Ultra‐short‐term PV power prediction model based on HP‐OVMD and enhanced emotional neural network

Accurate photovoltaic (PV) power prediction plays an increasingly crucial role to maintain the safety and reliability of grid operation. However, fluctuation non-stationarity PV make it a challenging task optimize accurate results. This paper presents novel model which is combination Hodrick–Prescott (HP) filter, optimized variational mode decomposition (OVMD), enhanced emotional neural network (EENN). It overcomes adverse effects random changes under highly volatile weather conditions. First, trend component are screened through HP filter as pre-step alleviate non-linearity impact data. Then, OVMD used decompose residual time series into relatively stationary intrinsic modes. Finally, EENN by grey wolf optimization (GWO) established predict each subseries, results subseries reconstructed obtain final predicted The numerical based on actual data show that accuracy proposed significantly improved compared with contrast models, achieves best against OVMD-GWO-EENN, VMD-GWO-EENN, GWO-EENN models.