Research Analyzer
← Back ICRA 2026

Adaptive Non-Linear Centroidal MPC with Stability Guarantees for Robust Locomotion of Legged Robots

Mohamed Elobaid, Giulio Turrisi, Lorenzo Rapetti, Giulio Romualdi, Stefano Dafarra, Tomohiro Kawakami, Tomohiro Chaki, Takahide Yoshiike, Claudio Semini, Daniele Pucci

PDF

AI summary

Key figure (auto-extracted from paper)
Reformulating online centroidal MPC with adaptive control and Lyapunov-based constraints guarantees closed-loop stability and robustness against constant disturbances for legged robots.
Centroidal MPC Adaptive Control Legged Robots Stability Guarantees Robust Locomotion Control Lyapunov Functions

Problem

Standard nonlinear centroidal MPC controllers for legged robots lack rigorous closed-loop stability and robustness guarantees, particularly when facing unmeasured disturbances like unknown payloads or external pushes.

Approach

The authors combine adaptive control theory and Control Lyapunov Functions to reformulate the online centroidal MPC, adding explicit stability constraints and an adaptation law for unknown constant disturbances while preserving friction cone and contact constraints.

Key results

  • Rigorous closed-loop stability and robustness guarantees for centroidal MPC under constant disturbances
  • A streamlined adaptation law and stabilizing constraints without linearization or multi-stage optimization
  • Experimental validation on a 56.7 kg humanoid (ergoCub) and a 21 kg quadruped (Aliengo) under various disturbances
  • Open-source code for reproducing locomotion experiments

Why it matters

Enables safe, robust, and provably stable locomotion for general-purpose legged robots operating in uncertain real-world environments, benefiting both robotics researchers and industry developers.

Abstract

Nonlinear model predictive locomotion controllers based on the reduced centroidal dynamics are nowadays ubiq- uitous in legged robots. These schemes, even if they assume an inherent simplification of the robot’s dynamics, were shown to endow robots with a step-adjustment capability in reaction to small pushes, and in the case of uncertain parameters - as un- known payloads - they were shown to provide some “practical”, albeit limited, robustness. In this work, we provide rigorous certificates of their closed-loop stability via reformulating the online centroidal MPC controller. This is achieved thanks to a systematic procedure inspired by the machinery of adaptive control, together with ideas coming from Control Lyapunov Functions. Our reformulation, in addition, provides robustness for a class of unmeasured constant disturbances. To demonstrate the generality of our approach, we validated our formulation on a new generation of humanoid robots - the 56.7 kg ergoCub, as well as on the commercially available 21 kg quadruped robot Aliengo.

Index terms

Legged Robots Optimization and Optimal Control Robust/Adaptive Control

Related papers