Development Of On-line Partial Discharge Monitoring System For Ring Main Unit

In China,the use of ring main units is increasing year by year.To ensure the safe and reliable operation of the ring main units is the necessary condition to ensure the safe operation of the power system.At present,the cost of PD on-line monitoring system of ring network cabinet is too high,which makes it difficult to widely use the monitoring system.In addition,the on-line monitoring system of the ring main unit is not comprehensive enough,and the function integration degree is single.In order to solve the above problems,this paper designs a set of on-line partial discharge monitoring system of ring network cabinet.In addition to the consideration of temperature,humidity,smoke and other environmental parameters acquisition,the system also designed HFCT and TEV sensor for partial discharge detection of ring main unit,and verified the effectiveness of the sensor.The frequency reduction and conditioning circuit is designed to realize the frequency reduction of HFCT PD signal.After the frequency reduction,the HFCT PD signal can meet the AD port sampling rate of DSP.The DSP can sample the PD signal directly,which effectively reduces the cost of online monitoring system.The signal conditioning circuit of TEV partial discharge is designed.The conditioning circuit reconstructs TEV partial discharge signal into square wave.After conditioning,TEV partial discharge signal can be counted by ECAP module of DSP.Through the processing of HFCT PD signal and TEV PD signal,the acquisition of PD signal by DSP processor is realized.In the aspect of PD signal recognition,three kinds of PD signals with defects are collected in this paper,and the feature extraction effects of three feature extraction schemes of wavelet energy spectrum method,statistical feature parameter method and fractal feature parameter method are studied.Combined with BP neural network,statistical feature parameters and fractal feature parameters are identified.Experiments show that the recognition effect of statistical feature is better than that of fractal feature.BP neural network is used to verify that the mixed feature vector composed of statistical feature and fractal feature has the best recognition effect.In this paper,a set of on-line monitoring system software of ring main unit is developed.Based on the development of Lab VIEW software of NI Company,the remote on-line monitoring of the state of ring main unit is realized by using wireless communication scheme.The software comprehensively displays the status information of the ring network cabinet uploaded by the lower computer through the chart,which has the functions of real-time display of status data,monitoring information storage,historical data query,discharge type identification,etc.,providing a reference for the monitoring personnel to analyze the fault development trend.