Long-term wind speed prediction based on optimized support vector regression

Wind energy is considered as one of the most remarkably renewable energy origins that reduce the expenditure of electricity production. In the last decade, there are several forecasting speed of wind algorithms that have been to improve prediction reliability. Support Vector Regression (SVR) parameters such as kernel parameter, penalty factor (C) have a great effect on the complexity and reliability of forecasting algorithm. This paper proposed a hybrid approach based on Whale Optimization Algorithm (WOA) and SVR namely WOA-SVR for fixing issues which traditional methods cannot handle effectively and have shown high performance in many respects. The performance of proposed algorithm (WOA-SVR) is evaluated using several different aspects as well as the daily average wind speed data from Space Weather Monitoring Center (SWMC) in Egypt as a case study is used. For verification, the results of the proposed algorithm are compared with Particle Swarm Optimization (PSO) and the original SVR without parameters optimization. The experimental results showed that the proposed WOA-SVR algorithm is capable of finding the optimal values of SVR parameters, avoid local optima problem, and it is competitive for forecasting speed of the wind.