The toxic and heavy gas leakage accidents have the characteristics of suddenness and seriousness of the damage. An optimal scheduling model with multiple demand points, multiple supply points and multiple material types with the goal of minimizing emergency total time and total emergency cost was established. In order to enhance the practicability of the model, vehicle load and optimal path selection are added. The improved Dijkstra algorithm is used as a sub-algorithm of MOPSO to solve the model. Aiming at the weakness of MOPSO algorithm that is easy to fall into local optimal solution, the update mode of inertia weight is designed to enhance the global search ability of the algorithm. Meanwhile, the adaptive grid method is used for reference in particle selection operation, which enriches the diversity of particle swarm. Finally, simulation experiment are performed to verifies the effectiveness of the proposed model and algorithm.