Conditional Scenario Generation Method of New Energy Prediction Error Based on GMM

This paper proposed a novel conditional new energy scenario generation method, characterized by error-based conditional framework, and GMM is applied to fit the prediction error distribution. Aiming at the inherent and random uncertainties in short-term new energy forecasting, the proposed method can effectively reflect the changing trend and fluctuation characteristics of photovoltaic and wind power. The framework of error-based conditional scenario generation improves the adaptability and computational efficiency of the algorithm. Case studies using real datasets demonstrated superior performance in coverage and bandwidth metrics compared to existing methods. The proposed method provides a powerful tool for power system operators to assess operational risks, supporting grid planning and decision-making under renewable energy uncertainty.