Research And Application Of The Freight Logistics Forecasting System

Quickly expanding logistics brings important influence on all kinds ofindustry in global economy .As an important part in integrated transportationchain, airport changes into freight logistics rapidly. There are many influent factorson freight logostics, including country or area polity, economic level of airport hinterland, geography condition and weather status. These factors have uncertainty and variability by time, and bring large difficulty to forecast. So, launching the forecasting study on freight logistics system has very good realistic meanings, aiming at fact of domestic airport.The Artificial Neural Network (abbreviated as ANN) has very strongrobustness, fault-tolerance and learning ability. Because ANN is good at associating, summarizing, analogy and reasoning, it is used in various fields including forecast. Relations of factors in freight logistics system are nonlinear. Classed forecast method is difficult using nonlinear system; while ANN's excellent non-linear character is applicable to the system. For this reason, we can use intelligence method base on ANN into freight logistics system.This paper discuss the characteristic of the freight logistics system, look for the way to improve the precision of the Forecast, work over ANN and Gray-system for forecast, then , analyse their excellence and disadvantage so that offer the date for forecast. The main studying contents and results of this paper are as follows:(1)It studies thoroughly the logistics forecast, especially character of frieght logistics system and requirement of prediction; then provides the function and structure of freight logistics system and method of software realizing.(2) It analyzes roundly the solution process and characters of classed forecast and ANN model, founds model of ANN forecast method and comes true relevant arithmetic with MATLAR tnnlhnx(3) It studies the methods of the programming interface between VB and MATLAB, and puts in practice.(4) It discusses data warehouse and on-line analytical processing technology and building progress, analyzes the method and course of data obtainment in logistics forecast system based on data warehouse, which is used for decision-making.