Forecasting China’s Carbon Dioxide Emissions Using a Novel Structure-Adaptive Conformable Fractional Gray Bernoulli Model

Accurate forecasting of CO2 emissions can provide theoretical support for the Chinese government in formulating carbon reduction policies. However, China’s CO2 emission sequences are constrained by limited available samples and complex characteristics. To address these issues, this study develops a novel structure-adaptive conformable fractional gray Bernoulli model (SAFGBM (1, 1)). Specifically, the newly designed model incorporates a new fractional-order accumulation operator to extract the sequence features and further uses the Simpson formula to optimize the background value. Subsequently, the particle swarm optimization (PSO) algorithm is employed to solve the nonlinear parameters in the model. Comparative verification with various competing models demonstrates that the proposed model exhibits significant accuracy advantages in forecasting China’s carbon emission sequences. Finally, this model is applied to predict China’s CO2 emissions from 2025 to 2030. The results indicate that although the growth rate of China’s CO2 emissions has slowed down, the dual carbon goals still face substantial challenges; the government needs to accelerate the implementation of relevant policies to advance the realization of the carbon peaking target and high-quality green economic development.