Abstract
The present paper treats the identification of nonlinear dynamical systems using Koopman-based deep state-space encoders. Through this method, the usual drawback of needing to choose a dictionary of lifting functions a priori is circumvented. The encoder represents the lifting function to the space where the dynamics are linearly propagated using the Koopman operator. An input-affine formulation is considered for the lifted model structure and we address both full and partial state availability. The approach is implemented using the the deepSI toolbox in Python. To lower the computational need of the simulation error-based training, the data is split into subsections where multi-step prediction errors are calculated independently. This formulation allows for efficient batch optimization of the network parameters and, at the same time, excellent long term prediction capabilities of the obtained models. The performance of the approach is illustrated by nonlinear benchmark examples.
Original language | English |
---|---|
Title of host publication | 60th IEEE Conference on Decision and Control |
Publisher | Institute of Electrical and Electronics Engineers |
Pages | 2288-2293 |
Number of pages | 6 |
ISBN (Electronic) | 978-1-6654-3659-5 |
DOIs | |
Publication status | Published - 1 Feb 2022 |
Event | 60th IEEE Conference on Decision and Control, CDC 2021 - Austin, TX, USA, Austin, United States Duration: 13 Dec 2021 → 17 Dec 2021 Conference number: 60 https://2021.ieeecdc.org/ |
Conference
Conference | 60th IEEE Conference on Decision and Control, CDC 2021 |
---|---|
Abbreviated title | CDC 2021 |
Country/Territory | United States |
City | Austin |
Period | 13/12/21 → 17/12/21 |
Internet address |