The sigmoid iteration ACO is optimized for the problems of poor environmental adaptability, high number of inflection points and high computational complexity that exist in the traditional ACO in route planning. Firstly, the Sigmoid activation function distribution strategy is adopted to improve the initial pheromone spread through the position of the mesh nodes, and the initial concentration of the pheromone is assigned by the sigmoid, which reduces the blindness of the algorithm's pre-search; secondly, the adaptive factor is introduced to dynamically regulate the heuristic function, which increases the degree of expectation of the ants in choosing the globally optimal node, and reduces the convergence time of the algorithm; lastly, a statistical analysis is carried out in each generation of the ant, and the three characteristic parameters of ant path optimal, worst and average are extracted in each generation, and the pheromone updating function is dynamically adjusted according to the number of iterations to give full play to the parallelism characteristics of the algorithm. The results prove that the improved algorithm shortens the optimal path length by 2.7%, 3.2%, and 5.4%, reduces the average number of iterations by 42%, 53%, and 62%, and shortens the worst path length by 49%, 62%, and 73%, respectively, when compared with the ant colony system, the elite ranking algorithm, and the traditional ACO. The study prove that the optimized algorithm has stronger global optimality seeking ability and better application value.