This paper deals with the ordering and transportation of raw materials. Firstly, the order completion rate, out of stock rate and stability were taken as three indicators affecting the supply characteristics, and a comprehensive evaluation model was established to select the companies with high scores as suppliers. BP neural network is used to establish the order quantity prediction model, and the neural network toolbox provided by MATLAB is used to program. Then take the different demand as the objective function, establish the multi-objective programming model, and give different priority factors, convert it into a single objective programming, so as to find the optimal ordering scheme and transport scheme.