Forward-Backward Extended DMD with an Asymptotic Stability Constraint

TL;DR

This paper introduces a data-driven approach to identify asymptotically stable Koopman models from noisy data by combining forward and backward dynamics and enforcing stability constraints, improving model reliability.

Contribution

It proposes a novel semidefinite programming method that ensures asymptotic stability in Koopman system identification from noisy data, addressing stability issues in existing models.

Findings

Outperforms state-of-the-art methods on simulated Duffing oscillator data.

Successfully applied to experimental soft robot data.

Ensures stability despite noisy measurements.

Abstract

This paper presents a data-driven method to identify an asymptotically stable Koopman system from noisy data. In particular, the proposed approach combines approximations of the system's forward- and backward-in-time dynamics to reduce bias caused by noisy data while enforcing asymptotic stability. A Koopman model of an inherently asymptotically stable system can be unstable due to noisy data and a poor choice of lifting functions. To prevent identifying an unstable model, the proposed approach imposes an asymptotic stability constraint on the Koopman model. The proposed method is formulated as a semidefinite program and its performance is compared to state-of-the-art methods with a simulated Duffing oscillator dataset and experimental soft robot dataset.