Details
| Original language | English |
|---|---|
| Pages (from-to) | 656-661 |
| Number of pages | 6 |
| Journal | Procedia CIRP |
| Volume | 99 |
| Early online date | 3 May 2021 |
| Publication status | Published - 2021 |
| Event | 14th CIRP Conference on Intelligent Computation in Manufacturing Engineering, CIRP ICME 2020 - Naples, Italy Duration: 15 Jul 2020 → 17 Jul 2020 |
Abstract
In the field of industry 4.0 the number of sensors increases steadily. The sensor data is often used for system observation and estimation of the system parameters. Typically, Kalman filtering is used for determination of the internal system parameters. Their accuracy and robustness depends on the system knowledge, which is described by differential equations. We propose a self-configurable filter (FNN-EKF) which estimates the internal system behavior without knowledge of the differential equations and the noise power. Our filter is based on Kalman filtering with a constantly adapting neural network for state estimation. Applications are denoising sensor data or time series. Several bouncing ball simulations are realized to compare the estimation performance of the Extended Kalman Filter to the presented FNN-EKF.
Keywords
- Adaptive extended Kalman filter, Denoising, Neural Network, Robustness, stability metric
ASJC Scopus subject areas
- Engineering(all)
- Control and Systems Engineering
- Engineering(all)
- Industrial and Manufacturing Engineering
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Procedia CIRP, Vol. 99, 2021, p. 656-661.
Research output: Contribution to journal › Conference article › Research › peer review
}
TY - JOUR
T1 - Estimation of unknown system states based on an adaptive neural network and Kalman filter
AU - Kellermann, Christoph
AU - Ostermann, Jörn
PY - 2021
Y1 - 2021
N2 - In the field of industry 4.0 the number of sensors increases steadily. The sensor data is often used for system observation and estimation of the system parameters. Typically, Kalman filtering is used for determination of the internal system parameters. Their accuracy and robustness depends on the system knowledge, which is described by differential equations. We propose a self-configurable filter (FNN-EKF) which estimates the internal system behavior without knowledge of the differential equations and the noise power. Our filter is based on Kalman filtering with a constantly adapting neural network for state estimation. Applications are denoising sensor data or time series. Several bouncing ball simulations are realized to compare the estimation performance of the Extended Kalman Filter to the presented FNN-EKF.
AB - In the field of industry 4.0 the number of sensors increases steadily. The sensor data is often used for system observation and estimation of the system parameters. Typically, Kalman filtering is used for determination of the internal system parameters. Their accuracy and robustness depends on the system knowledge, which is described by differential equations. We propose a self-configurable filter (FNN-EKF) which estimates the internal system behavior without knowledge of the differential equations and the noise power. Our filter is based on Kalman filtering with a constantly adapting neural network for state estimation. Applications are denoising sensor data or time series. Several bouncing ball simulations are realized to compare the estimation performance of the Extended Kalman Filter to the presented FNN-EKF.
KW - Adaptive extended Kalman filter
KW - Denoising
KW - Neural Network
KW - Robustness
KW - stability metric
UR - http://www.scopus.com/inward/record.url?scp=85106445656&partnerID=8YFLogxK
U2 - 10.1016/j.procir.2021.03.089
DO - 10.1016/j.procir.2021.03.089
M3 - Conference article
AN - SCOPUS:85106445656
VL - 99
SP - 656
EP - 661
JO - Procedia CIRP
JF - Procedia CIRP
SN - 2212-8271
T2 - 14th CIRP Conference on Intelligent Computation in Manufacturing Engineering, CIRP ICME 2020
Y2 - 15 July 2020 through 17 July 2020
ER -