Ant Colony Optimization (ACO) Algorithm, a probabilistic-based algorithm for seeking optimized paths in graphs, which is originated from the behavior of ants hunting for food. In nature, ants would emit a type of secretion called pheromone when they walk. Other ants may follow the paths where the concentration of pheromone is relatively high, and this would further increase the possibility of choosing the paths. However, traditional ACO algorithm possesses some defect that it might fall into the local optimum. Our strategy is to integrate artificial potential field for traditional ACO algorithm. Consequently, our improved version has prominent improvements with respect to its running time and how the path is chosen. Also, we proposed a new pheromone volatilization mechanism, which efficiently solve the problem of falling into the local optimal and would speed up the convergence of the algorithm.