Novel nonlinear wind power prediction based on improved iterative algorithm

To effectively improve the accuracy of wind power prediction and reduce the load on the power grid, a new nonlinear wind power prediction model based on an improved iterative learning algorithm was investigated. Firstly, the actual wind conditions are equated to a non-linear model. Using the concept of nonlinear decomposition, the nonlinear model is divided into many linear subdomain models while taking into account the nonlinear impacts of temperature, wind direction, altitude, and speed on the model. Next, using CRITIC weight analysis, the ideal weights are determined. Then, the linear sub-domain model is fitted into the full non-linear wind power prediction model equation by utilizing the least squares method. And the objective function of the iterative algorithm for iterative optimization search is derived from the prediction equations that were previously developed. The final enhanced iterative technique for nonlinear wind power prediction is produced by merging the iterative algorithm with the nonlinear decomposition. Finally, a comparative study of wind power prediction under different prediction models was carried out. The research results showed that the average absolute error of wind power prediction and the root mean square error were 4.5841% and 0.2301%, respectively. In particular, the prediction accuracy improved by 8.28%.