Planning a highefficiency and lowcost threedimensional (3D) flight track has become an urgent problem to be solved for UAV extensive application. Aiming at the problems of track length and lack of smoothness of ant colony algorithm in the flight track planning, this paper proposes the ant colony particle swarm fusion algorithm, which improves the node movement rules in ant colony system, constructs multiple heuristic information and combines the global search ability of particle swarm optimization algorithm. Meanwhile, to solve the problems of dynamic obstacle avoidance and target point change in the flight track, an improved bioinspired neural dynamics model algorithm is proposed, which realizes local online flight track adjustment for the obstacles and target point change in the 3D static optimal flight track. Experiment simulation results show that the ant colony particle swarm fusion algorithm can plan an expected track in 3D static environment. At the same time, the improved bioinspired neural dynamics model algorithm can not only dynamically avoid sudden obstacles, but also track the changes of dynamic target points in real time.