Abstract: Due to computation complexity, lower accuracy and premature convergence of the conventional algorithm, the ant colony algorithm was improved to solve the routing problem of logistics. In order to eliminate the problems that ant colony algorithm is easy to be stagnant and its convergence is slow, the algorithm was improved including the following points such as the transfer strategies of the ants, the pheromone update method and the integration of genetic algorithm. To solve the path optimization of simultaneous delivery and pickup, the heuristic function was added and the transfer strategies were designed, which made the algorithm conduct a search by considering a combination of factors reasonably, and could find the next right point more comprehensively and more accurately, then, a better path could be gotten.