基于多模态信息融合的高压开关柜局部放电缺陷识别与定位OA

Aimed at the problem of imprecise detection and localization of partial discharges in switchgear when a single signal is used,this paper designs an integrated sensor that fuses three types of signals—Transient Earth Voltage(TEV),Ultra-High Frequency(UHF),and Acoustic Arrays(AA).Subsequently,through the partial discharge simulation test of the real-type switchgear,the paper uses the partial discharge signals detected by the three-in-one sensor to calculate the 5-dimensional time-domain signal characteristics respectively and construct a multimodal information fusion model to achieve the identification of different types of insulation faults.Furthermore,based on the designed sensor,the paper combines with the time difference of arrival(TDOA)method to make a study on the location of the partial discharge power supply for the designed sensor switchgear,and also proposes the sampling period conditions that need to be satisfied when the time difference method is used.