Continuous and discrete time robust control for bipedal robot assuming minimal knowledge of the plant

Abstract

In this paper, a control approach in continuous and discrete time, for the robust tracking of walking patterns in a bipedal robot, is proposed. A discrete time model of the robot is obtained using the symplectic Euler method, in order to define a comparison between the proposed continuous and discrete time controllers in presence of parametric variations. Two robust controllers are designed using the Sliding Mode approach, one is based on the discrete time model and the other one on the continuous time model of the bipedal robot. The performance and robustness with respect to external disturbances and parametric variations of the proposed controllers, are demonstrated via Lyapunov stability analysis and simulations. By means of a discrete event model, the switching logic is simulated to represent the switching in the walking phases of the robot.