Author Email: arpit171294@gmail.com
Abstract
Work presented in this paper elaborates use of Artificial Neural Network (ANN) for detection, classification and location of faults in an interconnected power system. Transmission lines chunks of power to the required loads; it is an essential link between consumers and power stations. Hence transmission lines protection has become critical. A different fault classification technique is proposed in this paper, which yields high accuracy in fault classification with decreased training time and dimension of ANN. The three phase currents and voltages of an end are taken as input to the proposed system. All possible types of faults at various locations were considered. The variation of fault resistance is also considered. The collected data is then used to train the ANN, to locate the fault. The network is modelled and simulated in the MATLAB/Simulink environment. It is apparent in the simulation results obtained that proposed algorithm has satisfactory performance to achieve desired goal of fault identification, classification and determining location.
Introduction
An overhead electric power transmission line is an essential link between the generating plants and different substations. Transmission lines exclusively responsible for conveying electrical power from different generating plants to the grid and from the grid to different substations and from one substation to another substation in a varying degree of voltages in order to meet the extremely large number of load demands. Thus, Overhead power transmission lines are commonly found to be subjected to faults as they carry power over long distances in vividly extreme environmental conditions. [9] Depending on the duration, faults can be segregated into permanent and temporary/transient. Although a transient fault is self-cleared by the line by auto reclosing after a few cycles, it is important to determine fault condition and the location of the same in the transmission line. In the case of a permanent fault that persists even after 2-3 auto recloses, control engineers are equipped to determine fault location enabling maintenance crew to efficiently and quickly take necessary measures. Thus, for optimal utilization of electric power generated and reliability of electric power supplied, it is highly desirable and advantageous to detect, classify, locate and clear these faults at the fastest speed. The artificial neural network is a promising online and offline approach to achieving speedy fault detection, classification, and location in overhead transmission lines. Thus this paper aims to develop a fault analysing algorithm that uses an artificial neural network (ANN) which will detect, classify and locate a fault as it occurs.
Conclusion
Detection, classification and location of the fault in 3-ϕ transmission lines is analysed using ANN. The developed systems utilized inputs as values of instantaneous values of currents and voltages which were processed for detection, fault classification and fault location. The inherent capability of pattern classification and recognition of ANN is utilized in the schemes. ANNs for different application were developed using multilayer perceptron feedforward neural network with backpropagation algorithm. The proposed method is fast, adaptive and accurate due to the application of ANN. Architecture of ANN is found to be accurate, reliable and effective for the problem of detection, classification and location of the faults
1,704 total views, no views today