Study On The Key Techniques Of Trajectory Planning And Integrated Navigation Of UAV

UAV (Unmanned Aerial Vehicle) has been paid much attention because of its importance in reconnaissance and battle in recent years, and the related techniques of autonomous UAV are getting more and more attractive. In this dissertation, some key techniques about trajectory planning and integrated navigation are studied in detail. The main contributions are as follows:1. A trajectory-planning algorithm for UAV is put forward, where a shortest path is attained based on Voronoi-search, and then the path is smoothed. By use of the Voronoi-map, the search may be simplified from infinite to the finite space, thus the computation burden may be reduced greatly. The method may be used to the real-time trajectory planning, although the result is not optimal.2. Aiming at the flight environment composed of different power hostile radars, a global trajectory planning method is put forward using virtual potential fields and GA (genetic algorithms). Firstly, the radar net is carved up through Delaunay triangle, and the binary codes are designed according to the traversing direction and sequence. Then threats are taken as repulsion sources and the initial path as mass-spring- damper. The local optimal path of UAV is obtained by solving the mass-spring- damper equations. Finally, the global optimal trajectory is gotten by using GA. The method is proper to the off-line trajectory planning of UAV.3. A decentralized-integrated structure of DR/GPS/RP navigation is presented. Taking a certain UAV as example, filters are designed for each subsystem based on the respective characteristic of the subsystem, and then the filtering results are fused. Finally, the DR is corrected by using the fused information. The simulation results confirm the effectiveness of this method.4. A simplified model of the UAV is built. The yaw and the air-speed are filtered by FKF to improve the precision of DR. Effectiveness of this algorithm is validated by computer simulation.5. A least squared estimation (LSE) filter algorithm is presented and used to...