Design And Simulation Of Flight Control Method For Coaxial Octorotor Unmanned Aerial Vehicle

With the flight mission becoming more and more complex,researchers proposed a coaxial octorotor unmanned aerial vehicle(UAV)to improve the flight performance of quadrotor.Based on the space structure of quadrotor,this type of UAV adds four execution units in a coaxial way,which significantly increases the maneuverability and load capacity of the UAV.However,the multi-variable,nonlinearity,underactuation and strong coupling characteristics of coaxial octorotor UAV make its flight control a challenge.For its trajectory tracking control problem,three sliding mode control methods are designed in this thesis by sufficiently considering model uncertainties and external disturbances.The main contents are as follows:(1)The dynamic model of coaxial octorotor UAV is established based on Newton Euler formulaFirstly,the physical structure and motion form of coaxial octorotor UAV are analyzed.Secondly,the complete dynamic model of UAV is established based on Newton Euler formula,which provides the basis for the design of following control methods.(2)Trajectory tracking control of coaxial octorotor UAV based on a finite time sliding mode controlFirstly,the dynamic system is divided into position loop and attitude loop.Secondly,an extended state observer with finite time convergence is introduced to estimate the system velocities and the total disturbance terms in real time.Then,finite time sliding mode controllers are designed.Lastly,the effectiveness of the method is validated by a comparative simulation analysis.(3)Trajectory tracking control of coaxial octorotor UAV based on a neuro-adaptive sliding mode controlFirstly,in order to avoid the decoupling process,the dynamic system is divided into fully actuated subsystem and underactuated subsystem.Secondly,the radial basis function neural network is used to estimate the sliding mode surface coefficients and the total disturbance terms in real time.Then,neuro-adaptive sliding mode controllers are designed.Lastly,the effectiveness of the method is validated by a comparative simulation analysis.(4)Trajectory tracking control of coaxial octorotor UAV based on a self-tuning sliding mode controlFirstly,the system dynamics model of UAV with actuator fault is derived by defining the rotor failure factor.Secondly,in order to reduce the algorithm complexity of neural adaptive sliding mode control,the gradient descent method is used to self-tuning the sliding mode surface coefficients and reaching law coefficients.Then,the effectiveness of the method is validated by a comparative simulation analysis.Lastly,the three control methods proposed in this thesis are evaluated through quantitative indexes,and their advantages and disadvantages are described.