Research On Yingkou Port Logistics Demand Forecasting Based On Combined Forecasting Model

With the rapid development of logistics, as necessary maritime transport channel, large coastal ports have been upgraded into an international logistics center. In the process of trade, except provide transshipment services for goods, it also provide a series of value-added logistics services, such as transportation, warehousing, customs, distribution, information processing. Port plays an important role of the modern and international development of logistics industry. In order to make full use of the function of the port logistics hub, attract more foreign investment and improve the level of economic, Countries, as well as major port city begin to draw up the scientific development strategy of port logistics, logistics demand forecast as a basis and prerequisite for the development of port logistics strategy, point out the direction for the construction of port logistics, to avoid repetitive construction of the port logistics infrastructure, greatly reducing the vacancy rate of port resources, and promote the development of economic and raise the level of logistics in the port city and hinterland area.Currently, due to the results of theoretical studies is less in the port logistics, in the rapid development of ports, lack of the systematic development planning and strategic, lead to the blind development and construction of the basic implementation of port logistics, resulting in the imbalance of the port logistics supply capacity and demand levels, scientific and rational planning of port logistics is very necessary.The paper begins the study from the influencing factors of port logistics demand, the establishment of the index system and demand forecasting method. Firstly, conduct a comprehensive analysis for the macroeconomic situation, hinterland regional economic and port environmental conditions of port logistics, Secondly, built the combined index system of port logistics demand forecasting, and give the index set of Yingkou port logistics demand forecasting. Finally, As an example of Yingkou Port, with the software of SPSS17.0and MATLAB2011, to predict the demands of Yingkou port by trend extrapolation, GM(1,1) model, multiple regression, BP neural network theory and combined model, and through compare the error to prove that combined model has better prediction accuracy, and predict the Yingkou port logistics needs in2012-2015with the combined model, the results show the huge potential for development of Yingkou port logistics, and propose the recommendations of the development of Yingkou port logistics based on the forecast results.