Design and Implementation of Two-wheeled Self-Balancing Robot based on LQR

The two-wheeled self-balancing robot is a typical under-actuated system, which has the characteristics of nonlinearity, multivariable, strong coupling and high instability. This paper constructs a relatively precise mathematical model for the two-wheeled self-balancing robot system by applying the Newton-Euler method, specifically tailored to its unique characteristics. Then, based on this model, a Linear Quadratic Regulator (LQR) controller is designed, and the theoretical effectiveness of the LQR controller is verified through simulation experiments. On this basis, this paper further proposes a practical control scheme for practical applications, and elaborates on the implementation details based on the LQR controller. The experimental results show that this control scheme exhibits excellent control performance, providing a computationally efficient, low-cost, and effective solution for others to implement two-wheeled self-balancing robot.