A new pheromone updating rule is used to improve the traditional ant colony algorithm, which effectively solves the vehicle routing problem with hard time windows. A logistics distribution vehicle routing optimization model considering the impact of fire strikes from the enemy in battlefield environment is established, and the optimization routes are obtained by using this algorithm. Considering the possible changes of material demands of different combat units, a queuing strategy is used in the sovling process of the algorithm, and the subsequent distribution routes adapting to the demand changes are obtained. The rationality of the corresponding distribution routes is verified by simulation experiments.