Study On Greynet Forecast Algorithmic For Tianjin Port Logistics Demand And Development

With the sustained rapid growth of China’s foreign trade volume in recent years,the large port hub all around the world were playing important roles on promoting theintegration of the world economy, as well as on the optimization of resourcesconfiguration. And under the background that global production and manufacturingindustry is gradually transferred to Asia, especially China, coastal ports of China willassume an important cargo evacuation mission, so to speed up the development of theport of modern logistics industry is the objective requirements. In order to seize theopportunity, we must do research on establishment of port logistics demand.Considering the lack and nonlinear variation of the logistics data, this papercreatively get the gray theory and neural network algorithm combined to overcomeprediction difficult. Firstly, the article describes the basic theory and methods oflogistics demand forecasting, including the logistics demand forecasting indexselection, gray theory, neural network algorithm. Secondly, the paper gives focus ofanalysis on five factors which affect the demand of port logistics, namely theeconomic level, the industrial structure, the level of consumption and regional trade.After that analysis, author extracts the secondary index sets from the five aspects.Then, on the basis of the set of indicators and the feasibility analysis, authorconstructs a gray neural network combined model for port logistics demand forecast.There are nine indicators input as well as one port cargo output. Empirical studieshave shown that the model of the non-linear relationship between the input and outputbetter fit. Finally, this paper give five aspects of the development of the Tianjin Portlogistics development strategy.The research results show that the combination of gray prediction theory andnonlinear prediction function neural network algorithm can effectively find portlogistics demand affect the link between the factors and output indicators. Theempirical study effectively verify the reliability of the algorithm and feasibility studyof port logistics demand forecast and even regional logistics demand forecastinganother idea.