Fault Analysis Of Execution Units And Flight Control For Hex-Rotor Unmanned Aerial Vehicle

Over the last decade, Unmanned Aerial Vehicle (UAV) has been drawing more ind more attention from engineers and scientists due to its potentials in both military ind civilian areas. This paper studies a Hex-Rotor UAV developed on the basis of olanar multi-rotor unmanned aircraft. In this structure, the six rotors tilts from its axis thus improve the weakness in the yaw control of the UAV. The weight of the rcraft’s payload, endurance and fault redundancy were also increased. This paper will focus on the mathematical model of Hex-Rotor UAV, failure analysis, fault letection and diagnosis systems, and control systems research related issues. The nain contents include the following aspects:The dynamics of Hex-Rotor UAV were studied. Dynamics analysis and dynamic nodel is the basis for the design of flight control system. The multi-rotor UAV structural configuration was studied thoroughly and several reasonable simplifications were made using relevant knowledge. The dynamics model of the UAV’s translation and rotation motion around its center of mass was built. Finally, rdware architecture of the Hex-Rotor UAV prototype (computing layer, communication layer and task layer) was established, and the basic methods to mprove the reliability of the prototype’s hardware and software was describes. TheThe lift fault model of the execution units was established. First, the model of brushless DC motor was studied and analyzed, and an Extended State Observer (ESO) was designed to estimate the load resistance torque of the motor. Then the common types of fault on drive circuit board and their hazards on the aircraft systems were explained. Analysis of the rotor anti-torque and twist lift model shows that the lift factor and anti-twist torque factor will fluctuate with external factors, and set them as constants will reduce the control effect. In addition, this paper also points out the effect of rotor’s rotation dynamic imbalance on lift model, and eliminates the lift fluctuations noise by a FIR filter. Finally, the execution unit fault type was analyzed (motor failure, the drive circuit board failure and rotor failure), and the lift fault model was established (gain type faults and failure conditions).Fault detection and diagnosis system of the execution unit was studied, which is the basis of active fault tolerant control method. This paper designs appropriate fault detection and fault diagnosis system against different fault types (consists of fault diagnosis algorithm based on optimal hyperplane and fault observer based on Extended Kalman Filter). Fault diagnosis algorithm based on the optimal hyperplane was majorly designed for the fault with the motor drive circuit board, which extracts its own state and monitors working conditions of the motor drive circuit board in real-time. While fault observer based on Extended Kalman Filter was majorly designed against rotor failure, real-time estimates the lift factor of each execution unit. This fault detection and diagnosis system can effectively complete fault isolation, fault identification and real-time monitoring of UAV’s status, improving aircraft flight safety levels.A Hex-Rotor UAV control system was established. First, the software and hardware of execution units and a speed-stabilizing control algorithm was designed to improve reliability. Next, the Hex-Rotor UAV attitude stabilization control problem was studied.A control algorithms based on backstepping sliding mode method was proposed,and control algorithm was proved convergent theoretically. O and was used to adjust sliding mode controller of the input control matrix, in order to improve attitude control stability. A dual-loop nested structure of pace tracking algorithm was raised, and virtual target point was used to reduce the influence of yaw angle error. Finally, the Hex-Rotor UAV’s self-reconfigurable control algorithm was researched, which is the core of active fault tolerant control. Self-reconfigurable controller will reconfigure the control structure depending on the fault information provided by fault detection and diagnosis systems. In this paper, two different reconfiguration controllers were adopted for gain type faults and failure conditions, and were proven stable.Finally, the work that had been done was summarized and issues for further research were raised.