perfecting a rich decision-making background, modeling with actual simulation data, and solving (mainly linear programming models and SPREADSHEET solutions). It is sufficient to create a large linear programming model. If necessary, individual small models can be added, but on the premise of better problem solving, the content cannot be split to destroy the overall unity. criteria: degree of innovation in specific topic selection, rationality of model background and amount of information, richness of model description and setting (type and number of parameters and variables of model design), rationality of parameter setting (whether it is in line with industry reality), Sufficiency of literature and data citation, accuracy of model solution