Single-Stage Optimization of Open-loop Stable Limit Cycles with Smooth, Symbolic Derivatives
Authors:
Muhammad Saud Ul Hassan,
Christian Hubicki
Abstract:
Open-loop stable limit cycles are foundational to legged robotics, providing inherent self-stabilization that minimizes the need for computationally intensive feedback-based gait correction. While previous methods have primarily targeted specific robotic models, this paper introduces a general framework for rapidly generating limit cycles across various dynamical systems, with the flexibility to i…
▽ More
Open-loop stable limit cycles are foundational to legged robotics, providing inherent self-stabilization that minimizes the need for computationally intensive feedback-based gait correction. While previous methods have primarily targeted specific robotic models, this paper introduces a general framework for rapidly generating limit cycles across various dynamical systems, with the flexibility to impose arbitrarily tight stability bounds. We formulate the problem as a single-stage constrained optimization problem and use Direct Collocation to transcribe it into a nonlinear program with closed-form expressions for constraints, objectives, and their gradients.
Our method supports multiple stability formulations. In particular, we tested two popular formulations for limit cycle stability in robotics: (1) based on the spectral radius of a discrete return map, and (2) based on the spectral radius of the monodromy matrix, and tested five different constraint-satisfaction formulations of the eigenvalue problem to bound the spectral radius. We compare the performance and solution quality of the various formulations on a robotic swing-leg model, highlighting the Schur decomposition of the monodromy matrix as a method with broader applicability due to weaker assumptions and stronger numerical convergence properties.
As a case study, we apply our method on a hopping robot model, generating open-loop stable gaits in under 2 seconds on an Intel Core i7-6700K, while simultaneously minimizing energy consumption even under tight stability constraints.
△ Less
Submitted 17 September, 2024; v1 submitted 17 December, 2023;
originally announced December 2023.
