In order to solve the problem that the inherent strong uncertainty in PV power data leads to the low prediction accuracy of a single prediction model,a short-term PV power prediction method based on quadratic modal decomposition and hybrid bi-directional gated recurrent unit(HBiGRU)model is proposed.Firstly,to cope with the uncertainty of PV power data,a series of smoother eigenfunction components are obtained by processing PV power data based on adaptive noise-complete ensemble empirical modal decomposition,sample entropy and variational modal decomposition.Secondly,the HBiGRU model is constructed to fully explore the characteristic relationship between each component and PV power influencing factors,and the prediction results of each component are obtained.Finally,the prediction results of each component are superimposed to obtain the short-term PV power prediction results.The simulation results show that the proposed integrated prediction model can be adopted to effectively improve the short-term PV power prediction accuracy.Compared with the results from other prediction models,the normalized mean absolute error and root mean square error are reduced by 3.21%and 5.04%,respectively,and the coefficient of determination is improved by 22.7%.
万方数据