Impact of Renewable Energy Penetration on PMU Based Grid Event Detection Using Machine Learning Framework

Abstract

This paper is a systematic investigation and comprehensive study on ML based grid event classification in presence of renewable energy resources. In this data-driven approach, PMU data with reporting rate of 24 samples/cycle have directly been processed by various ML algorithms. Finally, it has clearly demonstrated the success of 1-D CNN for multichannel signal processing by wiping out the challenges of inter class similarities due to RE integration. The efficacy of the algorithm has been established with IEEE −9 bus system. Preliminary results obtained so far are very encouraging, projecting the promising future 1-D CNN in this area due to its excellent feature extraction using convolution process.