The advantages and disadvantages of kalman filtering in repetitive control
Conference proceedings article
Authors/Editors
Strategic Research Themes
No matching items found.
Publication Details
Author list: Panomruttanarug B., Longman R.W.
Publication year: 2008
Volume number: 129 PART 2
Start page: 1433
End page: 1452
Number of pages: 20
ISBN: 9780877035435
ISSN: 0065-3438
eISSN: 0065-3438
Languages: English-Great Britain (EN-GB)
Abstract
Repetitive control (RC) can eliminate deterministic tracking errors of a control system in tracking a periodic command or errors as a result of a periodic disturbance. When there is substantial plant and measurement noise it is natural to consider employing a Kalman filter to improve the error signals used by the RC law, and the performance is analyzed here. Introducing a Kalman filter does substantially decrease the steady state error due to noise, and it also improves the robustness of good learning transients to model errors. Using a small learning gain or doing averaging, without a filter, can also decrease the error, but not to the same level as the Kalman filter. However, errors in the model used introduce deterministic errors in the response that can easily be larger than the decrease in error from random noise. Hence, one should carefully analyze the situation before deciding to use a Kalman filter. Furthermore, RC is capable of getting zero deterministic error when the model used in the design is inaccurate, and this capability is forfeited when introducing a Kalman filter. There are two classes of applications of RC, tracking periodic commands, and eliminating periodic disturbances. It is shown here that one cannot use a Kalman filter in this latter class of applications, which is perhaps the majority of RC applications.
Keywords
No matching items found.
Contact Information