Machine Learning Approach for Wind Power Forecasting and Fault Detection for FACTS Device Integration to Provide Grid Support

Introduction of wind energy into the current grid system is being complicated by numerous challenges as the energy is intermittent. The machine learning (ML) techniques and power electronics will be used together to help in the reliable and efficient operation of wind powered grid system in this paper. Authentic SCADA-based information has been employed by us to come up with a supervised regression model that estimates active power production of wind generation. We have also designed a fault detection system that can identify the fault in power generation in case the wind condition is fluctuating. We have demonstrated the predictive capacity and to some extent trend identification that will assist in an earlier fault detection to make smarter operational decisions. We have also indicated, further, where we now can incorporate Flexible AC Transmission System (FACTS) equipment like STATCOM, SVC and UPFC to reduce voltage collapse and generally enhance grid performance in the ML–FACTS system. ML–FACTS system offers a robust system to scalable solution to run wind energy systems and also to stabilize power characteristics to the grid.