In this paper we give an overview of errors-in-variables methods in system identification. Background and motivation are given. Simple examples illustrate why the identification problem can be difficult. Under general weak assumptions, the systems are not uniquely identifiable, but can be parameterized using one degree of freedom. Examples where identifiability is achieved under additional assumptions are also provided. A number of approaches for parameter estimation of errors-in-variables models are presented. The underlying assumptions and principles for each approach are highlighted.
AgüeroJCGoodwinGC (2008) Identifiability of errors in variables dynamic systems. Automatica44: 371–382.
2.
AgüeroJCGoodwinGCSöderströmTYuzJI (2009) Sampled data errors in variables systems. In SYSID 2009, IFAC 15th Symposium on System Identification, Saint-Malo, France, 6–8 July 2009.
3.
AndersonBDO (1985) Identification of scalar errors-in-variables models with dynamics. Automatica21: 709–716.
4.
BeghelliSGuidorziRPSoveriniU (1990) The Frisch scheme in dynamic system identification. Automatica26: 171–176.
5.
ChenH-F (2007) Recursive identification for multivariate errors-in-variables systems. Automatica43: 1234–1242.
6.
ChengCLVan NessJW (1999) Kendall’s library of statistics: Statistical Regression With Measurement Error. Volume 6, Edward Arnold.
7.
CunhaFHeckmanJJSchennachSM (2010) Estimating the technology of cognitive and noncognitive skill formation. Econometrica78: 883–931.
8.
DeistlerM (1986) Linear dynamic errors-in-varibles models. Journal of Applied Probability23A: 23–39.
9.
DelaigleAHallP (2008) Using SIMEX for smoothing-parameter choice in errors-in-variables problems. Journal of the American Statistical Association103: 280–287.
10.
DiversiRGuidorziRSoveriniU (2003a) A new criterion in EIV identification and filtering applications. In Proceedings 13th IFAC Symposium on System Identification, Rotterdam, The Netherlands, 27–29 August 2003, pp. 1993–1998.
11.
DiversiRGuidorziRSoveriniU (2003b) Algorithms for optimal errors-in-variables filtering. Systems and Control Letters48: 1–13.
12.
DiversiRGuidorziRSoveriniU (2005) Frisch scheme-based algorithms for EIV identification. In Proceedings 12th IEEE Mediterranean Conference on Control and Automation, Kusadasi, Turkey, June 2004.
13.
DiversiRGuidorziRSoveriniU (2005) Kalman filtering in extended noise environments. IEEE Transactions on Automatic Control50: 1396–1402.
14.
DiversiRSoveriniUGuidorziR (2006) Yule–Walker equations in the Frisch scheme solution of errors-in-variables identification problems. In Proceedings 17th International Symposium on Mathematical Theory of Networks and Systems (MTNS), Kyoto, Japan, July 2006, pp. 391–395.
15.
DiversiRGuidorziRSoveriniU (2007) Maximum likelihood identification of noisy input-output models. Automatica43: 464–472.
16.
EkmanM (2005) Identification of linear systems with errors in variables using separable nonlinear least squares. In 16th IFAC World Congress. Prague, Czech Republic, 4–8 July 2005.
17.
FullerWA (1987) Measurement Error Models. New York: Wiley, 1987.
18.
GolubGHPereyraV (2003) Separable nonlinear least squares: The variable projection method and its applications. Inverse Problems19: R1–R26.
19.
GolubGHPereyraV (1973) The differentiation of pseudoinverses and nonlinear least squares problems whose variables separate. SIAM Journal of Numerical Analysis10: 413–432.
20.
GuidorziRDiversiRSoveriniU (2008) The Frisch scheme in algebraic and dynamic identification problems. Kybernetica44: 585–616.
21.
HongMSöderströmT (2009) Relations between bias-eliminating least squares, the Frisch scheme and extended compensated least squares methods for identifying errors-in-variables systems. Automatica45: 277–282.
JunB-EBernsteinDS (2007) Extended least-correlation estimates for errors-in-variables non-linear models. International Journal of Control80: 256–267.
24.
KlepperSLeamerEE (1984) Consistent sets of estimates for regressions with errors in all variables. Econometrics52: 163–184.
25.
KrishnamurthyVLogothetisA (1996) Iterative and recursive estimators for hidden Markov errors-in-variables models. IEEE Transactions on Signal Processing44: 629–639.
26.
KukushAMarkovskyIVan HuffelS (2004) Consistent estimation in an implicit quadratic measurement model. Computational Statistics and Data Analysis47: 123–147.
27.
KukushAMarkovskyIVan HuffelS (2007) Estimation in a linear multivariate measurement error model with a change point in the data. Computational Statistics and Data Analysis52: 1167–1182.
28.
LarkowskiTLindenJGBurnhamKJ (2009a) Errors-in-variables identification of dynamic bilinear systems in the case of coloured output noise. In Proceedings European Control Conference, Budapest, Hungary, August 2009.
29.
LarkowskiTLindenJGVinsonneauBBurnhamKJ (2009b) Frisch scheme identification for dynamic diagonal bilinear models. International Journal of Control82: 1591–1604.
30.
LindenJGBurnhamKJ (2009) A recursive Frisch scheme algorithm for coloured output noise. In ICINCO, Proceedings 5th International Conference on Informatics in Control, Automation and Robotics, Funchal, Portugal, 2009.
31.
LindenJGVinsonneauBBurnhamKJ (2008) Gradient based approaches for recursive Frisch scheme identification. In IFAC 17th World Congress. Seoul, Korea, 6–1 July 2008.
32.
LindenJGLarkowskiTBurnhamKJ (2009) An improved recursive Frisch scheme identification algorithm. In ICSENG 2008, International Conference on Systems Engineering, Las Vegas, NV, 2009.
33.
LjungL (1999) System Identification—Theory for the User, 2nd ed.Upper Saddle River, NJ: Prentice Hall, 1999.
34.
MahataK (2007) An improved bias-compensation approach for errors-in-variables model identification. Automatica43: 1339–1354.
35.
MahataKGarnierH (2005) Direct identification of continuous-time errors-in-variables models. In 14th IFAC World Congress. Prague, Czech Republic, 3–8 July 2005.
36.
MarkovskyI (2008) Structured low-rank approximation and its application. Automatica44: 891–909.
37.
MarkovskyIDe MoorB (2005) Linear dynamic filtering with noisy input and output. Automatica41: 167–171.
38.
MarkovskyIVan HuffelS (2007) Overview of total least-squares methods. Signal Processing87: 2283–2302.
39.
MarkovskyIWillemsJCVan HuffelSDe MoorBPintelonR (2005) Application of structured total least squares for system identification and model reduction. IEEE Transactions on Automatic Control50: 1490–1500.
40.
MateiBMeerP (2006) Estimation of nonlinear errors-in-variables models for computer vision applications. IEEE Transactions on Pattern Analysis and Machine Intelligence28: 1537–1552.
41.
MilaneseMVicinoA (1991) Optimal estimation theory for dynamic systems with set membership uncertainty: an overview. Automatica27: 997–1009.
42.
MossbergMSöderströmT (2009) Continuous-time errors-in-variables system identification through covariance matching without input signal modeling. In 2009 American Control Conference, St Louis, MO, 10–12 June 2009.
43.
MossbergMSöderströmT (2011) Covariance matching for continuous-time errors-in-variables problems. IEEE Transactions on Automatic Control2011; 56, 1478–1483.
44.
PintelonRSchoukensJ (2007) Frequency domain maximum likelihood estimation of linear dynamic errors-in-variables models. Automatica43: 621–630.
45.
ReiersølO (1950) Identifiability of a linear relation between variables which are subject to error. Econometrica18: 375–389.
46.
SchennachS (2004) Estimation of nonlinear models with measurement errors. Econometrica72: 33–75.
ScherrerWDeistlerM (1998) A structure theory for linear dynamic errors-in-variables models. SIAM Journal on Control and Optimization36: 2148–2175.
49.
SchoukensJPintelonRVandersteenGGuillaumeP (1997) Frequency domain system identification using non-parametric noise models estimated from a small number of data sets. Automatica33: 1073–1086.
50.
SöderströmT (2002) Discrete-time Stochastic Systems—Estimation and Control, 2nd ed.London: Springer-Verlag.
51.
SöderströmT (2006) On computing the Cramér–Rao bound and covariance matrices for PEM estimates in linear state space models. In Proceedings 14th IFAC Symposium on System Identification, Newcastle, Australia, 29–31 March 2006.
52.
SöderströmT (2007) A Frisch scheme for correlated output noise errors-in-variables identification. In European Control Conference, ECC’07, Kos, Greece, 2–5 July 2007.
53.
SöderströmT (2009) Sampling approximations for continuous-time identification. In SYSID 2009, IFAC 15th Symposium on System Identification. Saint-Malo, France, 6–8 July 2009.
54.
SöderströmT (2010) A generalized instrumental variable estimator for multivariable errors-in-variables problems. International Journal of Control, submitted.
55.
SöderströmT (2011) A generalized instrumental variable method for errors-in-variables. Automatica47: 1656–1666.
56.
SöderströmTHongM (2005) Identification of dynamic errors-in-variables systems with periodic data. In Proceedings 16th IFAC World Congress. Prague, Czech Republic, 4–8 July 2005.
57.
SöderströmTMossbergM (2011) Accuracy analysis of a covariance matching approach for identifying errors-in-variables systems. Automatica47: 272–282.
58.
SöderströmTStoicaP (1989) System Identification. Hemel Hempstead: Prentice Hall International, 1989.
59.
SöderströmTHongMZhengWX (2005) Convergence properties of bias-eliminating algorithms for errors-in-variables identification. International Journal of Adaptive Control and Signal Processing19: 703–722.
60.
SöderströmTLarssonEKMahataKMossbergM (2006) Using continuous-time modeling for errors-in-variables identification. In Proceedings 14th IFAC Symposium on System Identification. Newcastle, Australia, 29–31 March 2006.
61.
SöderströmTMossbergMHongM (2009) A covariance matching approach for identifying errors-in-variables systems. Automatica45: 2018–2031.
62.
SöderströmTHongMSchoukensJPintelonR (2010) Accuracy analysis of time-domain maximum likelihood method and sample maximum likelihood method for errors-in-variables and output error identification. Automatica46: 721–727.
63.
StoicaPLiJ (2005) On nonexistence of the maximum likelihood estimate in blind multichannel identification. IEEE Signal Processing Letters22: 99–101.
64.
ThilSGarnierHGilsonM (2008a) Third-order cumulants based methods for continuous-time errors-in-variables model identification. Automatica44: 647–658.
65.
ThilSHongMSöderströmTGilsonMGarnierH (2008b) Statistical analysis of a third-order cumulants based algorithm for discrete errors-in-variables identification. In IFAC 17th World Congress. Seoul, Korea, 6–11 July 2008.
66.
TugnaitJK (1992) Stochastic system identification with noisy input using cumulant statistics. IEEE Transactions on Automatic Control37: 476–485.
67.
VajkIHetthéssyJ (2003) Identification of nonlinear errors-in-variables models. Automatica39: 2099–2107.
68.
van HeusdenKKarimiASöderströmT (2011) On identification methods for direct data-driven controller tuning. International Journal of Adaptive Control and Signal Processing25, 448–465.
69.
Van HuffelS (ed.) (1997) Recent Advances in Total Least Squares Techniques and Errors-in-Variables Modelling. Philadelphia, PA: SIAM.
70.
Van HuffelSLemmerlingPh (eds) (2002) Total Least Squares and Errors-in-Variables Modelling. Analysis, Algorithms and Applications. Dordrecht: Kluwer.
71.
Van HuffelSVandewalleJ (1991) The Total Least Squares Problem: Computation Aspects and Analysis. Philadelphia, PA: SIAM.
72.
Van SchuppenJH (1989) Stochastic realization problems. In: NijmeijerNSchumacherJM (eds) Three Decades of Mathematical System Theory. Berlin: Springer-Verlag.
73.
WaldA (1940) The fitting of straight lines if both variables are subject to error. The Annals of Mathematical Statistics, Series B11: 284–300.
74.
ZhengWX (1998) Transfer function estimation form noisy input and output data. International Journal of Adaptive Control and Signal Processing12: 365–380.
75.
ZhengWXFengCB (1989) Unbiased parameter estimation of linear systems in presence of input and output noise. International Journal of Adaptive Control and Signal Processing3: 231–251.
