2024 |
Hyuntae Kim / Hamin Chang / Hyungbo Shim Evaluating MR-GPR and MR-NN: An Exploration of Data-driven Control Methods for Nonlinear Systems Journal Article In: International Journal of Control, Automation, and Systems, vol. 22, pp. 2934–2941, 2024, ISSN: 1598-6446. Abstract | Links | BibTeX | Tags: Data-driven control, Gaussian process, Neural Network, nonlinear systems, Stability, System identification @article{nokey, This paper addresses the challenge of data-driven control of nonlinear systems, focusing on the limi-tations and capabilities of model reference Gaussian process regression (MR-GPR) and its evolved counterpart,model reference neural networks (MR-NN). MR-GPR, based on Gaussian processes renowned for their adaptabil-ity to diverse data structures, encounters scalability issues especially when handling large datasets. To address theselimitations, this paper introduces MR-NN, an extension of MR-GPR, leveraging neural networks (NN) to managelarge datasets and capture complex nonlinear dynamics effectively. We present a comprehensive evaluation of bothmethods through a classical control problem of the inverted pendulum, a system well-recognized for its nonlinearbehavior. Numerical experiments are conducted to compare the methods in terms of control performance, compu-tational efficiency, and reliability. |
List of English Publication
2024 |
Evaluating MR-GPR and MR-NN: An Exploration of Data-driven Control Methods for Nonlinear Systems Journal Article In: International Journal of Control, Automation, and Systems, vol. 22, pp. 2934–2941, 2024, ISSN: 1598-6446. |