Research on Intelligent Verification of Relay Protection Setting Based on Improved Machine Learning
[Purposes]Due to the constantly changing operation mode and control strategy of the power system,relay protection setting methods need to adapt to these changes to ensure correct operation in various situations.However,in the process of adjusting adaptation,it is easily affected by external fac-tors,resulting in low accuracy of fixed value verification.Therefore,the research on intelligent verifica-tion of relay protection setting based on improved machine learning is proposed.[Methods]Firstly,a re-lay protection fault dataset is constructed for training and optimizing machine learning models.Then,an improved machine learning method is used to establish an intelligent sorting model for relay protection settings,which automatically performs intelligent sorting on the protection settings.Finally,by intelli-gently verifying the relay protection settings,fast response and accurate protection against power system faults can be achieved.[Findings]The experimental results show that the proposed verification method is not affected by external environmental interference,and the accuracy of relay protection setting verifi-cation is high.[Conclusions]The use of the designed method ensures stable operation of the power sys-tem while improving calibration accuracy.
improved machine learningrelay protection settingintelligent verificationpower system
NETL
NSTL
万方数据
