Machine Learning Studies for PD Monitoring

Partial discharges (PDs) are crucial to CMD (Condition Monitoring and Diagnosis) of high voltage and high field electrical insulation of power apparatuses and power electronics. Recent development of machine learning (ML) methods shall enable to accelerate such PD studies. There are lots of ML algorithms available to select. ML classification is more useful to investigate PD monitoring rather than ML regression. A recent review is given for this contemporary topic in the paper.