Application And Improvement Of Grey Prediction Algorithm In Logistic Parks' Demand Forecasting

Demand forecasting is the weakness of domestic logistic park's planning. In view of this, the thesis chooses demand forecasting as the object of study. It redesigns the process of parks'planning, sums up the general ideas of demand forecasting and evaluates common modes along with algorithms inland and abroad as well. The paper starts with the practical application of the grey prediction algorithm in parks'demand forecasting. According to the characteristic of the Markov chain, it builds the transition probability matrix and defines the happening probability vector of relative errors to refine the result, which makes the section of the grey prediction more reasonable than the pure grey prediction. Furthermore it introduces an approach on the base of the characteristic of the Jordan matrix to simplify the calculation, and still proves that the approach can reduce the time for calculation as well errors effectively, which can raise the precision of prediction greatly. In addition, it combines the algorithm with the practical research on the demand forecasting of logistic park at the Lianyungang port to evaluate the algorithm as well as the improved calculation, and expects to provide some reference for other logistic park planning.