Renewable Energy Generation Forecasting Model for Reducing Power System Load and Solving Output Control Issues

Global population growth and technological advancements have led to a surge in energy consumption, resulting in substantial greenhouse gas emissions and severe climate change. In response, various countries have implemented climate policies to combat climate change, which have accelerated the development of renewable energy. While renewable energy is more environmentally friendly than conventional fossil fuel-based energy, it still faces numerous challenges due to the lack of technological advancement. This study aims to alleviate output control issues and reduce the burden on the power grid by forecasting renewable energy generation using meteorological data, generation data, and power demand data. Unlike conventional fossil fuel-based power facilities, which can adjust output in real-time, renewable energy facilities must forecast power generation to supply electricity effectively. Unchecked expansion of renewable energy facilities without considering power generation could lead to severe issues, including large-scale blackouts, beyond simple output control problems. This research employs the Long Short-Term Memory (LSTM) model, demonstrating through collected and analyzed data that solar and wind power generation can be effectively forecasted. By utilizing the developed renewable energy generation forecasting model and power demand data, this study can aid in forecasting curtailment and facilitate solutions such as optimal placement of Energy Storage Systems (ESS) and expansion of power grid infrastructure to mitigate grid load and address generation curtailment issues. Ultimately, it contributes not only to the effective achievement of Korea's CFI 2030, aimed at carbon neutrality, but also to the expansion of renewable energy facility deployment.