| As an auxiliary device,attitude adjustment mechanism can control the balance and stability of unstable robots.In recent years,it has become an important research topic in the field of robotics.Common attitude adjustment mechanisms can be divided into the following types:inertia wheel,mass balancer,gyroscope,etc.Attitude regulating mechanism is a kind of multi-variable,nonlinear,strong coupling system,so it is of high value to study its balance control,position control and so on.In this paper,based on the two-wheeled robot,an attitude adjustment mechanism composed of weight and two connecting rods is designed.When the robot falls over,the connecting rod can be driven by the motor to rotate,so as to drive the end weight to move,and then the robot can be balanced control.This attitude regulating mechanism belongs to the mass balancer.Different from the inertia wheel and gyroscope,the attitude regulating mechanism designed in this paper does not need to change the torque and moment of inertia,so there is no need to rotate the acting wheel at high speed,which can greatly save energy consumption.Compared with the commonly used single-link mass balancer such as pendulum,rider model,etc.,the attitude adjustment mechanism designed in this paper controls the end weight through the structure of two-link,so that the end weight has a larger trajectory space,and the attitude adjustment is more flexible.This paper studies the attitude regulating mechanism from the aspects of mathematical modeling,model verification and control algorithm design.Firstly,the kinematic analysis of the robot system with attitude regulating mechanism is carried out to obtain the angular velocity and linear velocity of each joint of the robot.In addition,considering the non-holonomic constraints of the wheel and the ground,the Newton-Euler method is used to carry out the dynamics analysis of the system,and the mathematical model of the system is obtained.In this paper,the accuracy of mathematical model is verified by simulation experiment and real experiment.First,given different initial state values and input initial values of the mathematical model,observe its state response curve,so as to judge whether the mathematical model conforms to the physical law.Then the measurement and control system of attitude adjustment mechanism is designed and verified.Given the same initial value of the physical model and the mathematical model,their state curves are observed and compared to verify that the mathematical model is accurate near the equilibrium point.In this paper,some idealized assumptions are made in the modeling,so there are uncertainties in the model.When the uncertainty of the system appears in the input channel,it is the matching uncertainty.In order to solve the matching uncertainty,the robust adaptive dynamic programming algorithm transforms the design of the robust controller into the design of the optimal controller of the nominal system,and designs the optimal controller of the nominal system through the adaptive dynamic programming algorithm.When the uncertainty of the system does not appear in the input channel,it is a mismatch uncertainty.In order to solve the mismatch uncertainty,the robust adaptive dynamic programming algorithm transforms the design of the robust controller into the design of the optimal controller of the auxiliary system,and the optimal controller of the auxiliary system is designed by the adaptive dynamic programming algorithm.In this paper,the robust controller designed based on the robust adaptive dynamic programming algorithm is added to the attitude regulating mechanism system with uncertainty.Through simulation,it can be obtained that the system state trajectory under the action of the controller can achieve the final uniformly bounded,and compared with the adaptive dynamic programming algorithm,the effectiveness of the robust controller designed in this paper is proved. |
