నైరూప్య

Prediction research on food cold-chain logistics demand based on grey and AW-BP

Bi Ya, Chu Ye-Ping, Tao Jun-Cheng

The food cold-chain logistics is a typical nonlinear complex system, for which the traditional prediction method has been stretched thin. In view of the characteristics of Chinese economic development and the availability of the statistical data, the paper took the food cold-chain logistics system of Hubei province as the research object and designed the prediction method based on grey and AW-BP. The prediction method combined the advantages of low requirement of grey prediction method for statistical data and strong nonlinear capacity of BP neural network and overcome the disadvantages of too slow convergence and being easily caught in local optimum of general BP neural network by methods of correction of error function, introduction of dynamic adaptive weight etc. Through a large number of experiments, it is proved that the new prediction method is greatly improved in rate and precision of convergence and the capability of getting rid of local extremum and is a practical and efficient prediction method; simultaneously, the prediction data about future demands of the food cold-chain logistics of Hubei province were obtained and the analysis on the development trend and the scale of the food coldchain logistics system of Hubei province was made according to the prediction data.