Research On Attitude Control Of Quadrotor Inverted Pendulum System

The quad-rotor inverted pendulum system is a nonlinear system with under-driving,strong coupling and multiple variable characteristics.And because the inverted pendulum system contains two degrees of freedom,and the four-rotor UAV system contains six degrees of freedom,the combined system includes a total of eight degrees of freedom.The controller experiment platform studied in this thesis not only has the uncertainty of the quad-rotor UAV model system,but also the instability of the inverted pendulum system,which can meet the current needs for verification of the control design of complex systems.It can also enhance the working efficiency of the quad-rotor UAV.It can be found in the existing results that most of the linearized mathematical models are used to design the control system.However,most of the controllers in the actual environment have complex structures,many parameters and are difficult to integrate,and when the controller parameters are selected improperly,the system will be disturbed by the outside world.In response to the above problems,this thesis proposes an LQR controller and particle swarm auto-disturbance rejection control optimized based on particle swarm optimization,and the algorithm is highly effective through simulation.The main research of this thesis has the following aspects:(1)First,the research significance and background of the four-rotor inverted pendulum system is introduced,and the equation of system dynamics is established.The system dynamics model mainly has two aspects,namely the mathematical model of the inverted pendulum and the quadrotor UAV,and the linearization is implemented at the model balance point of the established combined system.(2)Secondly,in order to solve the situation that the controller parameters are difficult to integrate,the particle swarm algorithm is used to optimize the LQR controller.Compared with the empirical method,the particle swarm algorithm,as a kind of artificial intelligence algorithm,aims at the premature convergence of the particle swarm and the weakening of the local search ability of the algorithm,the inertia weight,learning factor and current local optimal value are optimized to enhance the local search ability of the algorithm,and suppress the problem of premature convergence of the algorithm,and the effectiveness of the particle swarm optimization is fully verified in the simulation.(3)Third,the active disturbance rejection control is introduced,because the mathematical model is designed under ideal conditions,there will be some unavoidable internal and external disturbances.Active disturbance rejection control can control the system to reach a stable and balanced state without relying on precise mathematical models,but there are many parameters,and there is no practical parameter adjustment method.This thesis uses an improved particle swarm optimization to adjust the parameters.It is verified by simulation that the particle swarm auto-disturbance rejection control can make the system run more stable,with stronger anti-interference and robust characteristics.(4)Finally,verify whether the system under the two control methods can stabilize motion in the presence of disturbance and noise,analyze and compare the anti-interference and robustness of the system under the two control methods.It shows that the active disturbance rejection controller based on improved particle swarm optimization can control the quad-rotor inverted pendulum system more stably,making the system have stronger anti-interference ability and robustness,and better meet the system requirements.