Wind power correction model designed by the quantitative assessment for the impacts of forecasted wind speed error

Abstract

The errors in wind power forecast will incur additional cost. It is critical to quantify the relationship between forecasting error in wind speed and power output. Unlike previous works that have rarely considered the speed error, we propose a comprehensive and repeatable wind power forecast correction model that quantitatively assess the impacts of speed error on power error, based on the power curves, speed predictions and distribution of speed forecast error. In this correction model, the power forecast error is obtained by calculating the mathematical expectation. The mathematical expectation of the wind power error is equal to the integral of the wind power error multiplied by its associated probability. Additionally, power forecast error and its probability are constructed as a function of speed forecast error and speed forecast error probability, respectively. To evaluate the model performance, numerical simulations are carried out in Guilin, Xiangyang and Xihai. The results suggest that the model can reduce the biases between observed and forecasted power, with the correlation coefficients increasing by over 15% in Guilin and Xihai. Furthermore, the root mean square error exhibits notable decline, with a reduction of over 35%, from 0.34 to 0.21 MW, from 0.42 to 0.27 MW and from 0.39 to 0.24 MW in the three aforementioned locations, respectively. This study contributes to enhancing the efficiency of wind power generation.