Recent Progress in Legged Robots Locomotion Control

Abstract

Purpose of review In recent years, legged robots locomotion has been transitioning from mostly flat ground in controlled settings to generic indoor and outdoor environments, approaching now real industrial scenarios. This paper aims at documenting some of the key progress made in legged locomotion control that enabled this transition.Recent findings Legged locomotion control makes extensive use of numerical trajectory optimization and its online implementation, model predictive control. A key progress has been how this optimization is handled, with refined models and refined numerical methods. This led the legged locomotion research community to heavily invest in and contribute to the development of new optimization methods and efficient numerical software. SummaryWe present an overview of the typical approach to legged locomotion control, which involves primarily planning a sequence of contacts with the environment and computing a corresponding dynamically feasible trajectory of the Center of Mass of the robot. We then detail recent progress in contact planning and trajectory optimization with either the full Lagrangian dynamics of legged robots or with reduced models.

Full-Body Torque-Level Non-linear Model Predictive Control for Aerial Manipulation

“…Alternatively, for differentially flat 1 systems, one could generate a sufficiently smooth trajectory in the flat output 1 Differential flatness is a property of a system that allows us to express the states and inputs as a function of several outputs and their higher order derivatives. These are the so-called flat outputs space.…”

“…The humanoids and the legged robots communities have undergone all these steps and converged to strategies of full-body optimal predictive control at the torque level. These formulations are at the same time generic, intuitive and powerful, have demonstrated their fitness to board many of the control challenges for such complex robots [1], especially when high dynamics are involved. In contrast, we have not seen these methods used in UAMs.…”

“…In view of this, in this paper we walk through the pro-cess of adapting the full-body, torque-level, optimal control methods [1], [2], used in the communities of humanoids and legged robots to the field of UAMs. The use of full-body nonlinear model predictive control (nMPC) simplifies the control architecture since it unifies several functional blocks.…”

Implicit Differential Dynamic Programming

“…Optimal control is a very convenient way to formally and practically compute the movement that a robot has to achieve [1], [2], [3], [4]. Among the various possibilities to numerically solve an optimal control problem, the differential dynamic programming [5], [6] algorithm provides several advantages that makes it a frequent choice to solve optimization trajectory problems.…”

Terrain-Perception-Free Quadrupedal Spinning Locomotion on Versatile Terrains: Modeling, Analysis, and Experimental Validation

“…The current quadruped robots can mimic not only static gaits of animals but also highly agile and dynamic behaviors, such as galloping, jumping, and back-flipping ( Katz et al, 2019 ; Kim et al, 2019 ), which enable them to traverse unstructured terrains ( Bledt et al, 2018 ; Kim et al, 2020 ; Jenelten et al, 2020 ). Yet, certain locomotion behaviors have not been explored, e.g., the circular spinning locomotion ( Carpentier and Wieber, 2021 ). Dogs often spin to inspect the environment and search for potential threats ( Park et al, 2005 ; Chen et al, 2017 ).…”