The emission of greenhouse gases poses enormous pressure on current carbon emissions and carbon reduction. Accurate quantification of carbon emissions from coal-fired power plants is of great significance for achieving the dual carbon goal. To enable enterprises to better understand their carbon emissions, this study constructs a carbon emission model and carbon emission data accounting model for coal-fired power plants. Case data calculations and a carbon emission reduction analysis were conducted. The experiment showcases that the carbon sensitivity of the inner side of the boiler under control conditions is higher than that of the operating parameters controlled on the inner side of the steam turbine, with a maximum total value of 16.67 g/MJ; the annual average low calorific value of coal remains between 16,000 kJ/kg; the activity level of coal remains between 30,000 TJ; and the oxidation probability of coal char during combustion fluctuates, with a maximum of 99.8%. In the calculation of coal-fired carbon emissions, the fitting difference between the emissions of generator unit 1 and generator unit 2 is maintained within 2%. Overall, the CO₂ emissions of power plants involved in the study are generally high. The model built through this study has well analyzed the carbon emissions of power plants. It is of great significance for the actual carbon emission reduction of coal-fired power plants.

E-commerce platforms involve interactions among platform governance, manufacturer pricing, consumer perception, and managerial cognition. This study examines how platform overconfidence and consumer product-fit perception accuracy affect pricing, profitability, and operational mode selection. Product-fit perception accuracy denotes the probability that consumers correctly perceive the mismatch between their preferences and product attributes. We develop a two-period dynamic game involving two competing platforms and two manufacturers and compare resale and agency modes under simultaneous and sequential decisions. Equilibrium analysis shows that, under the agency mode, overconfidence increases promotional allowances and manufacturers’ first-period and switching-purchase prices, while its effect on repeat-purchase prices depends on a mismatch-cost threshold. Under the resale mode, overconfidence raises wholesale prices only within an intermediate range, and product-fit perception accuracy can have either a positive or negative effect depending on the level of overconfidence. Sequential decision-making further produces different pricing responses for leader and follower platforms due to differences in commitment and information conditions. Numerical analysis under the specified parameter settings indicates that resale tends to generate higher platform profits when consumers’ perceived product-quality level is high. When this level is low, the relative profitability of the two modes becomes more sensitive to overconfidence, product-fit perception accuracy, and decision position, with agency becoming more profitable in some parameter regions. These results characterize how governance mode and decision timing shape the economic consequences of managerial overconfidence and product-fit information within the modeled platform setting.