The invention relates to the technical field of unmanned aerial vehicle patrol
route intelligent planning, in particular to a continuous
rigid frame bridge unmanned aerial vehicle patrol
route intelligent planning method, which comprises the following five steps of: acquiring data by using an oblique photography camera and an RTK (Real-Time Kinematic) positioning module to generate a bridge high-precision three-dimensional
space model; carrying out structural semantic segmentation on the model to identify a key area, and analyzing and arranging patrol waypoints in combination with view
reachability; an improved artificial
potential field method is adopted to plan a three-dimensional
route, a dynamic
gravitational field and an optimized repulsive force field are constructed, and after
resultant force is synthesized, a trajectory conforming to unmanned aerial
vehicle dynamics constraints is generated through a
smoothing algorithm; when a patrol task is executed, sensing a dynamic obstacle in real time, predicting a position and triggering an online re-planning adjustment track; and
processing the acquired image data, and updating the three-dimensional
space model of the bridge. According to the method, the patrol
data accuracy and the route safety are improved, the task continuity is enhanced, the model dynamic updating is realized, the patrol cost is reduced, and reliable support is provided for bridge
safety monitoring.