Minimalistic control of biped walking in rough terrain

Abstract

Toward our comprehensive understanding of legged locomotion in animals and machines, the compass gait model has been intensively studied for a systematic investigation of complex biped locomotion dynamics. While most of the previous studies focused only on the locomotion on flat surfaces, in this article, we tackle with the problem of bipedal locomotion in rough terrains by using a minimalistic control architecture for the compass gait walking model. This controller utilizes an open-loop sinusoidal oscillation of hip motor, which induces basic walking stability without sensory feedback. A set of simulation analyses show that the underlying mechanism lies in the "phase locking" mechanism that compensates phase delays between mechanical dynamics and the open-loop motor oscillation resulting in a relatively large basin of attraction in dynamic bipedal walking. By exploiting this mechanism, we also explain how the basin of attraction can be controlled by manipulating the parameters of oscillator not only on a flat terrain but also in various inclined slopes. Based on the simulation analysis, the proposed controller is implemented in a real-world robotic platform to confirm the plausibility of the approach. In addition, by using these basic principles of self-stability and gait variability, we demonstrate how the proposed controller can be extended with a simple sensory feedback such that the robot is able to control gait patterns autonomously for traversing a rough terrain.

Dynamic stability of individuals with transtibial amputation walking in destabilizing environments

“…Visual and somatosensory feedback are also thought to be critical for walking (Gandevia, 2001). However, walking and standing are very different tasks (Kang and Dingwell, 2006) and passive dynamics likely play a greater role during walking (Srinivasan and Ruina, 2006; Iida and Tedrake, 2010). Patients with unilateral transtibial amputation retain intact visual and vestibular input, and mostly intact proprioception and motor function in their proximal impaired limb and contralateral limb.…”

Toward fast policy search for learning legged locomotion

“…The hip can exert a torque of [−20, 20] Nm. The swingleg dynamics of the compass gait model are given as [9] M(q)q + C(q,q)q + G(q) = Bu , where M is the mass matrix, C subsumes the centripetal and Coriolis forces, and G is the gravity matrix. The contact point of the stance leg with the ground is a zero-torque, frictionless joint.…”

Human-inspired walking via unified PD and impedance control

“…Due to the complexity present in achieving bipedal robotic walking, its study is often split into two extremes: theoretical results aimed at developing torque controllers (e.g., controlled symmetries [18], geometric reduction [6], [17], inverted pendulum [14], [10]) that are provably correct, and simulation/experimental results guided by heuristics (e.g., ZMP methods [21], [22], passivity based control [16], [11], reinforcement learning [12] and the central pattern generators [15]) that provide better real world behavior than complex nonlinear controllers can achieve. Both of these extremes are important in the study of robotic walking, yet to achieve truly human-like walking it is necessary to bridge the gap between these two methodologies.…”