Main content area

A multi-objective wind speed and wind power prediction interval forecasting using variational modes decomposition based Multi-kernel robust ridge regression

Naik, Jyotirmayee, Dash, Pradipta Kishore, Dhar, Snehamoy
Renewable energy 2019 v.136 pp. 701-731
algorithms, hydrologic cycle, models, prediction, time series analysis, wind farms, wind power, wind speed, Spain
This paper presents a new hybrid multi-objective wind speed and wind power prediction interval forecasting (PIs) model which is the combination of variational mode decomposition (VMD), Multi-kernel robust ridge regression (MKRR) and a multi-objective Chaotic water cycle algorithm (MOCWCA). VMD is applied to decompose the main time series signals into appropriate number of modes that avoids the mutual effects present in between the modes. The VMD based MKRR method is applied to estimate the wind speed and wind power prediction intervals at a prediction interval nominal confidence levels (PINC) of 95%, 90%,85% and 80%, respectively. Further to improve the performance of the proposed prediction model MOCWCA is introduced for the optimization of the prediction models parameters in such a way that multiple objectives are satisfied to produce Pareto-optimal solutions. The wind speed and power data samples for prediction interval forecasting are collected at 30 min and 1 hour time intervals from the Sotavento wind farm located in Spain.