Maneuvering and stabilization control of a bipedal robot with a universal-spatial robotic tail

Abstract

This paper analyzes control methodologies to implement maneuvering and stabilization behaviors in a bipedal robot using a bioinspired robotic tail. Looking to nature, numerous animals augment their legs' functionality using a tail nature, numerous animals augment their legs' functionality using a tail to assist with both maneuvering and stabilization; looking to the robotics literature, previous research primarily focuses on single-mass, pendulum-like tails designed to perform a specific task. The overarching goal of this research is to study how bioinspired tail designs may be used in conjunction with low-complexity leg designs to achieve high-performance behaviors. In pursuit of this goal, this paper connects the serpentine universal-spatial robotic tail (USRT) with a biped consisting of a pair of Robotic Modular Legs to study the outer- and inner-loop control considerations necessary to achieve yaw-angle turning and stable leg lifting. The design and modeling of the tail and leg subsystems are presented, along with considerations for sensing the USRT's configuration in real-time. In addition, two inner-loop controllers that map desired tail trajectories into actuation commands are presented: a prescribed velocity approach that only utilizes motor feedback, and a prescribed torque approach that incorporates both feedforward consideration of the tail dynamics and feedback consideration from the tail sensing. Two outer-loop controllers-one for yaw-angle steering (maneuvering), and one for roll-angle disturbance rejection when lifting a foot (stabilization)-are also defined. Case studies including simulation and experimental results are used to validate the outer-loop control approaches.