Embedded Autonomous Flight Control System Design For Multi-Rotor UAVs

This thesis concentrated on the study of multi-rotor UAVs. According to the technical requirements of multi-rotor UAVs, an embedded autonomous flight control hardware system was designed for the purpose of autonomous flight of series of multi-rotor UAVs. The main works done in this thesis is as follows.(1) According to the requirements of highly real-time capability and multitasking ability, the hardware circuit of the control system was designed. A new type of dual-chip processor architecture was designed and the sensor circuit, power circuit and isolation circuit were given to build up the hardware control system.(2) uC/OS II(Micro Control Operation System Two) was a ROMable, scalable, portable, real-time multitasking kernel, which was suitable for flight control system especially. In order to improve the maintainability and real-time performance of the system, uC/OS Ⅱ was adopted in this thesis as the embedded operating system for system management. This thesis presented detailed requirements and steps of transplantation of uC/OS Ⅱ and proved that this system could be transplanted in the proposed hardware control system successfully.(3) The applications of navigation algorithm of UAVs were studied, which involved the ARHS (Attitude Heading Reference System) design and the INS (Inertial Navigation System) design based on Extended Kalman Filter. ARHS mainly provided the information of attitude angle. To improve the accuracy of attitude angle calculation, the dynamic acceleration was added into the observation equation as the noise and the offset error of gyro was added into the status equation to jointly calculate the attitude angle information. INS mainly calculated the information of velocity and position. Considering the inaccuracy of the GPS signal, a calculating algorithm was proposed based on the combined compensation using the accelerometers and barometer. Then the designed algorithm was embedded into the control system. The outputs of the designed algorithm were compared with those of the commercial sensor MT1-G, and the result proved the feasibility of the proposed algorithm.(4) The attitude, velocity and position model were established for quad-rotor UAV and the incomplete differential PID controller was designed. Experimental results verified the effectiveness of the system.By above-mentioned work, the dual-chip control system designed in this thesis has been mounted on various multi-rotor UAVs to control the fight. It provided a hardware-scalable, maintainable and affordable control platform for the study of UAVs and made it easy for secondary development.