Biped gait controller with active perturbation recovery.

Abstract

The goals of this work are to develop a biped gait controller and an active fall recovery algorithm, both based on the capture point stability theory with the criterion of N-step capturability, validate this algorithms via simulations; and to design a biped robot that represents gait on the sagittal plane, capable of serving as a test-bed for legged locomotion research. The effectiveness of the controller and fall recovery are assessed through simulations, with the designed robot, that imposes different types of perturbations, such as: pushes, ramps and irregular terrain. The controller ability to modulate the robot's velocity is also demonstrated. The fall recovery algorithm is addressed as a solution to increase robustness without significant detriment to energy efficiency. The controller was proved successful in the simulations, it only demonstrated speed limitation on descending slopes. The fall recovery manages to significantly increase robustness of a standard, energy efficient gait.