Research on Combined Forecasting Model for Logistic Material Demand Based on BP Neural Network and Grey System Theory

Abstract

Logistic material possesses the characteristics of high consumption, tight time and multiple uncertain factors. Precise demand forecasting is an important prerequisite for implementing active and meticulous material support. The rule of material demand can be studied based on self-learning and self-adaptation ability of BP neural network and the accumulated principle of grey system theory is applied to weaken the randomness of system data. A combined forecasting model for logistic material demand based on BP neural network and grey system theory is designed and applied in real cases. The forecasting results indicate that this method has advantages like high forecasting precision and rapid convergence rate. This method is superior to simple BP neural network or grey system theory method, which can narrow forecasting error, making the forecasting results closer to reality.