Research on wind power prediction based on IGOABP neural network

Aiming at the local minimum problem of Grasshopper optimization algorithm (GOA), an improved IGOA algorithm was obtained by chaotic initial solution, adding nonlinear adaptive scale factor, Gaussian variation domain perturbation and firework mapping boundary variation in the position updating process. The experimental findings indicate that the IGOA algorithm outperforms its predecessor, the GOA algorithm, by enhancing both its global search capability and local optimization precision. Furthermore, it effectively mitigates the issue of the GOA algorithm getting trapped in local minima, thereby achieving a higher level of optimization accuracy. The IGOA algorithm is integrated with the BP algorithm to establish the IGOA-BP algorithm model and is applied in wind power prediction. Show the experimental results that in comparison to the GOA-BP, PSO-BP, and GA-BP algorithms, when the BP neural network is employed, the proposed IGOA algorithm can resolve effectively the concerns of output oscillation and low accuracy in wind power prediction, and the model performs more productively