The Proposal of Wind Speed Scenario Generation using a Forecasting Error based on Bivariate Copula Function; [이변량 코퓰라 함수 기반의 예측 오차 모델링을 활용한 풍속 시나리오 생성 방법의 제안]

Abstract

With the target of global Net-Zero, the future power system needs to consider planning and operation approaches based on the stochastic properties of renewable energy. In this paper, we present the methods of wind speed scenario generation using conditional forecasting error. The proposed method utilized LSTM model to forecast the wind speed. In addition, it employed a copula function to cope with a dependency of wind speed and its wind forecasting value and estimated the conditional distribution of forecasting error. The effectiveness of the presented method was verified using a real wind speed data. The simulation results demonstrated that the various scenario generation can be obtained from the estimated conditional forecasting error, which is expected to be used for detailed power system analysis. Copyright © The Korean Institute of Electrical Engineers.