Bipedal Walking with Corrective Actions in the Tilt Phase Space

Abstract

Many methods exist for a bipedal robot to keep its balance while walking. In addition to step size and timing, other strategies are possible that influence the stability of the robot without interfering with the target direction and speed of locomotion. This paper introduces a multifaceted feedback controller that uses numerous different feedback mechanisms, collectively termed corrective actions, to stabilise a core keypointbased gait. The feedback controller is experimentally effective, yet free of any physical model of the robot, very computationally inexpensive, and requires only a single 6-axis IMU sensor. Due to these low requirements, the approach is deemed to be highly portable between robots, and was specifically also designed to target lower cost robots that have suboptimal sensing, actuation and computational resources. The IMU data is used to estimate the yaw-independent tilt orientation of the robot, expressed in the so-called tilt phase space, and is the source of all feedback provided by the controller. Experimental validation is performed in simulation as well as on real robot hardware.

Mapless Humanoid Navigation Using Learned Latent Dynamics

“…The policy frequency is 10 Hz, while the simulation runs at 1 kHz. The robot incorporates a bipedal gait engine, which generates leg motions based on a target gait velocity V t [27,28]. The gait runs at 100 Hz.…”

Tilt Rotations and the Tilt Phase Space

“…Tilt rotations have been used in previous works, for example for the modelling of heading-independent balance states while walking [3], and for the modelling of foot orientations and ground contacts [4]. The tilt phase space has also been used as the fundamental basis for an entire bipedal walking feedback controller [2], which was possible only due to its unique set of properties. Tilt rotations were first formulated as part of the definition of fused angles and tilt angles, but were not significantly further investigated in their own right.…”

“…Due to the many advantageous properties of the tilt phase space (see Section V-C), and in particular due to magnitude axisymmetry, the tilt phase space was chosen as the basis of a developed feedback controller for the stabilisation of bipedal walking [2]. It is in the nature of feedback controllers, e.g.…”

“…The concepts of tilt rotations and the tilt phase space have also been applied in the keypoint gait generator that underlies the feedback gait presented in [2]. Most notably, the tilt phase space is used to separate the yaw and tilt of the feet at each gait keypoint, and then interpolate between them using the orientation cubic spline interpolation described in Section V-E.…”

“…They are highly relevant for situations like the analysis of balancing bodies in 3D-e.g. bipedal robots [2] [3], as the yaw component does not contribute to the body-local state of balance-but can also be used for other purposes. For example, tilt rotations can be used to represent the relative orientations of contacting bodies, or the relative rotations of two planar surfaces, such as for instance the foot of a robot and the ground.…”