The Inner Overheat Fault Recognition Method of Distribution Power Transformers Based on the Supplerting Vector Machines
In order to reduce the problem of mor error inner overheat fault recognition of the distribution power transforme,in the paper,the inner overheat fault recognition method of distribution power transformers based on the supporting vector machine is proposed.The method analyzes the inner overbeat fault for the transformer and based on the analysis result,conduct charcecteristic quantity selection and normalized treatment to eliminate existential difference of gas volume value,ensuring follow-up recognition result accuracy to lay the foundation.Then,by cross-ing subrroup exchange tractics and optimal population self-variance tractics,improve leapfrog algorithm.Finally,based on improved mixed leapfrog algorithm to achieve supporting vector machine improvement,finishing the inner overheat fault recognition of the power transformer.The test result shows the adopted method conducts the inner o-verheat fault recognition of the distribution power transformer,its recognition accuracy is always up to 95%,and has better applicability.
power transformerinner overheatimproved supporting vector machinefault recognition