WMF self-tuning Kalman estimators for multisensor singular system

Abstract

For the multisensor linear stochastic singular system with unknown noise variances, the weighted measurement fusion (WMF) self-tuning Kalman estimation problem is solved in this paper. The consistent estimates of these unknown noise variances are obtained based on the correlation method. Applying the WMF method and the singular value decomposition (SVD) method yields the WMF reduced-order subsystems. Based on these consistent estimates of unknown noise variances and the new non-singular systems, the WMF self-tuning Kalman estimators of the state components and white noise deconvolution estimators are presented. Then the WMF self-tuning Kalman estimators of the original state are presented, and their convergence has been proved by dynamic error system analysis (DESA) method and dynamic variance error system analysis (DVESA) method. A simulation example of 3-sensors circuits systems verifies the effectiveness, the accuracy relationship and the convergence.

Keywords:

• Multisensor singular system
• weighted measurement fusion
• self-tuning Kalman estimators
• convergence analysis

Acknowledgments

The authors would like to thank the reviewers, associate editor, and editor for their helpful and constructive comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work is supported by Program for Young Creative Talents in Heilongjiang Provincial University (UNPYSCT-2016021), Science Fund for Outstanding Young Scholars of Heilongjiang University (JCL201304), Open Research Fund of Key Laboratory Electronic Engineering in Heilongjiang Provincial University (DZGC201605), 2017 Basic Scientific Research Project in Heilongjiang Province (KJCXZD201701). National Natural Science Foundation of China [61203121].

Notes on contributors

Yinfeng Dou

Yinfeng Dou was born in Harbin, China in 1982. He received the M.S. degree in department of Physical Electronics in 2009 and the Ph.D degree in department of Automation in 2017 from the Heilongjiang University. His research interest covers pattern recognition, information fusion and adaptive control.

Chenjian Ran

Chenjian Ran was born in Chongqing, China in 1981. She received her B.Sc. degree in Nanjing University of Science and Technology in 2005, and M.Sc. and Ph. D. degree in department of Automation, Heilongjiang University in 2008 and in 2011, respectively. Currently she is an associate professor at the Department of Automation, Heilongjiang University. Her research interests include multisensor information fusion, state estimation, descriptor system and self-tuning filtering.