optimal aspects of the artificial neural network were chosen for optimal results with the least possible error. Levenberg-Marquardt was established as the ideal training technique for the suggested system comprising the differential relay. Levenberg-Marquardt was the optimal training technique for the proposed framework consisting of the differential relay. Fault detection and categorization were performed using 20 and 50 hidden layers, and the corresponding error rates were 9.9873e-3 and 1.1953e-29. In the context of fault detection by the distance relay, the hidden layer neuron counts were 400, 250, and 300 for fault detection, categorization, and location; training error rates were 7.8761e-2, 1.2063e-6, and 1.1616e-26, respectively.

Application of back propagation neural network in complex diagnostics and forecasting loss of life of cellulose paper insulation in oil-immersed transformers

“…Twenty-six specific transformers utilized in distribution substations in South Africa, Mpumalanga, were exposed to the proposed diagnostics. The diagnostics were divided into five phases, which are described below 25 , 65 – 68 : Assessment of the transformer’s current condition: involves visual inspection and all DGA and chemical tests conducted on transformer oil. Gathering historical data: involves gathering information from prior faults, repairs, and maintenance.…”

“…A successful network will have an R -value close to 1, indicating that there is a notable relationship between the BPNN network output and the targeted output. The overall efficiency of the BPNN network is expressed by the value of R and the best BPNN network is identified based on its closest relationship value to 1 68 , 95 . The Levenberg–Marquardt (LM) training approach 64 , 65 was used since it is recommended as the preferable supervised algorithm in the MATLAB environment due to its fast training rate 116 .…”

Transformer faults identification via fuzzy logic approach

“…It plays a key role in the survival of our educational, health, financial, and industrial sectors and also in our homes which we use to solve our daily needs [1]. Modern electrical power systems are complicated networks that usually require highly and more précised protection schemes that would prevent them from failing or breaking down completely [2], [3]. Transformer being the substantial and crucial element of an electrical system, is subjected to a series of problems, resulting in power outages, which when they occur, can affect the economic and social development of any nation across the globe [4].…”