Construction Of Time-Delay GM (1,N) Power Model Considering Interaction Effects And Its Application In Carbon Emission Forecasting

The grey prediction theory is one of the important methods to solve uncertainty problems,and the grey multivariate prediction model,as one of its branches,has attracted much attention and research since it was proposed by Professor Deng Julong.However,as the uncertainty system becomes more and more complex,the existing grey multivariate prediction model has the problem of simple model structure,such as not considering the nonlinear relationship between the system and related influencing factors,time-delay effect and interaction effect.Therefore,the grey multivariate prediction model needs to be continuously improved and perfected to meet the practical needs.This paper focuses on the optimization methods and applications of multivariate grey prediction models,mainly including:For the prediction problem of multivariate time-delay nonlinear systems,the time-delay driving terms,interaction driving terms and power index are introduced into the conventional model in this paper.A timedelay GM(1,N)power model considering the interaction effect,referred to as IEDGPM(1,N),is constructed,and specific model construction methods and rigorous mathematical proofs are given.Firstly,to address the time delay phenomenon that exists between the effect of system behavior characteristics and related factors in reality,this paper introduces the time delay parameter into the system driver term,and uses the grey expanded dimensional identification method to identify and analyze the model driver term and the sequence of related factors.Secondly,considering the interaction effects of multiple relevant factors on the system behavior and the possible existence of nonlinear relationships,this paper constructs a nonlinear interaction term incorporated into the GM(1,N)power model.Finally,based on the superiority of the swarm intelligence optimization algorithm,the Dragonfly algorithm is introduced to solve the nonlinear parameters.The constructed time-delay GM(1,N)power model considering interaction effects is applied to the prediction of carbon emissions in China.First,this paper studies and analyzes the current situation of carbon emissions in China.Based on the overall characteristics of China’s carbon emissions,it is determined that the grey prediction model is suitable for predicting the trend of China’s carbon emissions,and in addition,based on the analysis of the factors influencing China’s carbon emissions,the main influencing factors are selected to be included in the construction of the prediction model.Secondly,five multivariate grey forecasting models are constructed to simulate and test China’s carbon emission data,compare the accuracy of the models,and select the model with the best performance to forecast China’s carbon emission during the 14th Five-Year Plan period.Finally,based on the analysis of the current situation and the forecasting results,policy recommendations are proposed for relevant departments to control carbon emission.The research results obtained after modeling and application are as follows:(1)In this paper,the IEDGPM(1,N)model is compared with the traditional GM(1,N)model,the time-delay GM(1,N)model,the IEGM(1,N)model considering linear interaction effects,and the GM(1,N)power model.The results show that the improved GM(1,N)power model performs better in both simulation and testing stages in the application of carbon emission problems in China.(2)Through the analysis of the current situation of carbon emissions in China,it can be concluded that China is facing a serious challenge of emission reduction,and the application of the grey multivariate prediction model based on data characteristics to predict the total carbon dioxide emissions in China is of great significance for the implementation of effective carbon reduction policies.(3)Using the grey correlation model to identify the main influencing factors,the results show that total energy consumption and industrial structure are the strong correlates of total carbon emissions in China,and both have a 3-period lag.(4)The IEDGPM(1,N)model was used to predict the trend of carbon emissions in China during the period 2022-2025.The results show that China’s carbon emissions will continue to grow during the 14th Five-Year Plan period.If no effective measures are taken,it will not be possible to achieve the binding target of an 18%reduction in carbon emissions intensity.Based on the forecast results,this paper puts forward corresponding policy recommendations.