State estimation in the case of loss of observations

Loss of information (observations) is a common problem in control and communication systems. Kalman filter is a versatile tool for state estimation, but would it still produce accurate estimation in such a case? In this paper we investigate this situation and propose several approaches to compensate the loss of information in employing Kalman filter to estimate the state of a system. Minimum error variance for these approaches is derived from the basic structure of the classical Kalman filer. Necessary discussion for all approaches regarding their applications and drawbacks are stated. Optimal Kalman gain matrix for these approaches is calculated. Selection criterion for the approaches are also presented. Details of the theoretical properties such as convergence and stabilization of Riccati equation are not, however, included due to limited length of a conference paper. Numerical example is included to illustrate the effectiveness of these approaches.

[1]  A. Goldsmith,et al.  Kalman filtering with partial observation losses , 2004, 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601).

[2]  L. Tanachaikhan,et al.  Declined angle analysis of shaft using magnetic field measurement , 2009, 2009 ICCAS-SICE.

[3]  M. Gevers,et al.  Fake algebraic Riccati techniques and stability , 1988 .

[4]  Ian Postlethwaite,et al.  A robust fault-tolerant tracking scheme , 2008 .

[5]  Da-Wei Gu,et al.  Properties of a robust Kalman filter , 2009, ICONS.

[6]  Arthur Gelb,et al.  Applied Optimal Estimation , 1974 .

[7]  P. P. Kanjilal,et al.  Adaptive Prediction and Predictive Control , 1995 .

[8]  A.H. Haddad,et al.  Applied optimal estimation , 1976, Proceedings of the IEEE.

[9]  Petros G. Voulgaris,et al.  On optimal ℓ∞ to ℓ∞ filtering , 1995, Autom..

[10]  Bruno Sinopoli,et al.  Kalman filtering with intermittent observations , 2004, IEEE Transactions on Automatic Control.