Online Planning for Autonomous Running Jumps Over Obstacles in High-Speed Quadrupeds

Abstract

Abstract-This paper presents a new framework for the generation of high-speed running jumps to clear terrain obstacles in quadrupedal robots. Our methods enable the quadruped to autonomously jump over obstacles up to 40 cm in height within a single control framework. Specifically, we propose new control system components, layered on top of a low-level running controller, which actively modify the approach and select stance force profiles as required to clear a sensed obstacle. The approach controller enables the quadruped to end in a preferable state relative to the obstacle just before the jump. This multi-step gait planning is formulated as a multiple-horizon model predictive control problem and solved at each step through quadratic programming. Ground reaction force profiles to execute the running jump are selected through constrained nonlinear optimization on a simplified model of the robot that possesses polynomial dynamics. Exploiting the simplified structure of these dynamics, the presented method greatly accelerates the computation of otherwise costly function and constraint evaluations that are required during optimization. With these considerations, the new algorithms allow for online planning that is critical for reliable response to unexpected situations. Experimental results, for a stand-alone quadruped with on-board power and computation, show the viability of this approach, and represent important steps towards broader dynamic maneuverability in experimental machines.

Trajectory Optimization With Implicit Hard Contacts

“…Each phase uses contact-consistent dynamics [8] and/or the corresponding contact constraints [1], [9]- [12]. While this eliminates the hybrid nature of the problem, the sequence of contact configurations (i.e., the gait) must be known a priori.…”

Dynamic Locomotion Through Online Nonlinear Motion Optimization for Quadrupedal Robots

“…Although robots are not yet fully capable of autonomous locomotion in such environments, the past years have shown that both the academia and the industry are taking big steps in the development of tools and methods which would allow this. The Cheetah 2 [18], a quadrupedal robot developed at MIT, has demonstrated its agile skills by jumping over unexpected obstacles using a fast online Model Predictive Controller (MPC). The IIT HyQ [20] quadruped has shown walking using a creeping gait [16] over stepping stones [23].…”

Core Actuation Promotes Self-manipulability on a Direct-Drive Quadrupedal Robot

“…Simplified models of quadrupedal platforms, such as the one depicted in Figure 1(a), often take the form of three-degree-of-freedom rigid bodies with common distances between the hips and mass center [16,17] and assume massless legs able to apply wrenches on the mass center when in contact with the ground subject to friction cone constraints. Following [6,7,8,9], we add core actuation to this model by introducing an actuated revolute joint to the body, depicted in Figure 1(b) (note that alternative formulations exist, such as [5]).…”