Multi-Source Signal Identification Method of Cable Based on Time-Frequency Energy Center Moment Feature Separation

To solve the difficult problem of multi-source signal recognition in cable equipment online monitoring and fault diagnosis. This paper proposes a cable multi-source signal identification method based on time-frequency energy center moment feature separation. The method firstly collect the time domain pulse in the cable multi-source high-frequency current time-domain signal by triggering, then calculates the timefrequency energy central moment of the time-domain pulse, then carries out the feature separation based on the time-frequency energy central moment and divides the pulse group, and then reduces it to the PRPS and PRPD maps, and finally carries out the signal source identification. After experiments, it is proved that this method can maximize the feature differences of multi-source signals and realize the accurate classification of multi-source signals in cables.