Research On The Control Technology Of Two-wheel Self-balancing Car

The self-balanced two-wheel vehicle is a kind of complex system featured by nonlinear, multivariable, strongly coupled and parameters-uncertain which is suitable for working in poky, space because of its small size, simple structure and flexible movement. It is an ideal device to test the processing ability of various control methods, and has gained great attentions from scientists as one of the challenging issues.The self-balanced two-wheel vehicle is a system with the collocation of the two wheels. As the traditional inverted pendulum, the vehicle system relies on a control algorithm employing orientation and rate of state-changing detected by the obliquity sensor to maintain the balance. In this thesis, the state of the art is summarized. The controller of a self-balanced two-wheel vehicle system is designed by properly choosing the controller, the administrable motor and the sensor. The current orientation information is obtained by the obliquity sensor and transferred to the controller by the external circuits, the motor is driven by the output control signals, and thus the vehicle system can maintain self-balanced. Reference to the modeling of robot systems, a vehicle system model is established based on kinematics and Newtonian mechanics. The model is linearized by the feedback linearity correction method, and the stability, controllability and observability of the model are verified. In the controller design, the feasibility of traditional PID and LQR control of a simple model is analyzed; then a variable structure controller which based on the exponential reaching law is designed according to the variable structure control theory, the effects of external interference on the vehicle are analyzed. A linear quadratic optimal control method is adopted and the system output responses of the angle and the control voltage are analyzed by using different values of the parameter Q and R. The analysis results provide basics for the design of a stable large-range controllable linear controller.The simulations show that the LQR control method is fairly better than the PID control method. The vehicle system has comparatively great quiver under the variable structure controller, but still possesses strong robustness. Large-range control and certain robustness can be achieved under the optimal controller.