The dynamic evolution mechanism of carbon emissions based on time-lag difference APSO-CKPLS

To address the problem that the quantitative calculation between the correlation system framework data and carbon emissions is affected by the small sample, multivariate, dynamic and time lag of carbon emissions, the time lag differential adaptive particle swarm-combined kernel partial least squares (APSO-CKPLS) method is proposed. The method first analyzes the time lag phenomenon of carbon emission in the actual production process, estimates the time lag parameters by using the sliding gray correlation method, reconstructs the model sample, and then constructs a time-lag difference model based on time difference to solve the problem of dynamic update and time lag of the model by reconstructing the sample. Finally, the APSO-CKPLS method is used to obtain the quantitative calculation of carbon emissions from the data collected in the framework of the correlation system, and to construct a dynamic evolution mechanism of carbon emissions for the whole process based on the integrated energy consumption at the edge of the key emission control enterprises in the framework of the correlation system of carbon emission sources. This method is compatible with the research on the dynamic evolution of carbon emissions of various carbon emission enterprises.