Research Analyzer
← Back ICRA 2026

EquiMus: Energy-Equivalent Dynamic Modeling and Simulation of Musculoskeletal Robots Driven by Linear Elastic Actuators

Yinglei Zhu, Xuguang Dong, Qiyao Wang, Qi Shao, Fugui Xie, Xin-Jun Liu, Huichan Zhao

PDF

AI summary

Key figure (auto-extracted from paper)
EquiMus accurately simulates complex rigid-soft musculoskeletal dynamics by mapping elastic actuators to discrete rigid-body elements, enabling real-time control and learning-based strategies.
Musculoskeletal robots dynamic modeling energy-equivalent simulation MuJoCo rigid-soft hybrid reinforcement learning

Problem

Classical rigid-body models and existing simulators fail to capture configuration-dependent mass redistribution, actuator inertia, and kinematic loops in hybrid musculoskeletal robots, making accurate dynamic modeling and real-time simulation difficult.

Approach

The authors propose EquiMus, an energy-equivalent modeling framework that discretizes linear elastic actuators into a three-mass-point lumped-mass system with matched stiffness, damping, and constraints, implemented in the MuJoCo physics engine.

Key results

  • Compact energy-equivalent lumped-mass formulation preserving actuator inertia
  • MuJoCo implementation with kinematic loop support and real-time speed
  • Close sim-to-real agreement validated on a pneumatic bionic robotic leg
  • Enables downstream applications including PID auto-tuning and reinforcement learning

Why it matters

Provides a computationally efficient and physically accurate simulation platform for designing and controlling complex bio-inspired musculoskeletal robots.

Abstract

Dynamic modeling and control are critical for un- leashing soft robots’ potential, yet remain challenging due to their complex constitutive behaviors and real-world operating conditions. Bio-inspired musculoskeletal robots, which integrate rigid skeletons with soft actuators, combine high load-bearing capacity with inherent flexibility. Although actuation dynamics have been studied through experimental methods and surrogate models, accurate and effective modeling and simulation remain a significant challenge, especially for large-scale hybrid rigid–soft robots with continuously distributed mass, kinematic loops, and diverse motion modes. To address these challenges, we propose EquiMus, an energy- equivalent dynamic modeling framework and MuJoCo-based simulation for musculoskeletal rigid–soft hybrid robots with linear elastic actuators. The equivalence and effectiveness of the proposed approach are validated and examined through both simulations and real-world experiments on a bionic robotic leg. EquiMus further demonstrates its utility for downstream tasks, including controller design and learning-based control strategies.

Index terms

Modeling Control and Learning for Soft Robots Biologically-Inspired Robots Dynamics

Related papers