Short-term wind power prediction optimized by multi-objective dragonfly algorithm based on variational mode decomposition

Short-term wind power prediction has a considerable effect on improving the productivity of wind energy systems and increasing economic benefits. In recently years, various wind velocity predictive models have been designed to raise the prediction effect. However, numerous predictive systems are limited by single type, and many ordinary predictive systems ignore the advantage of optimized parameters and the significance of data preparation, which bring about the lower predictive precision. To fill this gap, in this article, a novel predictive system is come up, which is on the basis of data denoising strategy, statistical predictive systems, artificial intelligence forecasting system and multi-objective optimization strategy. After using the data denoising strategy for denoising, the reconstructed data is used for the forecasting of different sub-systems, to obtain stable forecasting results, multi-objective dragonfly algorithm is used to estimate the weight coefficient of sub-systems. To evaluate the availability of the designed predictive system, five wind velocity datasets from different wind farms are used for the purpose of a case research. According four experiments and four analyses, it can be concluded that the designed combined system has a well predictive effect in short-term wind speed prediction. And it is in favor of grid regulation and operation.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

In this article, a novel predictive system based on a multi-objective optimization strategy is proposed to improve the accuracy and stability of short-term wind power prediction. By utilizing techniques such as data denoising, statistical predictive systems, and artificial intelligence forecasting, the designed system shows promising results for grid regulation and operation.