To constitute the scientific, effective logistics development strategy, construct the efficient operation of the logistics system to meet the economic development, ensure the supply of logistics for the demand to keep the logistics high efficiency and benefit, the regional logistics demand should be predicted to avoid being economic development bottleneck caused by the lack of supply logistics, or to prevent the phenomenon of over-investment and the repeated logistics infrastructure construction caused be the excess capacity of logistics supply.At the beginning, the related theories of logistics demand were introduced to analyze the necessity of the provincial regional logistics demand and the related impacts of logistics demand factors. Then taking principles of logistics demand quantitative index into account, the reason and the rationality of selecting the cargo volume and freight turnover quantity as logistics demand forecast quantitative index were described according to the status of lacking of historical statistic data. At last, the combining forecast method of Gray model and BP neural network model is employed to forecast the logistics demand.The Gray model has the characteristic of fewer samples, convenient calculation and can take full advantage of finding out the development rule by using the known information. And the BP neural network model, which can simulate any non-linear relationship of input and output, has the characteristic of simple construction, better maneuverability, and better approaching and generalization ability of non-linear as well as better adaptation. The combining forecast method can offset the weakness of approaching complicated non-linear function and enhances both strengths. It gives prominence to variation trends and weight relations between input and output. The new gray neural network model was used for provincial regional logistics demand because it directly shows the variation trends of influencing factors of logistics demand and the variation trends between the cargo volume and freight turnover quantity and weight relations between them. According to the variation trends and weight relations, the development direction was provided to adjust the logistics strategy and programming. Finally, Hunan province was taken as an example to empirical analysis. |