Offline Trajectory Generation for Bipedal Robot Using Linear Inverted Pendulum Model

Abstract

Reduced order model (ROM)-based controllers have proved to be effective to generate stable bipedal locomotion. However, it is important to understand the limitations and effectiveness of these models without implementing any controllers. This study highlights the versatility of the Linear Inverted Pendulum Model (LIPM) at various walking speeds. Firstly, the Centre of Mass (COM) trajectory has been generated using the LIPM model, and the foot motion trajectory has been created using a sixth-order polynomial function. The trajectory is generated using a predefined step length, speed of locomotion and COM height. Secondly, the task space trajectory has been converted into a joint space trajectory through inverse kinematics for a 6-degree-of-freedom leg. To facilitate the proper walking motion the contact between the foot sole and the ground is implemented. Finally, a simple bipedal robot in MATLAB/Simulink has been modelled and the generated trajectories were implemented.