Application Research Of Independent Component Analysis And Neural Network On Fault Detection Of Electric System

Electric system has become more and more important for our everyday life. They are indispensable to generate and provide electricity power for domestic appliance, public service, industrial machineries and other modern society demands. Especially, the distributed electric systems are often in the large-sized, such as offshore power system. Any failure of the equipment not only leads to loss in money in terms of production and time, but may lead to loss people's life. Detection of faults in those electric systems before the eventual breakdown of the system is very important as it can reduce the costs incurred from equipment failures. In order to make the electric system continues work well, it is necessary to study the fault detection as well as the maintenance of these systems using advance methods.Partial discharge is one of the main causes of breakdown of distributed energy system. This paper proposed a novel method for fault detection by analyzing the different partial discharge signals after various normal techniques for fault detection have been investigated. This method based on the principles of independent component analysis (ICA) and neural network. Using this method, it is possible to sieve through large amounts of data that contains relatively small amounts of useful information and extract out the useful one from the data.Independent Component Analysis (ICA) is an approach for blind signal separation,which has been developed during the past twenty years. It is a statistical method, which aims to recover independent original signals from observed signals given by sensors which are linear mixtures of the independent original signals, and attempts to make the separated signals as independent as possible. It is widely used in signal processing such as speech recognition system, telecommunications, and medical signal processing. Neural Networks is a cross-discipline which integrates neural science, information science, computer science, and it has been developing rapidly in recent years. Neural Networks is an information processing system through abstracting, simplifying, simulating the structure, function and other aspects of biological Neural Networks theory. With its unique advantages, Neural Networks has been applied to several areas, such as signal processing, feature extraction, pattern recognition, etc. Satisfied performance has been achieved, which arouses the attention of many researchers.In this paper, researches on fault detection in electric system have been done, which is based on ICA. Furthermore, the fault detection techniques combined with ICA and Neural Networks have also been studied. The research results demonstrate that this method is not only easy to apply but also can quickly identify the type and the location of partial discharge activities. This paper supplies a feasible method for detection of faults in electric system.