Forecasting clean energy generation volume in China with a novel fractional Time-Delay polynomial discrete grey model
Ye Li,Xue Bai,Bin Liu
DOI: https://doi.org/10.1016/j.enbuild.2022.112305
IF: 7.201
2022-09-15
Energy and Buildings
Abstract:Clean energy, represented by hydropower, wind power, and nuclear power, plays a crucial role in addressing the issue of power shortage and promoting the transition of the energy structure. Scientific and accurate forecasting of clean energy generation volume can provide the necessary reference for relevant authorities to formulate reasonable energy policies. This paper proposes a novel fractional time-delay polynomial discrete grey model, FTDP-DGM(1,1), for forecasting clean energy generation volume in China. Firstly, both the structural parameters and the time response function are derived directly from the discrete equation. Secondly, the genetic algorithm is employed to determine the optimal value of the nonlinear parameter in the model, with the optimization objective of minimizing the mean absolute percentage error. Then, the proposed model is applied to the simulation and prediction of hydro, wind, and nuclear energy generation volume in China. The results show the validity and superiority of the model. Finally, out-of-sample forecasts are performed, and the forecasting results show that the proposed model has good performance with values of 1.29%, 2.94%, and 2.32% for mean absolute percentage error(MAPE), respectively. It is expected that in 2021–2025, China's hydro, wind, and nuclear power generation volume will continue to grow at a rate of 1.45%, 10.74%, and 10.66% respectively. Based on the forecast results, policy recommendations are suggested.
energy & fuels,construction & building technology,engineering, civil