Prediction Based On The Combination Of Projected Regional Logistics Needs

Along with quickening step's of economic globalization, logistics industry becomes more and more important in optimizing the allocation of resources, and is contributing more and more to economy. Logistics was considered as the development artery and the basic industry of National Economy. The degree of logistics development has already become an important symbol to evaluate economic development degree of a nation or a district. Many departments of government in many countries and districts have carried on the planning work of logistics distribution system to improve the investment environment and enhance the comprehensive competitive power of a region or an urban. Forecast research on region logistics demand is an important link of logistics plans. Therefore, establishing scientific and reasonable forecast model and predicting demand of logistics quantitatively can provide policy-making foundation for formulating the region logistics development policies, determining scales of logistics infrastructural facilities construction, and analyzing logistics market situation.Firstly, this paper has done a detailed research on the elementary theory of region logistics demand, introduced the relation between regional economy and logistics industry, analyzing the economic impact factors of logistics demand and the predictability, content and so on of region logistics demand, these have established theory basis for the following logistics demand forecast modeling.Secondly, this paper has studied region logistics demand forecast target system. Through analyzing various target of logistics demand, region logistics demand forecast target and related economic target has been established.Thirdly, this paper has conducted the research on technology and the method of the region logistics forecasting. Through comparing and analyzing, this paper respectively selected the grey theory, the multiple regression theory, and the neural network theory to found the forecasting model, and combined each single model to establish combination forecast model by appropriate method.Finally, according to the former founded region logistics theoretical model, this paper has forecasted the logistics demand quantity of Yunnan Province.This characteristic of this paper has two points:First, it has made principal component regression model to solve multicollinearity among independent variables. Second, three forecast techniques have been combined appropriately, and then confirmed the combination forecast model are much closer to the actual situation, and improved the forecast precision.