Research On Navigation Method Of Quadrotor For Multi-target Operation

With the development of sensor technology and embedded computer technology,the uses of multi-rotor aircraft also expand further, it has a wide application in military and civilian aspects. The attitude control, localization and navigation and task execution of multi-rotor aircraft have become a research hotspots recently. And achieving autonomous flight rely on airborne navigation equipment is still a technical difficulties. The mission 7 of the International Aerial Robotics Competition(IARC)proposed higher requirements of autonomous capability of aircraft. Taking the competition as the research background, this paper gives a research on localization and navigation method of multi-rotor aircraftFirstly, according to the rules of mission 7 and unstructured indoor environment,designed a visual localization algorithm by using optical flow odometer. This algorithm uses the ground horizontal and vertical intersecting lines as the artificial landmark and use Kalman filter fuse with optical flow data to achieve global positioning.Secondly, according to the task of quadrotor UAV maneuvering ground targets,gives a path planning research on moving target intercept method. For multi-target decision-making, proposed a method of APF model, which is effective in this mission.Finally, based on the task layer, navigation layer and execution layer system structure, designed the entire control system and did experiments on the global navigation.According to the research, completed the program and participated in the IARC(Asia-Pacific Division) which held in Yantai in August 2014, the results show that the system designed in this paper has certain applicability.