Partial Discharge Angle Of Arrival Estimation Based UHF Wireless Sensor Array

Long-term partial discharge(PD)can cause equipment insulation aging and lead to equipment failure.Therefore,monitoring of partial discharge is one of the effective methods to ensure the safe and reliable operation of power system.Commonly ultra-high-frequency(UHF)space partial discharge detection and localization techniques in substation include two types of time difference method and received signal strength indicator(RSSI)method.The time difference method is simple in principle,but requires extremely high-precision synchronous data sampling devices and leads to a high cost.The UHF detection method based on the received signal strength has the advantages of low cost and is convenient for carriage,but the workload of establishing a fingerprint database in the early stage is large,and it is not easy to deploy and implement.And it only supports online monitoring mode.Therefore,this paper proposes a partial discharge direction finding technology based on UHF wireless sensor array,which uses UHF RSSI and sensor array technology to locate partial discharges in the substation.In this paper,the architecture and composition of UHF wireless are first introduced.Through measurement,the basic signal transmission model of the sensor is found that the received signal strength value in a particular direction is relatively large.Then,a circular array of sensors based on the model is formed to obtain the RSSI curves of the sensor array for partial discharges in different directions in space.Based on the basic received signal strength curve,a partial discharge angle of arrival estimation method based on a circular array is proposed.The number of sensors in the array is increased from four to twelve.According to the difference in the RSSI received by each sensor in different directions,the angle of partial discharge is achieved.Also,neural network is used to filter the data.Normalization,interpolation and clustering are used to improve results.At the same time,considering the cost and the size limitation of the field application,the number of sensors remains four.Then,angle of arrival estimation method based on maximum likelihood estimation is proposed.The received signal model of the array in more directions is obtained offline.Then,the maximum likelihood estimation in statistics is used to perform a probability analysis on the data to achieve the results.The two partial discharge angle of arrival estimation methods are carried out in a high-voltage laboratory.And the orientation accuracy under the discharge distance,the discharge angle,and the discharge intensity were studied respectively.The experimental results show that the maximum angle error of the two methods is within 9°,which satisfies the positioning of equipment or equipment parts with potential defects.