Share Email Print
cover

Proceedings Paper • new

A multiple model adaptive SVSF-KF estimation strategy
Author(s): Jacob M. Goodman; Stephen A. Wilkerson; Charles Eggleton; S. Andrew Gadsden
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

State estimation strategies play a critical role in obtaining accurate information about the state of dynamic systems as they develop. Such information can be important on its own and critical for precise and predictable control of such systems. The Kalman filter (KF) is a classic algorithm and among the most powerful tools in state estimation. The Kalman filter however can be sensitive to modeling uncertainty and sudden changes in system dynamics. The Smooth Variable Structure Filter (SVSF) is a relatively new estimation strategy that operates on variable structure concepts. In general, the SVSF has the advantage that is can be quite robust to modeling uncertainty and sudden fault conditions. Recent advancements to the SVSF, such as the addition of a covariance formulation, and the derivation of a time varying smoothing boundary layer (VBL), have allowed for combined SVSF – KF strategies. In a typical SVSF-KF approach, the VBL is used to detect the presence of a system fault, and switch from the more optimal KF gain to the more robust SVSF gain. While this approach has been proven effective in several cases, there are circumstances where the VBL will fail to indicate the presence of an ongoing fault. A new form of the SVSF-KF is proposed, based on the framework of the Multiple Model Adaptive Estimator.

Paper Details

Date Published: 7 May 2019
PDF: 12 pages
Proc. SPIE 11018, Signal Processing, Sensor/Information Fusion, and Target Recognition XXVIII, 110181K (7 May 2019); doi: 10.1117/12.2520018
Show Author Affiliations
Jacob M. Goodman, Univ. of Maryland, Baltimore County (United States)
Stephen A. Wilkerson, York College of Pennsylvania (United States)
Charles Eggleton, Univ. of Maryland, Baltimore County (United States)
S. Andrew Gadsden, Univ. of Guelph (Canada)


Published in SPIE Proceedings Vol. 11018:
Signal Processing, Sensor/Information Fusion, and Target Recognition XXVIII
Ivan Kadar; Erik P. Blasch; Lynne L. Grewe, Editor(s)

© SPIE. Terms of Use
Back to Top