Locomotion Control of a Bipedal Wheeled Robot Using Virtual Model Control and Linear Quadratic Regulator Techniques

This paper aims to develop a balance control technique and investigates its impact on the stability and disturbance rejection capability of a bipedal wheeled robot. The bipedal wheeled robot is equivalent to a wheeled inverted pendulum nonlinear model with a legs-airframe centroid variable rod. The nonlinear model is linearized and decoupled into two subsystems: straight-line control using linear-quadratic regulator (LQR) for balance and speed, and steering control employing proportional integral derivative (PID). Height control adjusts the virtual force with PID-Feedforward, while hip torque is determined by virtual model control (VMC). MATLAB simulation confirms effective control of height, linear motion, and steering, with decoupling enhancing steering performance.