Variable-Impedance Muscle Coordination under Slow-Rate Control Frequencies and Limited Observation Conditions Evaluated through Legged Locomotion

2025-12-03

AI Summary

This study addresses the poor robustness of monopedal hopping under low-frequency control and limited perception, along with excessive computational burden on high-level controllers. We propose a hierarchical control framework based on variable-impedance muscle synergies. Methodologically, we develop a biologically plausible musculoskeletal model incorporating both mono- and bi-articular muscles, integrate a reinforcement learning–based high-level controller with a morphology-computation-driven low-level muscle synergy mechanism, and explicitly simulate biological constraints—including sensory delay, partial observability, and surrogate sensory signals. Our key contribution is demonstrating that muscle synergies can actively offload high-frequency feedback tasks via mechanical-level morphological computation, thereby substantially reducing the high-level controller’s dependence on high sampling rates and complete sensory information. Experiments show stable hopping performance even under severely reduced control frequencies (≤10 Hz) and diverse perception-limited conditions, validating the framework’s enhanced control robustness and improved policy learning efficiency.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Human motor control remains agile and robust despite limited sensory information for feedback, a property attributed to the body's ability to perform morphological computation through muscle coordination with variable impedance. However, it remains unclear how such low-level mechanical computation reduces the control requirements of the high-level controller. In this study, we implement a hierarchical controller consisting of a high-level neural network trained by reinforcement learning and a low-level variable-impedance muscle coor dination model with mono- and biarticular muscles in monoped locomotion task. We systematically restrict the high-level controller by varying the control frequency and by introducing biologically inspired observation conditions: delayed, partial, and substituted observation. Under these conditions, we evaluate how the low-level variable-impedance muscle coordination contributes to learning process of high-level neural network. The results show that variable-impedance muscle coordination enables stable locomotion even under slow-rate control frequency and limited observation conditions. These findings demonstrate that the morphological computation of muscle coordination effectively offloads high-frequency feedback of the high-level controller and provide a design principle for the controller in motor control.

Problem

• Variable-impedance muscle coordination reduces high-level control requirements
• It enables stable locomotion under slow control frequencies and limited observations
• Morphological computation offloads high-frequency feedback from the neural controller

Innovation

• Hierarchical controller with high-level neural network and low-level variable-impedance muscle coordination
• Variable-impedance muscle coordination enabling stable locomotion under slow control frequencies
• Morphological computation offloading high-frequency feedback from high-level controller