In order to improve the efficiency of material transportation scheduling and reduce the operation time, the problem of 2 batches of material delivery was divided into 3 stages. The minimum total operation time of the 3 phases was used as an objective function, a mixed integer programming model was proposed. Single and twoway lane constraints and the constraints of vehicles meeting at intersections were defined, and the multiple population genetic algorithm was introduced to study the material delivery scheduling model and algorithm of multiple material transportation vehicles. Compared with the single population genetic algorithm and the historical data of a company, the results show that by using the algorithm, the accuracy and speed of the system are greatly improved on the basis of avoiding local search, the material delivery time is reduced, and the reliability is improved. The research results provide a reference for improving the economic benefits of enterprises, and also provide new ideas for the optimization and decisionmaking of multibatch material delivery operations