A novel seasonal grey Euler model with three parameter-interval grey numbers for forecasting power generation

Accurate forecasting of power generation helps to assess the stability of the supply of power generation and to better plan the use of electrical energy. China’s power generation exhibits seasonal oscillations, nonlinear growth, and uncertain fluctuations. The interval numbers can reflect the range of uncertainty fluctuations in the data. Therefore, the three-parameter interval grey numbers prediction of China’s power generation is studied. A matrixed Fourier grey Euler Bernoulli model MFGEBM(1,1) for three-parameter interval grey numbers is proposed. First, a seasonal factor is introduced into a new Caputo fractional accumulation generation operator to reduce the seasonal volatility of the sequence. Secondly, the Fourier series and Bernoulli’s equation are introduced into the grey Euler model to further improve the applicability to sequences with seasonal oscillations. Then, based on a new convergence factor and triangular walking strategy, the grey wolf algorithm is improved to optimize the model’s parameters, and its effectiveness is verified with algorithm comparison experiments. In order to test the accuracy of the model proposed in this paper, two cases with different development trends and related to power are studied, and four existing grey models for seasonal oscillation sequences are used as competing models. Finally, the proposed model is used to forecast China ’s power generation.