Prediction Of Cable Joint Temperature And Insulation Deterioration State Based On UHF-RFID And DRN

With the rapid development of social urban modernization,power cable has become an important part of the distribution network line and transmission power.As its weak point,cable joint can effectively reflect the hidden danger and health status of power cable operation,and temperature and insulation deterioration are important indicators to judge whether the normal operation of the cable.Therefore,real-time detection of cable joint temperature and insulation deterioration is a current research focus.At present,the conventional cable temperature monitoring means cost a lot of manpower and material resources,and the collected data is not used effectively,and the collected temperature data transmission is delayed.In view of this phenomenon,this thesis studies the prediction of power cable joint temperature and insulation deterioration.The main work is as follows:First of all,this thesis summarizes and analyzes the current main methods for detecting the temperature of power cable joint,proposes a power cable joint temperature detection system based on UHF-RFID(ultra-high frequency radio frequency identification technology),designs and selects its main modules,and sets up an experimental platform to verify its accuracy in detecting the temperature of power cable joint.Secondly,a power cable joint temperature prediction model based on DRN(deep residual network)is proposed in the background computer side of this system.The big data collected in the early stage of the system is used to train the prediction model and can be updated and improved in time.The DRN residual block is designed to make it more suitable for this study.Then the underground environment factors and the actual cable current load are fully considered to accurately predict the cable joint temperature.By using Python simulation and comparing the simulation results of DRN with those of CNN and RBF,the results show that the DRN prediction model is accurate and has practical engineering significance.Finally,taking polyvinyl chloride cable,which is commonly used in underground laying,as an example,the aging factors of insulation are summarized and analyzed,and the life prediction analysis of electric aging model and thermal aging model is carried out respectively.The results show that the limitations and accuracy of this model are not high.Based on this,a DRN model for predicting cable insulation deterioration is proposed in this thesis,which makes full use of the cable joint temperature data predicted by the system,and uses EEMD to decompose the temperature series to improve the prediction accuracy.The simulation results show that the DRN prediction model has high accuracy.