Proposing an Innovative Method for Predicting Optimum Wind Generator Output

Abstract

Renewable energy sources offer environmental benefits and a more consistent supply, making them essential for integrating into energy fusions. Network operators have challenges with intermittent and stochastic power sources like wind power owing to their fluctuating production. This study presents an innovative method for predicting wind turbine power output, from any location in a vast region with high efficiency. To ensure a representative sample of geographical locations, the research region is divided into a grid and sampled accordingly. The ideal production times for each sub-area are discovered and utilized to anticipate electricity output. It uses the LSTM model to predict meteorological data and the linear model to approximate wind turbine power curves. The sub-area showed one-day wind speed, direction, and power projections clustering scores. Wind and directional predictions yielded RMSE (0.35m\sec, 7.9rad) and R2 (94%, 71%) scores.