A Double-stage Hierarchical Hybrid PSO-ANFIS Model for Short-term Wind Power Forecasting

Assessment of the output power of wind generators is always associated with some uncertainties due to wind speed and other weather parameters alteration, and precise short-term forecasts are essential for their efficient operation. This can efficiently support transmission and distribution system operators and schedulers to improve the power network control and management. In this paper, we propose a double stage hierarchical particle swarm optimization based adaptive neuro-fuzzy inference system (double-stage hybrid PSO-ANFIS) for short-term wind power prediction of a microgrid wind farm in Beijing, China. The approach has two hierarchical stages. The first PSO-ANFIS stage employs numerical weather prediction (NWP) meteorological parameters to forecast wind speed at the wind farm exact site and turbine hub height. The second stage models the actual wind speed and power relationships. Then, the predicted next day's wind speed by the first stage is applied to the second stage to forecast next day's wind power. The influence of input data dependency on prediction accuracy has been analyzed by dividing the input data into five subsets. The proposed approach has attained significant prediction accuracy improvements. The performance of the proposed model is compared with five other prediction approaches and showed the best accuracy improvement of all.