Study On The Forecasting Of Regional Logistics Demand Base On ANN And Its Application

With the economic globalization and informationlization accelerating continuously, the logistics as the new service industry that has broad prospects and value-added function has been developing rapidly throughout the world since the middle of 1980s.The countries and areas regard logistics as the significant industry to develop. They hope to boost the economy, improve the investment environment, attract increased foreign investment and lessen the employment pressure by developing logistics industry. However, the sudden growth of logistics may be lead to imbalance between the logistics supply and the actual logistics-demand. In order to provide the basic data and decision-making supports in drawing up logistics industry policy, planning and investment for the planners, the paper is going to analyze and forecast the scale of regional logistics demand, the structure of logistics demand and the developing trend of logistics demand.The paper mainly study the problem how to predict the regional logistics demand, which is based on the forecasting research and related knowledge. Firstly, the article introduces the purpose and contents of this study. Secondly the paper based on the basic theory of regional logistics demand, analyzes the main factors and the correlation between them and sets up the demand forecast index system, including regional economic index and logistics demanding index. Thirdly, the paper based on the basic theory of artificial neural network (ANN), analyzes the feasibility of using ANN to forecast logistics demand, establishes ANN model for regional logistics demanding forecast and describes the process of implement. Finally the paper analyzes and forecasts the logistics demand of Shanghai by using quantitative forecast method and qualitative forecast method.The important conclusion can be drawn from this research is that there is a strong correlation between regional logistics demand and regional economy level. The intrinsic logic relationship decides the regional logistics demand can be forecasted by using the economy level. Because BP neural network has a powerful non-liner mapping function, it can reveals the non-liner relationship between regional economy and logistics well and gets reasonable results. So the paper introduced a scientific tool and mean for regional logistics demand forecast.