A comprehensive wind speed forecast correction strategy with an artificial intelligence algorithm

Wind speed forecasting is critical to renewable energy generation, agriculture, and disaster prevention. Due the uncertainty intermittence of wind, conventional methods with numerical weather prediction (NWP) models fall short achieving satisfactorily high accuracy. Post-processing predicted results necessary for enhancing The industry generally employs time-series (TSP) error correction, yet it time-consuming since repeated modeling needed if location changes. Aiming at addressing this problem, paper discusses application a deep learning algorithm in post-processing period wind prediction. NWP are utilized as basis, algorithms used minimizing errors. An experimental study conducted industrial data. functionality performance TSP-based including rolling mean, exponential smoothing, autoregressive integrated moving average compared learning-based algorithms, long-short term memory convolutional neural network. From results, both TSP deep-learning error-correction can effectively increase accuracy day-level model while data-driven, no process needed. This work also poses an insight into future development meteorology.