Details
| Original language | English |
|---|---|
| Title of host publication | Proceedings - IFAC-PapersOnLine |
| Pages | 692-698 |
| Number of pages | 7 |
| Volume | 54 |
| Publication status | Published - 2021 |
| Externally published | Yes |
Publication series
| Name | IFAC-PapersOnLine |
|---|---|
| Publisher | IFAC Secretariat |
| ISSN (Print) | 2405-8963 |
Abstract
The application of neural networks to non-linear dynamic system identification tasks has a long history, which consists mostly of autoregressive approaches. Autoregression, the usage of the model outputs of previous time steps, is a method of transferring a system state between time steps, which is not necessary for modeling dynamic systems with modern neural network structures, such as gated recurrent units (GRUs) and Temporal Convolutional Networks (TCNs). We compare the accuracy and execution performance of autoregressive and non-autoregressive implementations of a GRU and TCN on the simulation task of three publicly available system identification benchmarks. Our results show, that the non-autoregressive neural networks are significantly faster and at least as accurate as their autoregressive counterparts. Comparisons with other state-of-the-art black-box system identification methods show, that our implementation of the non-autoregressive GRU is the best performing neural network-based system identification method, and in the benchmarks without extrapolation, the best performing black-box method.
Keywords
- Benchmark Systems, Gated Recurrent Unit, Neural Networks, Nonlinear System Identification, Process Noise, Temporal Convolutional Network, Wiener-Hammerstein
ASJC Scopus subject areas
- Engineering(all)
- Control and Systems Engineering
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Proceedings - IFAC-PapersOnLine. Vol. 54 20. ed. 2021. p. 692-698 (IFAC-PapersOnLine).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Non-Autoregressive vs Autoregressive Neural Networks for System Identification
AU - Weber, Daniel
AU - Gühmann, Clemens
N1 - Publisher Copyright: Copyright © 2021 The Authors. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
PY - 2021
Y1 - 2021
N2 - The application of neural networks to non-linear dynamic system identification tasks has a long history, which consists mostly of autoregressive approaches. Autoregression, the usage of the model outputs of previous time steps, is a method of transferring a system state between time steps, which is not necessary for modeling dynamic systems with modern neural network structures, such as gated recurrent units (GRUs) and Temporal Convolutional Networks (TCNs). We compare the accuracy and execution performance of autoregressive and non-autoregressive implementations of a GRU and TCN on the simulation task of three publicly available system identification benchmarks. Our results show, that the non-autoregressive neural networks are significantly faster and at least as accurate as their autoregressive counterparts. Comparisons with other state-of-the-art black-box system identification methods show, that our implementation of the non-autoregressive GRU is the best performing neural network-based system identification method, and in the benchmarks without extrapolation, the best performing black-box method.
AB - The application of neural networks to non-linear dynamic system identification tasks has a long history, which consists mostly of autoregressive approaches. Autoregression, the usage of the model outputs of previous time steps, is a method of transferring a system state between time steps, which is not necessary for modeling dynamic systems with modern neural network structures, such as gated recurrent units (GRUs) and Temporal Convolutional Networks (TCNs). We compare the accuracy and execution performance of autoregressive and non-autoregressive implementations of a GRU and TCN on the simulation task of three publicly available system identification benchmarks. Our results show, that the non-autoregressive neural networks are significantly faster and at least as accurate as their autoregressive counterparts. Comparisons with other state-of-the-art black-box system identification methods show, that our implementation of the non-autoregressive GRU is the best performing neural network-based system identification method, and in the benchmarks without extrapolation, the best performing black-box method.
KW - Benchmark Systems
KW - Gated Recurrent Unit
KW - Neural Networks
KW - Nonlinear System Identification
KW - Process Noise
KW - Temporal Convolutional Network
KW - Wiener-Hammerstein
UR - http://www.scopus.com/inward/record.url?scp=85124626242&partnerID=8YFLogxK
U2 - 10.1016/j.ifacol.2021.11.252
DO - 10.1016/j.ifacol.2021.11.252
M3 - Conference contribution
VL - 54
T3 - IFAC-PapersOnLine
SP - 692
EP - 698
BT - Proceedings - IFAC-PapersOnLine
ER -