It is important to find the shortest path for manoeuvres of UAV, since the power consumed during manoeuvres is tightly coupled with the length of the flight path. In this paper, an algorithm that can find the shortest path during manoeuvres and improve the performance of UAV to follow waypoints is described. The shortest path for UAV during manoeuvres is derived firstly by the theory of Dubins curve. Secondly, in order to improve the ability of UAV to follow the derived optimal path, a real-time path planning algorithm is designed by transforming the constraints of Dubins curve into a dynamic equation. To demonstrate the applicability and performance of the proposed path planning algorithm, two numerical examples are presented. The results show that the proposed algorithm is promising to be applied in the path planning for manoeuvres of UAV.