Stable CoM Trajectory Generation for Biped Robots based on Heuristic Capture Point Tracking Approach

Biped locomotion is a complex phenomenon. As the bipedal robotic system possesses multiple degrees of freedom (DoF), the resulting dynamics is highly nonlinear and coupled. However for practical purposes, reduced and simplified dynamical models such as Linear Inverted Pendulum Model (LIPM) have been widely used for gait generation. In this work, an LIPM based Capture Point (CP) method has been worked out to achieve stable and robust walk for biped robot. An existing algorithm has been modified to avoid infeasibility of solution while generating the desired trajectories of Zero Moment Point (ZMP), CP and Center of Mass (CoM) for stable bipedal locomotion and the efficacy of the approach is validated through simulations. It is shown that the modification also offers an approach to adjust footprint up to a certain extent according to design requirements or terrain conditions.