Fault content and composition of the gas type transformer fault linkages between

This paper analyzes the characteristics of transformer fault based on the DGA valid information provided to study the composition of the gas content and the fault type of fault between phase transformer mutual contact. As the traditional  electric meter  fault diagnosis  Voltage stabilizer  detection accuracy is not high, requires a lot of experience and so insufficient, the urgent need to introduce some new  power transformers  fault diagnosis techniques to compensate for the current transformer fault diagnosis in a variety of defects. Support vector machine is in statistical theory developed on the basis of a new learning method, which not only can solve the neural network algorithm in high-dimensional problems and local minima problems, but also for small samples with good learning generalization ability.