Dynamic Adaptive Legged Locomotion Policy via Decoupling Reaction Force Control and Gait Control

2025-09-17

AI Summary

To address the poor adaptability and insufficient robustness of legged robots in unknown environments and sim-to-real transfer, this paper proposes a dynamic adaptive control strategy that decouples reaction force control from gait control. The core innovation lies in separately modeling the stance phase (force control) and swing phase (gait planning), enabling rapid online disturbance rejection and substantially reducing reliance on domain randomization. Within a reinforcement learning framework, the two controllers are independently optimized in simulation and deployed directly onto physical hardware without fine-tuning. Experiments demonstrate stable locomotion under challenging conditions—including horizontal external force disturbances, uneven terrain, heavy payloads, and asymmetric loading—while achieving significantly improved sim-to-real transfer success rates. The method thus exhibits strong generalization capability and practical engineering applicability.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

While Reinforcement Learning (RL) has achieved remarkable progress in legged locomotion control, it often suffers from performance degradation in out-of-distribution (OOD) conditions and discrepancies between the simulation and the real environments. Instead of mainly relying on domain randomization (DR) to best cover the real environments and thereby close the sim-to-real gap and enhance robustness, this work proposes an emerging decoupled framework that acquires fast online adaptation ability and mitigates the sim-to-real problems in unfamiliar environments by isolating stance-leg control and swing-leg control. Various simulation and real-world experiments demonstrate its effectiveness against horizontal force disturbances, uneven terrains, heavy and biased payloads, and sim-to-real gap.

Problem

• Addresses performance degradation in out-of-distribution legged locomotion
• Mitigates sim-to-real gap through decoupled stance and swing control
• Enhances robustness against force disturbances and uneven terrains

Innovation

• Decoupled stance and swing leg control
• Online adaptation to unfamiliar environments
• Mitigates sim-to-real gap without domain randomization