On The High Maneuver Flight Control Of Multi-rotor Unmanned Aerial Vehicle System

Multi-rotor unmanned aerial vehicle(MUAV)is a typical multiple-input-multipleoutput underactuated system with couplings and nonlinearity.The MUAV often executes near boundary states,which are defined by physical structure,task requirements and security needs.It is a great challenge in control field about how to enhance the high maneuver response capability of the MUAV to track the desired attitude and trajectory in different flight conditions.In summary,this paper deals with the high maneuver flight control problem for the twelve-rotor UAV,considering external disturbances,input saturation,response speed,attitude constraints and sensor faults in the attitude and trajectory tracking tasks under multiple operating conditions,which lays a good foundation for autonomous flight of the UAV.The main research contents include the following aspects1)A mathematical model of the twelve-rotor UAV is built.The premise of MUAV high maneuver flight control is to describe the dynamic characteristics of the MUAV.Under the physical structure analysis and reasonable assumptions,the twelve-rotor UAV is seen as a rigid body with six-degree of freedom.Based on the equivalent input caused by wind disturbances,the mechanism and flight principle analysis,the UAV is divided into translational and rotational subsystems and the dynamic model is established by Newton-Euler equation.The mathematical model lays a good foundation for high maneuver flight control of the UAV.2)Robust attitude and trajectory tracking control tasks for the twelve-rotor UAV are accomplished.In consideration of external disturbances,air drag,model uncertainties,gyro torques and other negative factors acting on the UAV,a robust control strategy is needed to enhance the anti-disturbance ability and ensure the control performance of the UAV high maneuver flight task.In order to solve the above problems,a linear extended state observer is introduced to estimate the total perturbation in each channel and compensate the value to the controller to improve the robustness of the UAV.Controllers based on linear active disturbance rejection control algorithm and backstepping algorithm are designed to acquire desired attitude convergent property and accomplish the trajectory tracking task in the presence of total disturbances.3)A finite-time trajectory tracking control algorithm is proposed for the twelve-rotor UAV under input saturation.The high maneuver autonomous flight mission of the UAV requires a high level of the response speed for the system.Finite-time control strategy can meet the requirement theoretically,since it has the properties of fast convergence speed,high tracking precision and strong anti-disturbance ability.In consequence,for translational system of the UAV,finite-time backtepping controllers are designed to ensure the finite-time stability of the closed-loop system,and a finite-time auxiliary system is introduced to compensate the input saturation impact.The proposed finite-time trajectory tracking control algorithm can gaurantee the finite-time convergent property of the position tracking errors and reduce the influence of the input saturation from a theoretical level.The simulation experiments,semi-physical hardware in loop experiments and actual flight experiments all get good control performance,which illustrates the effectiveness of our algorithm.4)To solve the attitude constrained control problem for the twelve-rotor UAV high maneuver flight in the presence of the external disturbances and model uncertainties,an adaptive robust backstepping controller combined with Asymmetric Barrier Lyapunov Function(ABLF),adaptive neural network and dynamic surface control(DSC)is proposed.By introducing ABLF,the attitude angles are strictly maintained in the time-varying constrained sets all the time.The adaptive neural network is employed to approximate the model uncertainties of the UAV.The DSC avoids calculating the analytical derivative of the virtual control function in backstepping,which simplifies the algorithm and reduces the computation burden.In order to balance out the external disturbances and further reduce the approximation error of the neural network,a robust term is designed to compensate the above factors.The proposed control strategy can guarantee all the signals in the closed-loop attitude system bounded,meet the time-varying attitude constraints and improve the dynamic performance and tracking accuracy of the UAV attitude system.5)Aimed at sensor fault tolerance control problem of the twelve-rotor UAV high maneuver flight task,a novel siding mode controller based on disturbance observer and double power reaching law is proposed.The twelve-rotor UAV has enough hardware redundancy to ensure safe operation under actuator faults.While sensor faults will lead to the inaccuracy of state signals measurement,which still degrades the control performance of the UAV.The sensor fault is regarded as the equivalent mismatched disturbance of attitude system in this paper,and a disturbance observer is designed to online estimate the value of the fault.A novel sliding mode surface is designed by adding sensor fault estimation value to get the invariance property of the mismatched disturbance.A double power reaching law is employed to reduce the chattering phenomenon.The proposed control strategy has no need of fault detection and controller reconfiguration,which makes it less computation and easy implementation.The closed-loop stability is proved by Lyapunov theory and simulation experiments acquire approving control performance.