Automatic Identification Method of Relay Protection Equipment Based on Enhanced Transfer Learning
The current method for automatic identification of information anomalies has low sensitivity and F-value.Therefore,a reinforcement transfer learning based automatic identification method for information anomalies in relay protection equipment is proposed.Perceive the information of relay protection equipment under different working conditions and preprocess it.Based on expert experience,use the logical process of"abnormal phenomenon abnormal evolution path abnormal cause abnormal solution"to establish an information anomaly recognition knowledge base.Through reinforcement transfer learning,learn the knowledge in knowledge base,extract abnormal features of equipment information,identify equipment information anomalies,and achieve automatic recognition of relay protection equipment information anomalies based on reinforcement transfer learning.Experimental results show that the sensitivity of the method proposed in this article can reach over 96%,and the F-value can reach over 97%.It can achieve accurate automatic identification of abnormal information in relay protection equipment.
enhanced transfer learningrelay protection equipmentinformation abnormalityautomatic identificationexpert experienceknowledge base
NETL
NSTL
万方数据
