| In this thesis, the attitude estimation algorithm for a RTLinux platform based miniature UAV(unmanned aerial vehicle) is designed and implemented.Miniature UAV generally refer to those which are equipped with embedded computer, GPS, IMU(inertial measurement unit), cameras and wireless comunication quipments, etc.. They could accomplish autonomous flight within limits, object recognition and location, so it is also called aerial robot.It is necessary to know its position and attitude relative to earth frame in order to locate a helicopter and control the dynamics of it. To estimate the realtime Euler angles, we make use of a compass three accelerometers and three gyros. because of the volume and weight limits of the helicopter, all sensors are mounted in a strapdown way in contrast to using a gimballed platform. Thereby, there are two problems which need more considerations. One is the attitude estimation, which is how to convert the data relative to the body frame received from sensors to the attitude relative to the earth frame. It is the basic of precise navigation and accurate control of the helicopter. The other problem is to fuse the data sets from different sensors.To solve the problems above, the thesis is arranged as follows. At first, we introduce comprehensive applications of autonomous miniature UAV. Then overseas and domestic studies and applications of attitude estimation algorithms are described. Furthermore, requirements of the miniature UAV designed in our lab is analysed and appropriate attitude estimation algorithm is presented.The design and implementation of the onboard embedded software platform for miniature UAV are also introduced in details. after the analysis of each subsystem of general Linux, How to obtain an embedded Linux satisfying our system requirements and transplant it onto PC/104 is explained. At last result of the attitude estimation algorithm running on the PC/104 platform is given.
|
PDF Full Text Request