Modeling And Control Of Drone-Inverted Pendulum System

With the continuous development of modern control theory,the ability of machine automation is also constantly improving.In order to adapt to the increasingly advanced control theory and technology,the inverted pendulum system has become a popular platform for researchers to verify the theory and technical feasibility.There are more and more control methods for inverted pendulum,such as traditional PID control,genetic optimization control,sliding mode control,linear quadratic optimal control and so on.In order to promote the continuous advancement of control technology,the inverted pendulum system is becoming more and more complicated,not only reflected in the increase in the number of inverted pendulums,but also in the diversity of the inverted pendulum platform,such as the inverted pendulum on the trolley,the inverted pendulum on the ball and inverted pendulum on a circular orbit,etc.The subject of this paper is the inverted pendulum on a quadrotor.This special form of inverted pendulum is more complex than the inverted pendulum on the straight line,so it has good research significance.The main research contents of this paper are as follows:(1)Modeling and analysis of the quadrotor-inverted pendulum system is mainly to divide the four-rotor inverted pendulum system into two separate systems for modeling and analysis,which greatly simplifies the modeling.When modeling the system,it is necessary to ignore the uncertainties in the actual environment,so the modeling analysis is carried out under the assumptions.Finally,the errors caused by these assumptions are analyzed to improve the accuracy of the model.(2)The control algorithm of the quadrotor-inverted pendulum system is analyzed and designed.Since the inverted pendulum and the quadrotor are two separate control systems,and the external interference is high,the control quality is required,so the cascade controller of three loops is used here.They are a four-rotor position loop using an auto disturbance rejection controller,an inverted pendulum control loop,and a four-rotor attitude loop based on a Kalman filter and a PID controller.In addition,in the actual conditions,the system involves many communication loops,so there will be communication delay problems.For this problem,communication delay compensation is added to improve the controllability of the system.(3)After system modeling and analysis,the Simulink system is simulated on matlab software.The control loop of the inverted pendulum and the position loop of the quadrotor are simulated firstly to verify that the auto disturbance rejection controller has good robustness on this system.Finally,the whole control system is integrated for simulation.The experimental results show that the above control algorithm has good control performance and robustness.(4)The physical test platform of the quadrotor-inverted pendulum is designed from hardware and software.The position information of the four-rotor and the inverted pendulum is captured by the indoor positioning system OptiTrack,and then the position information is sent to the quadrotor,and the inertial measurement of the drone itself.The unit also takes the attitude information as input,which causes the system to form a closed loop control.Finally,test the feasibility of the system experimental platform.