A novel Seasonal Fractional Incomplete Gamma Grey Bernoulli Model and its application in forecasting hydroelectric generation

Xin Xiong,Zhenghao Zhu,Junhao Tian,Huan Guo,Xi Hu
DOI: https://doi.org/10.1016/j.energy.2024.130257
IF: 9
2024-01-11
Energy
Abstract:With the arrival of the first truly global energy crisis, how to precisely forecast the hydroelectric generation becomes a hot spot for allowing governments to obtain more valuable information. A novel forecasting model, Seasonal Fractional Incomplete Gamma Nonlinear Grey Bernoulli Model (SFIGNGBM(1, 1)), is proposed in this paper to precisely forecast the hydroelectric generation in some countries. First, the seasonal raw data are classified into four seasonal groups based on their significant seasonal fluctuations. Second, a novel SFIGNGBM(1, 1) model is established by combining the Bernoulli equation, the fractional-order accumulation operator, and the incomplete gamma function to further optimize partial parameters in the forecasting model and improve the forecasting performance. Third, the Whale Optimized Algorithm (WOA) is employed to optimize the Bernoulli power exponent η , the fractional order parameter r , and the incomplete coefficient h for minimizing the MAPE values and enhancing the fitting precision. Finally, our results present that our proposed model outperforms a set of baseline forecasting models with the smallest three error measure values in all fitting results, and its MAPE values converge before 10 iterations. This indicates that our proposed model has a favorable forecasting performance with fast-convergence for hydroelectric generation in the elected countries.
energy & fuels,thermodynamics
What problem does this paper attempt to address?