In the complex dynamic environment, the UAV can be autonomous intelligent path optimization through the perception of the surrounding environment to reduce the track error and arrive to the destination safety with a minimum cost, this is called intelligent navigation. The intelligent navigation need solve the key techniques, such as self-positioning, environment awareness, navigation decision and route guidance strategy, so as to ensure that have the real-time and reliable navigation information, including position, speed, attitude and the safety path. Firstly, the agent is brought into the navigation system and the intelligent navigation decision method based on agent is established. Then the intelligent navigation system information, including composition, function, etc. is given. Based on the above study, a kind of path planning method based on artificial potential field is presented, which can make the intelligent path planning in any route point for UAV.