| With the increasing complexity of the modeling objects,the contradiction between the structural adaptability of the grey forecasting model and the complexity and variability of the modeling sequence is gradually deepening.Therefore,based on the existing grey forecasting modeling technology,this paper systematically studies the novel grey forecasting modeling technology from the aspects of data preprocessing optimization method,model expansion and model parameter optimization.A novel adaptive multi-variable grey forecasting model with adjustable parameters and variable structure was studied emphatically.The main works was summarized as follows.Firstly,the defects in structure and parameters of the OGM（1,N）model with relatively complete structure and high model accuracy were analyzed.Then,based on the above shortcomings,two kinds of grey optimization methods were discussed.The first one was to introduce real number field generation operator,which is beneficial to explore the function of obtaining sequence difference information of the order in the novel model.The second was to propose a smooth generation method of independent variables,which can weaken the influence of extreme values in the sequence on the modeling results of multi-variable grey forecasting model.Secondly,based on the above two proposed grey optimization methods,a novel multi-variable grey forecasting model with variable structure and parameter combination optimization,SPGM（1,N）for short,was constructed.Specifically,the modeling mechanism of SPGM（1,N）was described,the parameters of SPGM（1,N）were estimated and proved,the time-response formula and the final restored expression of SPGM（1,N）were derived,the modeling steps of SPGM（1,N）were sorted out,and the compatibility of SPGM（1,N）was proved.In addition,the performance and stability of SPGM（1,N）were tested by three groups of cases with different data characteristics.The results showed that the performance and stability of SPGM（1,N）are better than the mainstream multi-variable grey forecasting models of the same kind,which proved that the improvement of the novel SPGM（1,N）model is effective.At the same time,SPGM（1,N）is of positive significance to enrich the modeling method system,improve the modeling ability and broaden the application range of multi-variable grey forecasting model.Thirdly,based on the SPGM（1,N）model,the prediction of domestic waste clearing and transporting volume in Jiangsu Province was studied.On the basis of explaining the selection principles and steps of experimental objects and variables,SPGM（1,N）was firstly applied to simulate and analyze the domestic waste clearing and transporting volume in Jiangsu Province.The results showed that the performance and stability of SPGM（1,N）were better than other similar grey forecasting models.Secondly,the SPGM（1,N）model was applied to predict the domestic waste clearing and transporting volume in Jiangsu Province from 2021 to 2030,and the results showed that the average annual growth rate is 3.12%.The prediction results provide data reference for the countermeasures and suggestions of government management. |
