Performance Comparison of Metaheuristic Optimization-Based Parametric Methods in Wind Turbine Power Curve Modeling

Wind-based power generation, which is a safe and clean energy resource, has a widespread implementation around the world to reduce the environmental pollution, alleviate the energy crisis and provide the economic benefits. In the wind energy industry, robust and stable power curve modeling is an important task for the operational management of wind turbines. This work compares the power curve modeling performance of various metaheuristic optimization-based parametric methods, which have not been combined before in this field. African vultures optimization algorithm, Fick’s law algorithm, geometric mean optimizer and marine predators algorithm are used for the metaheuristic optimization, while 3- and 4-parameter logistic functions, 6th- and 7th-order polynomial functions and modified hyperbolic tangent function are utilized for the parametric approximation. As a result of the experimental analyses, marine predators algorithm-based modified hyperbolic tangent method provides the best goodness-of-fit results, while the mentioned metaheuristic algorithms-based 3-parameter logistic methods provide the worst ones.