Abstract:In view of the continuous growth of China's electricity generation, based on China's electricity generation data from 2012 to 2021, the short-term electricity generation of China from 2022 to 2030 is predicted by using the gray-arima coupling model, which is derived from the metabolic GM(1,1) model and the damped trend model.?By comparing the experimental results of metabolic GM(1,1), traditional GM(1,1) and new information GM(1,1) models.The SSE minimum model is coupled with ARIMA model, and time series analysis is introduced to fit the power generation.An equal-weight gray-arima coupling prediction model is constructed.?The reliable coupling prediction data of power generation from 2022 to 2030 is obtained, which has certain robustness.?It has important reference significance for China to formulate new energy measures and development plans to achieve carbon neutrality, carbon peak and sustainable development.