General Machine Learning based approach to pulse classification for separation of Partial Discharges and Interference

This paper describes a complete approach to filtering Partial Discharge (PD) pulses from interference in High Voltage electrical equipment using supervised Machine Learning (ML) techniques. The PD signals are registered Ultra Frequency radiation band with multisensor acquisition system composed of 4 antennae. proposed methodology focuses on the implementation ML algorithms and proposes novel this field onset detection incoming signals. goal was achieve high accuracy reasonably low compilation times classifier. That would allow use model edge sensor devices. In paper, different models training variants framework tested. presented results based robust measurement campaign performed laboratories GEIRI Europe. is validated through tests 3 separate test scenarios. Each represents complexity problem an increasing number active sources. show potential for utilization ANN other classifiers problems as achieves desired threshold 80% most tested variants. step forward toward fully online filter.