IDENTIFICATION OF FAULT AND ITS LOCATION USING NEURAL NETWORK IN DISTRIBUTION NETWORK WITH DISTRIBUTION GENERATION

Abstract

There are basically two ways by which electricity can be generated. One is by centralized generation scheme where electricity is generated far from the load and is supplied to the consumer using transmission and distribution network while the other is distributed generation in which electricity is generated near to the load center. These days distributed generation is quickly being implemented by most of the countries because of easy installation, transmission and distribution cost reduction, improve rural electrification system, improved energy efficiency, enhanced reliability and security, electricity deregulation and less impact on environment. Despite of several advantages associated with DG it seriously affects the protection scheme of an existing network. Distributed generation may affect the operations in a distribution network protection scheme that may lead to various unwanted situation such as reclosing failure. Also sometimes components which are not affected and are healthy can be disconnected from the system and it may even bring obstacles in carrying out protection procedure. Therefore it is very important to protect the system with DG so as to ensure continuous supply and safety of the equipment and personal staff. This works aims to identify the fault and its location using neural network in distribution network with DG. For this purpose, a distribution network having Distributed Generator has been considered for data generation. At different location of the network, different types of fault have been simulated and voltages as well as current samples have been collected. This collected data have been used to train and test the neural network. On successful training & testing of neural network, it identifies the type and location of fault.