Research On Regional Logistics Demand Forecasting Based On Set Pair Analysis

In such a comprehensive environment that the rapid development of big data," Internet plus" and e-commerce,in the modern society;the increasing update of the domestic infrastructure;the spiritual civilization of the people's living standards has been improved,it provides good conditions for the development of logistics industry.Whether to the country,the region,the enterprise or the individual has formed a certain scale to the demand of logistics.The understanding of logistics scale helps to understand the development of economy and the formulation of policies,which has certain guiding significance for future logistics planning.As the consequence,the prediction of logistics demand is very important.At the view of regional logistics,it is meaningful to predict regional logistics demand for logistics industry development,resource allocation for time and place,traditional industrial optimization and whole planning.Based on the current situation of regional logistics demand forecasting in China and abroad,considering the influence factors of logistics demand diversity as well as the uncertainty,nonlinear,complexity characteristics of logistics demand environment,using set pair analysis clustering prediction algorithm which satisfied with these characteristics to study logistics demand.The following contents are mainly studied:(1)This paper introduced some knowledge and characteristics about logistics demand.After studying the research on logistics demand,instead of the freight volume of previous study selected,this paper has selected the total social logistics as the forecast target,which can cover the various phases of the logistics and fully reflect the demand of the logistics.(2)This paper described the theoretical knowledge of the set pair analysis method.For regional logistics environment of uncertainty,unlike traditional prediction method of regional logistics,the theory of set pair analysis method with the uncertainty system theory and the identical-discrepancy-contrary system theory can study the logistics demand forecasting.(3)This paper analyzes the relevant factors that influence regional logistics demand,and establishes the index system of logistics demand predicting study based on these elements.Compared with the previous study of the set pair analysis regional logistics demand predict,this paper determined the weights of each index by correlation analysis and set pair potential analysis.It is the innovation point of this paper to use the set pair potential analysis to determine the weight coefficient.(4)In the case of Jilin province,an empirical study was carried out,which adopt the single weight coefficient and the comprehensive weight coefficient of the three groups,and the empirical results indicated that compared with the single wight coefficient the accuracy of the comprehensive weight coefficient prediction was good.