Physics-guided gated recurrent units for inversion-based feedforward control

TL;DR

This paper introduces a physics-guided gated recurrent unit (PG-GRU) framework for inversion-based feedforward control, combining physical models with neural networks to improve system identification and control accuracy in real-world experiments.

Contribution

It develops a novel PG-GRU architecture that integrates physical models with neural networks for improved inverse system identification and control.

Findings

Two-fold reduction in integral absolute error compared to linear feedforward.

Enhanced training convergence and interpretability through physics-guided learning.

Successful validation on a two-mass spring-damper system.

Abstract

Inversion-based feedforward control relies on an accurate model that describes the inverse system dynamics. The gated recurrent unit (GRU), which is a recent architecture in recurrent neural networks, is a strong candidate for obtaining such a model from data. However, due to their black-box nature, GRUs face challenges such as limited interpretability and vulnerability to overfitting. Recently, physics-guided neural networks (PGNNs) have been introduced, which integrate the prior physical model structure into the prediction process. This approach not only improves training convergence, but also facilitates the learning of a physics-based model. In this work, we integrate a GRU in the PGNN framework to obtain a PG-GRU, based on which we adopt a two-step approach to feedforward control design. First, we adopt stable inversion techniques to design a stable linear model of the inverse dynamics. Then, a GRU trained on the residual is tailored to inverse system identification. The resulting PG-GRU feedforward controller is validated by means of real-life experiments on a two-mass spring-damper system, where it demonstrates roughly a two-fold improvement compared to the linear feedforward and a preview-based GRU feedforward in terms of the integral absolute error.