Adaptive decision fusion based framework for short-term wind speed and turbulence intensity forecasting: case study for North West of Turkey

Abstract

In this paper, an online learning framework called adaptive decision fusion (ADF) is employed for short-term wind speed and turbulence intensity forecasting by use of wind speed data for each season for the city of İzmit, located in the northwest of Turkey. Fixed-weight (FW) linear combination is derived and used for comparison with ADF. Wind speeds and turbulence intensities are predicted from the existing wind speed data and computed turbulence intensities, respectively, using the ADF and FW methods. Simulations are carried out for each season and the results are tested on mean absolute percentage error criterion. It is shown that the proposed model captured the system dynamic behavior and made accurate predictions based on the seasonal wind speed characteristics of the site. The procedure described here can be used to estimate the local velocity and turbulence intensity in a wind power plant during a storm.