Reduced Basis Methods for Efficient Simulation of a Rigid Robot Hand Interacting with Soft Tissue

TL;DR

This paper develops reduced basis methods to enable real-time simulation and control of a robot hand interacting with soft tissue, significantly reducing computational costs while maintaining accuracy.

Contribution

It introduces structure-preserving and non-structure-preserving reduced basis techniques for coupled elastic-robot interaction problems, including efficient low-rank solutions for high-dimensional Riccati equations.

Findings

Reduced basis methods achieve significant speedups in simulations.

Low-rank solutions effectively approximate high-dimensional Riccati equations.

Numerical examples demonstrate high approximation quality and computational efficiency.

Abstract

We present efficient reduced basis (RB) methods for the simulation of the coupled problem consisting of a rigid robot hand interacting with soft tissue material which is modeled by the linear elasticity equation and discretized with the Finite Element Method. We look at two different scenarios: (i) the forward simulation and (ii) a feedback control formulation of the model. In both cases, large-scale systems of equations appear, which need to be solved in real-time. This is essential in practice for the implementation in a real robot. For the feedback-scenario, in the context of the linear quadratic regulator, we encounter a high-dimensional Algebraic Riccati Equation (ARE). To overcome the real-time constraint by significantly reducing the computational complexity, we use several structure-preserving and non-structure-preserving reduction methods. These include proper orthogonal decomposition-based reduced basis techniques. For the ARE, instead of solving a full dimensional problem we compute a low-rank-factor and hence a low-dimensional ARE is solved. Numerical examples for both cases are provided. These illustrate the approximation quality of the reduced solution and speedup factors of the different reduction approaches.