Research On Mission-based Adaptive Indoor Navigation Methods Of Quadrotor

As UAV(Unmanned Aerial Vehicle) is widely used in military and civilian frontiers, more and more people have devoted themselves to its research, especially the indoor autonomous navigation, which is becoming a hot topic. But now the indoor autonomous navigation is mostly researched based on the extern global navigation equipment and it's difficult to navigate based on airborne equipment. Also the indoor search mission of IARC VI(International Aerial Robotics Competition VI) challenges the indoor autonomous navigation with airborne equipment. To complete the mission mentioned above, an adaptive navigation method of the whole mission is studied.Firstly, according to the structured characteristic of environment where the competition will be held and also that the competition required that the doorplate and the U disk must be identified, airborne computer, laser range finder and camera are chosen to make up the quadrotor platform. Secondly, a three-layer system architecture of mission layer, navigation layer and control layer is mentioned according to the traditional automatic control system. The mission layer analyses and planes the whole mission and some sub-mission states are got; The navigation layer further divides the sub-mission states into some navigation states in order to be controlled easily; The control layer accomplishes every navigation states. Thirdly, an autonomous navigation method based on graded finite state rough set is mentioned. The finite state machine is extended to the graded finite state machine and the complex problem is simplified, and then rough set theory is used to get the knowledge reduction set. Finally, an adaptive navigation method based on the three-layer system architecture and graded finite state rough set theory is mentioned. The system formation of mission layer, navigation layer, and control layer is introduced respectively: the mission layer studied the basis of sub-missions, and sub-missions based on graded finite state machine is obtained; the navigation layer consists of hardware of navigation system, data process of sensors such as laser and camera, and the realization of adaptive navigation method; the control layer studied the hardware of flight control system and the attitude control method.The software program is written according to the method and experiments are conducted and during the International Aerial Robotics Competition held in August, 2013, the results shows that the method is appropriate to complete the competition mission.