Abstract: Ant colony algorithm (ACA) is a new heuristic algorithm, which is successfully used to solve some NP-hard combinatorial optimization problems through simulating the process of ants searching for food. In this contribution a new ACA, which is based on adaptively adjusting the pheromone on routes according to the solutions that artificial ants have found, will be proposed. Thus it can escape from the local maximum. Simulations demonstrate that the improved algorithm can achieve better performance than basic ant colony algorithm.