Transformer Fault Diagnosis Based On Neural Network

The power industry is the forerunner in the development of various fields and the cornerstone for the sustained and stable development of the national economy.The demand for power systems in the contemporary world is getting higher and higher,which also makes the power system gradually develop toward large capacity,ultra high pressure and cross-region.As the core equipment of the power system,the normal operation of the transformer is an important factor to ensure the safe,reliable and high-quality operation of the power system.Early fault diagnosis can be used to detect early faults inside the transformer and evaluate its operating status,so as to provide more decision-making reference for load dispatching and to formulate a state maintenance plan.This study also combines the analysis of dissolved gases in oil with neural network technology.By using the advantages of self-learning of artificial neural networks,it can realize the function of monitoring the state of transformers online,and it will also detect in the early stage of faults.Defects under development to achieve more timely,accurate and intelligent detection of transformer faults.At the same time,in order to improve the efficiency of fault diagnosis and the speed of diagnosis,this paper proposes a method combining neural network and rough set,which reduces the large fault samples by rough set theory and reduces the scale obtained by rough set reduction.The reduction decision table is used as a training sample of the neural network,and the process is established into a transformer fault diagnosis model to diagnose the fault of the transformer.In view of the low accuracy of current transformer fault diagnosis methods,this thesis uses neural network-based fault diagnosis and fault diagnosis based on rough set and neural network respectively,and compares the results,using the same fault sample data to establish nerves.The diagnostic simulation of the network gives the diagnostic accuracy and diagnostic step size of the two methods,and analyzes the simulation process and results of the two methods.The simulation results prove that the latter ismore efficient,more accurate,and has higher Practical value.