Improved Ant Colony Algorithm for Vehicle Scheduling Problems of Military Logistics Distribution

This paper is aimed to research into military vehicle scheduling problem (VSP) by using ant colony algorithm. A vehicle scheduling model with time windows was built up based on the objective of minimum transportation distance, and the model characteristics and application prospects was analyzed. Based on local search strategies, traditional ant colony algorithm was improved. Then the algorithmic procedures of the model was put forward, and the parameters of state transition function in ant colony algorithm were calibrated by test calculation. An example was given to demonstrate feasibility and actual application method of the model and algorithm program. The study indicates that the improved ant colony algorithm has higher algorithm efficiency and can effectively solve vehicle scheduling problems of military distribution centers; The parameter selection of ant colony algorithm significantly influences the algorithm convergence.