Fuzzy Neural Network for Classification Fault in Protection System

Novel intelligent technique is a combination of fuzzy and neural network techniques that can be used to classify faults in electric power system protection. There have two problems in the protection system, which are: undesired tripping and fail to operate. Loss of power supply to relays and circuit breakers or failure in protective devices may cause failures in protection system. Construction of neural networks to explore fact to identify fault component is from control center. The objective of this paper is to develop novel concept for classification failures protection system are using Fuzzy Neural Network technique. Methodology consists of Neural Network and Fuzzy. The Neural network is also conscientious for estimating degree of membership in system components from corresponding area in classification of disorders. The input variables of neural network formed of binary numbers. Value of 1 indicates if fault occurs and value of 0 indicates no-fault occurs. Fuzzy relations will represent by fuzzy. These Fuzzy relations can be represented by fuzzy diagram consisting of three sets of node that would be considered to represent components, relays and circuit breakers. Fuzzy diagram is built as causal operation of relays and circuit breakers on event of the fault in protection system. The causality is represented in arrows. Finally, the concept of Fuzzy Neural Network can be proposed as alternative to solve issue of failures that occur in the protection system.