Quantized identification of systems with networked packet losses

The quantized identification problem of systems with networked packet losses is studied in this paper. Specifically, the control signals are firstly quantized by the quantized sensor and then transmitted over a communication channel. There exist packet losses in this communication channel. The unknown parameter is identified by the quantized signals with packet losses. An empirical estimator of the interested parameter is obtained. The probability convergence, mean-square convergence of the estimator, and the upper bound and the lower bound of the estimation errors are analyzed in detail, respectively. In the end, a numerical example is presented to demonstrate the results obtained in this paper.

[1]  Feller William,et al.  An Introduction To Probability Theory And Its Applications , 1950 .

[2]  Qiang Zhang,et al.  Distributed Parameter Estimation Over Unreliable Networks With Markovian Switching Topologies , 2012, IEEE Transactions on Automatic Control.

[3]  Gerold Alsmeyer,et al.  Chebyshev's Inequality , 2011, International Encyclopedia of Statistical Science.

[4]  Graham C. Goodwin,et al.  System identification using quantized data , 2007, 2007 46th IEEE Conference on Decision and Control.

[5]  Minyue Fu,et al.  Identification of ARMA models using intermittent and quantized output observations , 2011, 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[6]  Fredrik Gustafsson,et al.  Generating dithering noise for maximum likelihood estimation from quantized data , 2013, Autom..

[7]  Torsten Söderström,et al.  System identification in a networked environment using second order statistical properties , 2013, Autom..

[8]  Tamer Başar,et al.  Quantization in ${H}^{ \infty}$ Parameter Identification , 2008, IEEE Transactions on Automatic Control.

[9]  B. Widrow,et al.  Statistical theory of quantization , 1996 .

[10]  Huazhen Fang,et al.  Kalman filter-based identification for systems with randomly missing measurements in a network environment , 2010, Int. J. Control.

[11]  Herman Chernoff.  Chernoff Bound , 2011, International Encyclopedia of Statistical Science.

[12]  Roberto Sanchis,et al.  Output prediction under scarce data operation: control applications , 1999, Autom..

[13]  Minrui Fei,et al.  A fast model identification method for networked control system , 2008, Appl. Math. Comput..

[14]  Jan M. Maciejowski,et al.  Optimal quantization of signals for system identification , 2003, ECC.

[15]  T. Sugie,et al.  System identification based on quantized I/O data corrupted with noises and its performance improvement , 2006, Proceedings of the 45th IEEE Conference on Decision and Control.

[16]  Jing Sun,et al.  Prediction of Oxygen Storage Capacity and Stored NOx by HEGO Sensors for Improved LNT Control Strategies , 2002 .

[17]  Ofer Zeitouni,et al.  Lectures on probability theory and statistics , 2004 .

[18]  Wei Xing Zheng,et al.  Identification of linear dynamic systems operating in a networked environment , 2009, Autom..

[19]  Torbjörn Wigren,et al.  Adaptive filtering using quantized output measurements , 1998, IEEE Trans. Signal Process..

[20]  G. Yin,et al.  System Identification with Quantized Observations , 2010 .

[21]  Fredrik Gustafsson,et al.  Statistical results for system identification based on quantized observations , 2009, Autom..