Data-Driven Soft Robot Control via Adiabatic Spectral Submanifolds

2025-03-13

AI Summary

Soft robotic systems exhibit strongly nonlinear, high-dimensional dynamics that severely limit the performance of model-based control—particularly for complex spatial trajectory tracking. To address this, we introduce, for the first time, the theory of adiabatic spectral submanifolds (aSSMs) into data-driven modeling and control of soft robots, proposing the aSSM-MPC framework. Leveraging the fast-decaying intrinsic vibrational modes, aSSMs automatically extract physically interpretable, low-dimensional attracting invariant manifolds, enabling equation-free model order reduction while rigorously guaranteeing closed-loop stability. The method synergistically integrates aSSM theory, data-driven manifold learning, and model predictive control. Validated on high-fidelity finite element simulations, the framework achieves accurate trajectory tracking using only a 4–5 dimensional aSSM model, improving tracking accuracy by up to one order of magnitude over state-of-the-art data-driven approaches.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Learning with Manifolds
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The mechanical complexity of soft robots creates significant challenges for their model-based control. Specifically, linear data-driven models have struggled to control soft robots on complex, spatially extended paths that explore regions with significant nonlinear behavior. To account for these nonlinearities, we develop here a model-predictive control strategy based on the recent theory of adiabatic spectral submanifolds (aSSMs). This theory is applicable because the internal vibrations of heavily overdamped robots decay at a speed that is much faster than the desired speed of the robot along its intended path. In that case, low-dimensional attracting invariant manifolds (aSSMs) emanate from the path and carry the dominant dynamics of the robot. Aided by this recent theory, we devise an aSSM-based model-predictive control scheme purely from data. We demonstrate the effectiveness of this data-driven model on various dynamic trajectory tracking tasks on a high-fidelity and high-dimensional finite-element model of a soft trunk robot. Notably, we find that four- or five-dimensional aSSM-reduced models outperform the tracking performance of other data-driven modeling methods by a factor up to 10 across all closed-loop control tasks.

Problem

• Addresses challenges in model-based control of soft robots.
• Develops control strategy using adiabatic spectral submanifolds (aSSMs).
• Demonstrates improved trajectory tracking with data-driven aSSM models.

Innovation

• Data-driven control using adiabatic spectral submanifolds
• Model-predictive strategy for nonlinear soft robot dynamics
• Low-dimensional aSSM models enhance trajectory tracking