A kind of bicycle robot dynamic modeling and nonlinear control

Abstract

The bicycle robot is a natural unstable nonlinear system. Its inherent nonlinear and deeply coupling characters of bicycle robot make the problem of control it much more difficult. Taking the steering torque and the driving torque of rear wheel as system inputs, the goal of bicycle robot self balancing was achieved. The rolling angle and the steering angle were taken as system generalized coordinates. A kind of two-input and two-output nonlinear dynamic model of the bicycle robot was presented based on Lagrange method. A multi-input and multi-output affine nonlinear system was presented to describe the dynamic model. And the input-output exactitude linearization theory was applied to analyze the affine nonlinear system. Then a control algorithm was proposed to achieve the goal of decoupling and accurate linearization. Computer simulation verified the validity of the dynamic model. The effect of the controller is testified by simulation experiments.