A coupled VECM-SD modeling approach for power demand forecasting in the new era

Under the background of new development concept in the new era and the goal of "double carbon", the power demand system presents the complex characteristics of multi-factor, nonlinear and strong feedback. Traditional forecasting methods struggle to comprehensively capture its dynamic feedback features and evolutionary driving mechanisms.Consequently, this study develops a novel VECM-SD (Vector Error Correction Model-System Dynamics) coupled forecasting model.The system is deconstructed into four key subsystems: socio-economic, electrification, green power structure, and policy response. This framework facilitates the analysis of its dynamic feedback mechanisms.The VECM component provides robust econometric calibration of variable relationships, while the SD component simulates complex policy impacts and feedback effects. This integrated approach enables sophisticated multi-scenario forecasting and policy simulation. The model offers valuable insights for formulating robust energy strategies and supporting the low-carbon transformation of the coal-powered electricity system, with particular relevance to China's context.