ore Accurate Wind Power Prediction Based on Intelligent Error Correction Model

EasyChair Preprint 9684

Abstract

In order to limit the negative impacts of wind energy fluctuations on power system performance, this research suggests an improved hybrid method to improve the accuracy and efficiency of wind energy forecasts. This strategy mainly consists of a two-phase modeling technique. First, a main model based on the wind power curve is developed, whose function is to anticipate the evolution of wind power using physical mechanisms. The errors of the initial model are then extracted and become the study objectives of the second phase. Using the modeling capabilities of data mining techniques, data-driven models for error correction are developed. The ultimate results of wind energy projection are a combination of these two steps. The analysis of a real wind farm demonstrates that the proposed method outperforms conventional models in terms of accuracy and cost analysis. Using a specific degree of improvement, the quantitative results reveal significant improvement over the baseline physical model and conventional statistical models.

Keyphrases: Data Mining, Hybrid, efficiency, errors, wind farm

@booklet{EasyChair:9684,
  author    = {Bellat Abdelouahad and Mansouri Abderrahman and Tyass Ilham and Mansouri Khalifa and Raihani Abdelhadi},
  title     = {ore Accurate Wind Power Prediction Based on Intelligent Error Correction Model},
  howpublished = {EasyChair Preprint 9684},
  year      = {EasyChair, 2023}}