A wind speed point-interval fuzzy forecasting system based on data decomposition and multiobjective optimizer

Despite the depletion of traditional energy resources and escalating environmental challenges, the importance of wind power in the energy sector has intensified. However, the inherent stochasticity of wind makes short-term forecasting a complex and essential task for grid stability and efficiency. This study proposes a wind speed fuzzy prediction system that integrates enhanced variational mode decomposition for data preprocessing, an optimal predictor selection strategy for selecting optimal submodels, and a modified multiobjective optimization algorithm for optimizing multiple forecasting objectives. The system employs fuzzy theory to construct fuzzification, aggregation, and defuzzification functions, leveraging the strengths of benchmark predictors to generate point and interval predictions. In addition, comparative experiments are conducted on three sampling intervals of data from five wind turbine sites in China. The proposed system achieved a mean absolute percentage error of 3.89% and a prediction interval coverage probability of 94.44% at site 1, which significantly outperformed the existing contrast models.