Mobile agent provides a novel paradigm for distributed computing. It has the potential to offer a single, general framework in which a wide range of distributed systems can be implemented efficiently, easily and robustly. The traveling agent problem is a complex combinatorial optimization problem, which solves the problem of planning out an optimal migration path according to the tasks and other restrictions when agents migrate to several hosts. Ant colony algorithm is a new evolutionary algorithm and extremely suit to solve the travelling agent problem, which has the characteristic of parallelism, positive feedback and heuristic search. To avoid the limitation of ant colony algorithm such as stagnation like other evolutionary algorithm, an improved ant colony algorithm is introduced to solve the travelling agent problem by modifying pheromone updating strategy, and a self-adaptive pheromone evaporation rate is proposed, which can accelerate the convergence rate and improve the ability of searching an optimum solution, so mobile agents can accomplish the migration task with high efficiency and short time. The results of contrastive experiments show that the algorithm is superior to other related methods both on the quality of solution and on the convergence rate.