RF Algorithm and MSCR-integrated Internal Fault Diagnosis and Relay Protection for Transformers
In order to solve the problems of high error rate and low accuracy in internal fault diagnosis and relay protection process of transformers,the study first uses random forest algorithm to mine the internal fault data of transformers,and then utilizes the characteristics of magnetic saturation controllable reactors to construct a simulative relaying model based on MSCR by analyzing the internal fault status and operating status of transformers.The results showed that the accuracy of the three fault classifiers in the training and testing sets differed by 1.03％,3.04％,and 4.68％,respectively.This indi-cates that the constructed relaying model can effectively identify and troubleshoot internal faults in transformers,thereby providing technical support for transformer protection.
random forest algorithmmagnetically-saturated controllable reactortransformerrelayingtroubleshooting
NETL
NSTL
万方数据
