Clustering-Based Short-Term Wind Speed Interval Prediction With Multi-Objective Ensemble Learning

As a renewable and green energy source, wind energy has attracted great attention from academia and industry in recent decades. However, it is challenging to integrate wind energy into smart grids due to the instability and randomness of wind speed. To solve this problem, this paper proposes a clustering-based short-term wind speed interval prediction with multi-objective ensemble learning, which can provide an accurate and reliable wind speed interval prediction to support energy dispatch planning. First, a clustering-based uncertainties estimation method segments the initial wind sequence into several groups and determines the estimated width for each group. Second, a variational mode decomposition is employed to acquire the sub-sequence matrix of wind speed, and then a Hurst exponent-based model selection method is used to choose and train an optimal model for each sub-sequence based on its long-term correlation. Finally, an improved multi-objective optimizer is utilized to determine the optimal superposition weights of the prediction results for each model. The proposed approach is evaluated using eight cases from two wind farms, which are published by the National Renewable Energy Laboratory. Experimental results indicate that the proposed approach outperforms several state-of-the-art studies, demonstrating a higher prediction interval coverage probability and a narrower prediction interval width.