Short-term prediction of photovoltaic power based on improved Wile Horse Optimizer
Guomin Xie, Chenxi Wang · 2023
According to the volatility of photovoltaic output, a multi- kernel extreme learning machine (MKELM) based on improved Wile Horse Optimizer is proposed. First, historically similar day data sets were generated using fuzzy C mean clustering (FCM). Secondly, in view of the problem that MKELM parameters are difficult to determine, the improved Wile Horse Optimizer (IWHO) is used to optimize the kernel parameters and regularization coefficient of MKELM, which enhances the generalization ability of the model. Taking the measured data as an example, the results show that the proposed prediction model can achieve satisfactory prediction accuracy under different weather conditions, compared with the traditional support vector machine, BP neural network and long and short-term neural network error index, which effectively improves the prediction accuracy of photovoltaic output.