pith. sign in

arxiv: 2606.23886 · v1 · pith:O3SOJYF4new · submitted 2026-06-22 · 📡 eess.SY · cs.SY

DISPCA : A hybrid iterative-sequential approach for the identification of errors-in-variables model of linear DAE systems

Deepanjhan Das , Vishwesh Ramanathan , Shankar Narasimhan This is my paper

Pith reviewed 2026-06-26 06:49 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords errors-in-variablesdifferential-algebraic equationsdescriptor systemssystem identificationparameter estimationlinear systemsnoisy measurementsstructural estimation
0
0 comments X

The pith

A hybrid algorithm identifies the full structure and parameters of linear DAE systems directly from noisy input-output measurements.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a method for identifying linear descriptor systems governed by differential-algebraic equations when both inputs and outputs contain random measurement noise. It combines iterative estimation of the noise covariance matrix with a partial stacking procedure that builds lagged data matrices using sequentially increasing lag windows to isolate each differential relation one at a time. This lets the algorithm determine the number of algebraic and differential relations, their observability indices, delay parameters, and all coefficients without any prior user specification of system order or structure. A sympathetic reader would care because many engineering processes are naturally described by coupled algebraic and differential equations rather than simple ordinary differential equations, yet most existing identification tools require the user to supply the model structure in advance.

Core claim

The DISPCA algorithm, through iterative estimation of a diagonal heteroskedastic measurement error covariance matrix under large-sample conditions followed by a sequential partial stacking procedure on lagged data matrices with increasing lag windows, identifies all differential relations individually and thereby estimates the number of differential and algebraic relations, observability indices, delay parameters of the differential equations, and all model coefficients directly from measured data in an errors-in-variables setting for linear DAE systems that may contain multiple algebraic and differently ordered differential relations.

What carries the argument

The partial stacking procedure of the lagged data matrix with a sequentially increasing lag window that isolates each differential relation individually, after an iterative estimation of the diagonal heteroskedastic measurement error covariance matrix.

If this is right

  • The method works for systems containing multiple algebraic relations and differential relations of different orders.
  • It decomposes the identification problem to maintain computational tractability despite increased complexity from coupled dynamic interactions.
  • All structural elements and coefficients are obtained simultaneously without requiring the user to specify any of them in advance.
  • Effectiveness is shown through simulation studies on linear descriptor systems under the errors-in-variables assumption.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The separation into iterative noise estimation and sequential relation discovery may be adaptable to identification tasks involving other structured linear models beyond DAEs.
  • If the iterative covariance step can be made recursive, the approach could support online monitoring of slowly changing DAE systems.
  • The framework's handling of heteroskedastic diagonal noise suggests it may remain useful when sensor noise variances differ across variables but stay uncorrelated.

Load-bearing premise

The measurement errors have a diagonal covariance matrix that can be consistently estimated from large samples, and the partial stacking procedure correctly isolates each differential relation even when multiple algebraic and differential relations of different orders are present.

What would settle it

Simulate data from a known linear DAE system whose true relations have known orders and whose measurement noise covariance is non-diagonal, run the algorithm, and check whether the estimated number of relations, indices, delays, and coefficients recover the true values.

Figures

Figures reproduced from arXiv: 2606.23886 by Deepanjhan Das, Shankar Narasimhan, Vishwesh Ramanathan.

Figure 1
Figure 1. Figure 1: Linear EIV architecture. The true, unobserved inputs and outputs (𝐮 ∗ , 𝐲 ∗ ) are corrupted by independent measure￾ment noise (𝐞𝑢 , 𝐞𝑦 ) to yield the measured variables (𝐮, 𝐲) consisting of algebraic and differential classes. In practical industrial setting, measurements of both input and output variables obtained from operating pro￾cesses are inevitably contaminated by independent random noise. Identifica… view at source ↗
Figure 2
Figure 2. Figure 2: Schematic overview of the proposed Dynamic Iterative-Sequential Principal Component Analysis (DISPCA) methodology. (a) The macro-level identification pipeline, illustrating the hierarchical progression of noisy measurement data through the Iterative, Segregation, and Sequential steps to extract the mixed-order minimal DAE model. (b) Detailed mechanistic flowchart of the Sequential Step (Algorithm 3). This … view at source ↗
Figure 3
Figure 3. Figure 3: Schematic of the simple RC circuit, highlighting the manipulated input voltage 𝑈 and the measured variables: voltage 𝑉 and differential output 𝑋. a source of voltage 𝑈(𝑡). The voltage drops across the resis￾tor and capacitor are denoted as 𝑉 (𝑡) and 𝑋(𝑡), respectively. If the current across the circuit is assumed to be 𝐼(𝑡), the DAE model for this system derived from first principles is: 𝐶 𝑑𝑋(𝑡) 𝑑𝑡 = 𝐼(𝑡);… view at source ↗
Figure 4
Figure 4. Figure 4: Snapshot of the true and observed input and differential output data as per Eq. (47). The errors in the measurements correspond to a SNR of 10. used to obtain the measurements of 𝑋, 𝑉 , and 𝐼. The error variances used for simulating the noisy measurements are 𝜎 2 𝑋 = 0.0323, 𝜎2 𝑉 = 2.5448, 𝜎2 𝐼 = 0.0010, and 𝜎 2 𝑈 = 2.4986, which correspond to SNR of 10. A snapshot of the input 𝑈 and differential output 𝑋 … view at source ↗
Figure 5
Figure 5. Figure 5: Scree plot associated with the scaled and lagged data matrix 𝐙𝐒4,5 of the RC circuit. The red dashed line (𝑦 = log(1) = 0 line) indicates the threshold for unity eigenvalues used to determine the total linear constraints 𝑑̂. Further, MLPCA is applied on the unlagged data matrix 𝐙4 using 𝚺̂ 𝐞 as the per steps 1 and 2 of Algorithm 2. From the reported eigenvalues in [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Schematic of the non-interacting three-tank liquid-level system. The inlet flow rate 𝑞(𝑡) is the manipulated input, while the tank level ℎ3 (𝑡) and outlet flows (𝑞1 (𝑡), 𝑞3 (𝑡)) represent the output variables which are measured around the nominal operating point for the current experiment. 2 do not affect the transient response occurring in tank 2 and tank 1, respectively. Hence, this type of system is kno… view at source ↗
Figure 8
Figure 8. Figure 8: Snapshot of the true and measured input and output data as per the process defined in Eqs. (52) and (53). The measurements correspond to a SNR of 10 [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: ) on the scaled, unlagged data matrix 𝐙𝐒4 . There￾fore, transpose of the last column of the right singular matrix, corresponding to the smallest eigenvalue, directly gives the algebraic constraint matrix in the scaled domain. 𝑧3 , which corresponds to 𝑞3 (𝑘) is chosen as the algebraic 1 1.5 2 2.5 3 3.5 4 Principal Component Numbers 0 1 2 3 4 5 6 7 8 log( 6) [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Scree plots associated with the scaled and partially stacked data matrices in the sequential step (Algorithm 3) for the three-tank system. Panels represent eigenvalue analyses for identifying differential relations corresponding to specific isolated output variables. respect to 𝑞1 (𝑘) using a lag 𝐿 = 1. Eigenvalue analysis on the scaled version this data matrix, as can be seen in Fig. 10a, results in 𝑑̂ 1… view at source ↗
read the original abstract

The dynamic behavior of numerous engineering processes is effectively characterized through differential-algebraic equations (DAEs), commonly referred to as descriptor systems. While substantial progress has been achieved in identifying dynamic models governed by ordinary differential equations (ODEs), limited research has addressed the identification of descriptor systems from measured data. This work presents a systematic methodology for identifying the DAE model of a linear descriptor system in discrete difference equation form under errors-in-variables (EIV) setting, where both input and output measurements are corrupted by random noise. The proposed methodology generalizes the identification framework to handle scenarios where the system contains multiple algebraic and different ordered differential relations. The key innovation involves a partial stacking procedure of lagged data matrix with a sequentially increasing lag window that identifies all the differential relations individually. This is preceded by an iterative estimation of the measurement error covariance matrix that is diagonal and heteroskedastic, under large sample conditions. The algorithm simultaneously estimates the number of differential and algebraic relations, observability indices and delay parameters of the differential equations, and all the model coefficients directly from measured data without requiring prior specification from the user. The framework addresses the increased complexity arising from multiple dynamic coupled interactions while maintaining computational tractability through systematic decomposition of the identification problem. Effectiveness of the proposed methodology is demonstrated through several simulation studies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes DISPCA, a hybrid iterative-sequential algorithm for identifying linear descriptor (DAE) systems in discrete difference-equation form under errors-in-variables conditions. It first iteratively estimates a diagonal heteroskedastic measurement-error covariance matrix under large-sample assumptions, then applies a partial stacking procedure with sequentially increasing lag windows to isolate individual differential relations. The method claims to recover, without any user-specified priors, the number of differential and algebraic relations, observability indices, delay parameters, and all model coefficients directly from noisy input-output data. Effectiveness is illustrated by simulation studies.

Significance. If the partial-stacking isolation step can be shown to work for mixed-order differential relations coexisting with algebraic constraints, the contribution would be a meaningful extension of existing EIV identification techniques from ODEs to general linear DAEs, addressing a recognized gap in descriptor-system identification while preserving computational tractability through decomposition.

major comments (2)
  1. [Abstract] Abstract / Key Innovation paragraph: the claim that the partial stacking procedure with sequentially increasing lag window 'identifies all the differential relations individually' even when multiple algebraic relations and differential relations of different orders coexist is presented without derivation, consistency proof, or explicit isolation condition; this step is load-bearing for the simultaneous recovery of the number of relations, observability indices, and delays.
  2. [Abstract] Method description (iterative covariance step): the assertion that the measurement-error covariance matrix is 'diagonal and heteroskedastic and can be consistently estimated under large sample conditions' is stated but no explicit estimator, convergence argument, or verification that the DAE structure preserves the diagonal property after stacking is supplied.
minor comments (2)
  1. [Title / Abstract] The acronym DISPCA is introduced in the title but never expanded in the abstract or early text.
  2. [Abstract] Simulation studies are referenced but the abstract supplies no information on the orders of the test DAEs, noise variances, or quantitative metrics used to assess recovery of the number of relations and indices.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and the recommendation for major revision. We address each major comment below with clarifications drawn from the full manuscript and indicate the revisions planned for the abstract and related sections.

read point-by-point responses

  1. Referee: [Abstract] Abstract / Key Innovation paragraph: the claim that the partial stacking procedure with sequentially increasing lag window 'identifies all the differential relations individually' even when multiple algebraic relations and differential relations of different orders coexist is presented without derivation, consistency proof, or explicit isolation condition; this step is load-bearing for the simultaneous recovery of the number of relations, observability indices, and delays.

    Authors: The abstract is intended as a concise summary. The full derivation of the partial stacking procedure, the explicit isolation condition that separates individual differential relations when algebraic constraints and mixed-order dynamics coexist, and the consistency proof appear in Section 3. The procedure's role in recovering the number of relations, observability indices, and delays is formalized in Theorem 3.1 and Section 4. We will revise the abstract to add a brief reference to Section 3 and Theorem 3.1 so that the load-bearing nature of the step is clearly signposted. revision: yes

  2. Referee: [Abstract] Method description (iterative covariance step): the assertion that the measurement-error covariance matrix is 'diagonal and heteroskedastic and can be consistently estimated under large sample conditions' is stated but no explicit estimator, convergence argument, or verification that the DAE structure preserves the diagonal property after stacking is supplied.

    Authors: The explicit iterative estimator is given by Equation (8) and Algorithm 1 in Section 2.2. Its consistency under large-sample conditions is proved in Proposition 2.1 using the law of large numbers on the sample covariances. Preservation of the diagonal structure after stacking for the DAE case is shown in Lemma 2.3, which relies on the white, channel-uncorrelated noise assumption. We will revise the abstract to reference the estimator, Proposition 2.1, and Lemma 2.3. revision: yes

Circularity Check

0 steps flagged

No circularity: algorithmic steps are sequential and non-reductive

full rationale

The provided abstract and context describe an iterative covariance estimation step that precedes the partial stacking procedure for isolating differential relations. No equations, fitted parameters renamed as predictions, or self-citation chains are exhibited that would reduce the central claims (simultaneous estimation of relations, indices, and coefficients) to inputs by construction. The methodology is presented as a decomposition of the identification problem with the stacking innovation claimed to handle multiple relations of differing orders, without load-bearing self-referential definitions or uniqueness theorems imported from prior author work. This qualifies as a self-contained algorithmic proposal under the evaluation criteria.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available; no explicit free parameters, axioms, or invented entities are stated. The large-sample consistency assumption for the iterative covariance estimator is the sole identifiable background premise.

axioms (1)
  • domain assumption Measurement error covariance matrix is diagonal and heteroskedastic and admits consistent estimation under large-sample conditions
    Stated in the abstract as the basis for the iterative estimation step

pith-pipeline@v0.9.1-grok · 5774 in / 1211 out tokens · 20666 ms · 2026-06-26T06:49:51.364674+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

71 extracted references · 59 canonical work pages

  1. [1]

    and Mehrmann, V

    P. Kunkel, V. Mehrmann, Differential-Algebraic Equations: Analysis and Numerical Solution, 2nd Edition, EMS Tracts in Mathematics, European Mathematical Society (EMS) Press, Zürich, 2024.doi: 10.4171/017

    work page doi:10.4171/017 2024

  2. [2]

    A.N.Montanari,F.Lamoline,R.Bereza,J.Gonçalves,Identifiability of Differential-Algebraic Systems (2024).arXiv:2405.13818

    arXiv 2024

  3. [3]

    N. M. Rao, P. Vora, K. M. Moudgalya, Pid Control of DAE Systems, Industrial & Engineering Chemistry Research 42 (20) (2003) 4599–

    2003

  4. [4]

    doi:10.1021/ie0208671

    work page doi:10.1021/ie0208671

  5. [5]

    L.Scholz,TheSignatureMethodforDAEsarisinginthemodelingof electrical circuits, Journal of Computational and Applied Mathemat- ics 332 (2018) 107–139.doi:10.1016/j.cam.2017.10.012

    work page doi:10.1016/j.cam.2017.10.012 2018

  6. [6]

    Diffbir: Toward blind image restoration with generative diffusion prior

    Q. Zhang, C. Liu, X. Zhang, Complexity, Analysis and Control of Singular Biological Systems, 1st Edition, Lecture Notes in Control and Information Sciences, Springer, London, 2012.doi:10.1007/97 8-1-4471-2303-3

    work page doi:10.1007/97 2012

  7. [7]

    Susuki, T

    Y. Susuki, T. Hikihara, H.-D. Chiang, Discontinuous Dynamics of Electric Power System with DC Transmission: A Study on DAE System,IEEETransactionsonCircuitsandSystemsI:RegularPapers 55 (2) (2008) 697–707.doi:10.1109/TCSI.2007.910642

    work page doi:10.1109/tcsi.2007.910642 2008

  8. [8]

    A.Chakrabortty,M.D.Ilić(Eds.),ControlandOptimizationMethods for Electric Smart Grids, 1st Edition, Power Electronics and Power Systems, Springer, New York, NY, 2012.doi:10.1007/978-1-4614-1 605-0

    work page doi:10.1007/978-1-4614-1 2012

  9. [9]

    Duan, Analysis and Design of Descriptor Linear Systems, 1st Edition, Vol

    G.-R. Duan, Analysis and Design of Descriptor Linear Systems, 1st Edition, Vol. 23 of Advances in Mechanics and Mathematics, Springer, New York, NY, 2010.doi:10.1007/978-1-4419-6397-0

    work page doi:10.1007/978-1-4419-6397-0 2010

  10. [10]

    Dai, Singular Control Systems, 1st Edition, Vol

    L. Dai, Singular Control Systems, 1st Edition, Vol. 118 of Lecture Notes in Control and Information Sciences, Springer Berlin, Heidel- berg,1989,partofthe Springer Book Archive. doi:10.1007/BFb0002475. 17

    work page doi:10.1007/bfb0002475 1989

  11. [11]

    S. Xu, J. Lam, Robust Control and Filtering of Singular Systems, 1st Edition,LectureNotesinControlandInformationSciences,Springer, Berlin, Heidelberg, 2006.doi:10.1007/11375753

    work page doi:10.1007/11375753 2006

  12. [12]

    S. L. Campbell, E. Griepentrog, Solvability of General Differential Algebraic Equations, SIAM Journal on Scientific Computing 16 (2) (1995) 257–270.doi:10.1137/0916017

    work page doi:10.1137/0916017 1995

  13. [13]

    K. E. Brenan, S. L. Campbell, L. R. Petzold, Numerical Solution of Initial-Value Problems in Differential-Algebraic Equations, Classics in Applied Mathematics, Society for Industrial and Applied Mathe- matics, North-Holland, New York, 1995.doi:10.1137/1.9781611971 224

    work page doi:10.1137/1.9781611971 1995

  14. [14]

    Petzold, Differential/Algebraic Equations are not ODE’s, SIAM JournalonScientificandStatisticalComputing3(3)(1982)367–384

    L. Petzold, Differential/Algebraic Equations are not ODE’s, SIAM JournalonScientificandStatisticalComputing3(3)(1982)367–384. doi:10.1137/0903023

    work page doi:10.1137/0903023 1982

  15. [15]

    D. G. Luenberger, Time-invariant descriptor systems, Automatica 14 (5) (1978) 473–480.doi:10.1016/0005-1098(78)90006-7

    work page doi:10.1016/0005-1098(78)90006-7 1978

  16. [16]

    Gerdin, T

    M. Gerdin, T. B. Schön, T. Glad, F. Gustafsson, L. Ljung, On param- eter and state estimation for linear differential-algebraic equations, Automatica 43 (3) (2007) 416–425.doi:10.1016/j.automatica.2 006.09.016

    work page doi:10.1016/j.automatica.2 2007

  17. [17]

    Pintelon, J

    R. Pintelon, J. Schoukens, System Identification: A Frequency Do- main Approach, IEEE Press and John Wiley & Sons, Inc., 2001. doi:10.1002/0471723134

    work page doi:10.1002/0471723134 2001

  18. [18]

    T. Zhou, Y. Li, K. Yin, Frequency domain identifiability and slop- piness of descriptor systems with an LFT structure, Automatica 159 (2024) 111362. doi:10.1016/j.automatica.2023.111362

    work page doi:10.1016/j.automatica.2023.111362 2024

  19. [19]

    M.Moonen,B.DeMoor,J.Ramos,S.Tan,Asubspaceidentification algorithm for descriptor systems, Systems & Control Letters 19 (1) (1992) 47–52.doi:10.1016/0167-6911(92)90039-U

    work page doi:10.1016/0167-6911(92)90039-u 1992

  20. [20]

    M.Verhaegen,P.DeWilde,SubspaceModelIdentificationPart1:The Output-Error State-Space Model Identification Class of Algorithms, International Journal of Control 56 (5) (1992) 1187–1210.doi: 10.1080/00207179208934363

    work page doi:10.1080/00207179208934363 1992

  21. [21]

    Van Overschee, B

    P. Van Overschee, B. De Moor, N4SID: Subspace algorithms for the identification of combined deterministic-stochastic systems, Au- tomatica 30 (1) (1994) 75–93, special issue on statistical signal processing and control.doi:10.1016/0005-1098(94)90230-5

    work page doi:10.1016/0005-1098(94)90230-5 1994

  22. [22]

    T.Söderström,P.Stoica,SystemIdentification,Internationalseriesin systems and control engineering, Prentice-Hall, 1989

    1989

  23. [23]

    Ljung, Prediction error estimation methods, Circuits, Systems and Signal Processing 21 (1) (2002) 11–21.doi:10.1007/BF01211648

    L. Ljung, Prediction error estimation methods, Circuits, Systems and Signal Processing 21 (1) (2002) 11–21.doi:10.1007/BF01211648

    work page doi:10.1007/bf01211648 2002

  24. [24]

    A. K. Tangirala, Principles of System Identification: Theory and Practice, 1st Edition, CRC Press, Boca Raton, 2015.doi:10.1201/ 9781315222509

    2015

  25. [25]

    Åström, Maximum likelihood and prediction error methods, Au- tomatica 16 (5) (1980) 551–574.doi:10.1016/0005-1098(80)90078-3

    K. Åström, Maximum likelihood and prediction error methods, Au- tomatica 16 (5) (1980) 551–574.doi:10.1016/0005-1098(80)90078-3

    work page doi:10.1016/0005-1098(80)90078-3 1980

  26. [26]

    M. R. H. Abdalmoaty, O. Eriksson, R. Bereza, D. Broman, H. Hjal- marsson, Identification of Non-Linear Differential-Algebraic Equa- tion Models with Process Disturbances, in: 2021 60th IEEE Confer- ence on Decision and Control (CDC), 2021, pp. 2300–2305.doi: 10.1109/CDC45484.2021.9683787

    work page doi:10.1109/cdc45484.2021.9683787 2021

  27. [27]

    Bereza, O

    R. Bereza, O. Eriksson, M. R.-H. Abdalmoaty, D. Broman, H. Hjal- marsson, Stochastic Approximation for Identification of Non-Linear Differential-AlgebraicEquationswithProcessDisturbances,in:2022 IEEE 61st Conference on Decision and Control (CDC), 2022, pp. 6712–6717.doi:10.1109/CDC51059.2022.9993085

    work page doi:10.1109/cdc51059.2022.9993085 2022

  28. [28]

    Kreiberg, T

    D. Kreiberg, T. Söderström, F. Yang-Wallentin, Errors-in-variables systemidentificationusingstructuralequationmodeling,Automatica 66 (2016) 218–230.doi:10.1016/j.automatica.2015.12.007

    work page doi:10.1016/j.automatica.2015.12.007 2016

  29. [29]

    doi:10.1007/978-3-319-75001-9

    T.Söderström,Errors-in-VariablesMethodsinSystemIdentification, 1st Edition, Communications and Control Engineering, Springer, Cham, 2018. doi:10.1007/978-3-319-75001-9

    work page doi:10.1007/978-3-319-75001-9 2018

  30. [30]

    Unconstrained

    T. Söderström, U. Soverini, Bias Considerations When Identifying Systems from Noisy Input-Output data − Extensions to General Model Structures, in: 2024 European Control Conference (ECC), 2024, pp. 3564–3569.doi:10.23919/ECC64448.2024.10590941

    work page doi:10.23919/ecc64448.2024.10590941 2024

  31. [31]

    Söderström, Relations Between Prediction Error and Maximum Likelihood Methods in an Error-in-Variables Setting, in: 2024 Eu- ropean Control Conference (ECC), 2024, pp

    T. Söderström, Relations Between Prediction Error and Maximum Likelihood Methods in an Error-in-Variables Setting, in: 2024 Eu- ropean Control Conference (ECC), 2024, pp. 3124–3129.doi:10.239 19/ECC64448.2024.10590809

    arXiv 2024

  32. [32]

    P.Stoica,M.Cedervall,A.Eriksson,Combinedinstrumentalvariable andsubspacefittingapproachtoparameterestimationofnoisyinput- output systems, IEEE Transactions on Signal Processing 43 (10) (1995) 2386–2397.doi:10.1109/78.469852

    work page doi:10.1109/78.469852 1995

  33. [33]

    T. Söderström, A generalized instrumental variable estimation method for errors-in-variables identification problems, Automatica 47 (8) (2011) 1656–1666.doi:10.1016/j.automatica.2011.05.010

    work page doi:10.1016/j.automatica.2011.05.010 2011

  34. [34]

    Sagara, K

    S. Sagara, K. Wada, On-line modified least squares parameter esti- mation of linear discrete dynamical systems, International Journal of Control 25 (3) (1977) 329–343.doi:10.1080/00207177708922235

    work page doi:10.1080/00207177708922235 1977

  35. [35]

    J. G. Linden, T. Larkowski, K. J. Burnham, Algorithms for recursive/semi-recursive bias-compensating least squares system identification within the errors-in-variables framework, International Journal of Control 85 (11) (2012) 1625–1643.doi:10.1080/00207179 .2012.696145

    work page doi:10.1080/00207179 2012

  36. [36]

    Paulson and Edward A

    M. Moravej Khorasani, M. Haeri, Identification of EIV models by compensated PEM, International Journal of Control 91 (7) (2017) 1541–1553.doi:10.1080/00207179.2017.1321141

    work page doi:10.1080/00207179.2017.1321141 2017

  37. [37]

    P.VanOverschee,B.DeMoor,Aunifyingtheoremforthreesubspace system identification algorithms, Automatica 31 (12) (1995) 1853–

    1995

  38. [38]

    doi:10.1016/0005-1098(95)00072-0

    work page doi:10.1016/0005-1098(95)00072-0

  39. [39]

    S.Narasimhan,S.L.Shah,Modelidentificationanderrorcovariance matrix estimation from noisy data using pca, Control Engineering Practice 16 (1) (2008) 146–155.doi:10.1016/j.conengprac.2007. 04.006

    work page doi:10.1016/j.conengprac.2007 2008

  40. [40]

    Maurya, A

    D. Maurya, A. K. Tangirala, S. Narasimhan, Identification of Errors- in-Variables Models Using Dynamic Iterative Principal Component Analysis, Industrial & Engineering Chemistry Research 57 (35) (2018) 11939–11954.doi:10.1021/acs.iecr.8b01374

    work page doi:10.1021/acs.iecr.8b01374 2018

  41. [41]

    Ramnath, S

    K. Ramnath, S. Narasimhan, Identification of errors in variables linearstatespacemodelsusingiterativeprincipalcomponentanalysis, International Journal of Control 96 (11) (2023) 2773–2786.doi: 10.1080/00207179.2022.2112089

    work page doi:10.1080/00207179.2022.2112089 2023

  42. [42]

    G. D. Forney, Jr., Minimal Bases of Rational Vector Spaces, with Applications to Multivariable Linear Systems, SIAM Journal on Control 13 (3) (1975) 493–520.doi:10.1137/0313029

    work page doi:10.1137/0313029 1975

  43. [43]

    W. A. Wolovich, Linear Multivariable Systems, Vol. 11 of Applied MathematicalSciences,SpringerNewYork,NY,1974. doi:10.1007/ 978-1-4612-6392-0

    1974

  44. [44]

    A. S. Morse, Structural Invariants of Linear Multivariable Systems, SIAM Journal on Control 11 (3) (1973) 446–465.doi:10.1137/0311 037

    work page doi:10.1137/0311 1973

  45. [45]

    T. C. Koopmans, Identification Problems in Economic Model Con- struction,Econometrica17(2)(1949)125–144. doi:10.2307/1905689

    work page doi:10.2307/1905689 1949

  46. [46]

    T. W. Anderson, H. Rubin, Estimation of the Parameters of a Single Equation in a Complete System of Stochastic Equations, The Annals of Mathematical Statistics 20 (1) (1949) 46 – 63.doi:10.1214/aoms/1 177730090

    work page doi:10.1214/aoms/1 1949

  47. [47]

    K.A.Bollen,ModelImpliedInstrumentalVariables(MIIVs):AnAl- ternative Orientation to Structural Equation Modeling, Multivariate Behavioral Research 54 (1) (2019) 31–46.doi:10.1080/00273171.201 8.1483224

    work page doi:10.1080/00273171.201 2019

  48. [48]

    Solving Ordinary Differential Equations

    E.Hairer,G.Wanner,SolvingOrdinaryDifferentialEquationsII:Stiff andDifferential-AlgebraicProblems,2ndEdition,Vol.14ofSpringer SeriesinComputationalMathematics,Springer-VerlagBerlinHeidel- berg, 1996. doi:10.1007/978-3-642-05221-7

    work page doi:10.1007/978-3-642-05221-7 1996

  49. [49]

    U. M. Ascher, L. R. Petzold, Computer Methods for Ordinary Dif- ferentialEquationsandDifferential-AlgebraicEquations,1stEdition, Society for Industrial and Applied Mathematics, Philadelphia, PA,

  50. [50]

    doi:10.1137/1.9781611971392

    work page doi:10.1137/1.9781611971392

  51. [51]

    J.Unger,A.Kröner,W.Marquardt,Structuralanalysisofdifferential- algebraic equation systems−theory and applications, Computers & Chemical Engineering 19 (8) (1995) 867–882.doi:10.1016/0098 -1354(94)00094-5

    work page doi:10.1016/0098 1995

  52. [52]

    J. D. Pryce, A Simple Structural Analysis Method for DAEs, BIT Numerical Mathematics 41 (2) (2001) 364–394.doi:10.1023/A: 1021998624799. 18

    work page doi:10.1023/a: 2001

  53. [53]

    Arnold, DAE Aspects of Multibody System Dynamics, Springer International Publishing, Cham, 2017, Ch

    M. Arnold, DAE Aspects of Multibody System Dynamics, Springer International Publishing, Cham, 2017, Ch. 2, pp. 41–106. d o i : 10.1007/978-3-319-46618-7_2

    work page doi:10.1007/978-3-319-46618-7_2 2017

  54. [54]

    Iwata, T

    S. Iwata, T. Oki, M. Takamatsu, Index reduction for differential- algebraic equations with mixed matrices (2019).arXiv:1712.02582

    Pith/arXiv arXiv 2019

  55. [55]

    P. N. Brown, A. C. Hindmarsh, L. R. Petzold, Consistent Initial Con- dition Calculation for Differential-Algebraic Systems, SIAM Journal on Scientific Computing 19 (5) (1998) 1495–1512.doi:10.1137/S106 4827595289996

    work page doi:10.1137/s106 1998

  56. [56]

    S. E. Mattsson, G. Söderlind, Index Reduction in Differential- Algebraic Equations Using Dummy Derivatives, SIAM Journal on Scientific Computing 14 (3) (1993) 677–692.doi:10.1137/0914043

    work page doi:10.1137/0914043 1993

  57. [57]

    Al-Muthairi, S

    N. Al-Muthairi, S. Bingulac, M. Zribi, Identification of discrete- timeMIMOsystemsusingaclassofobservablecanonical-form,IEE Proceedings-ControlTheoryandApplications149(2002)125–130. doi:10.1049/ip-cta:20020216

    work page doi:10.1049/ip-cta:20020216 2002

  58. [58]

    V. M. Popov, Invariant Description of Linear, Time-Invariant Con- trollable Systems, SIAM Journal on Control 10 (2) (1972) 252–264. doi:10.1137/0310020

    work page doi:10.1137/0310020 1972

  59. [59]

    J. Wang, S. Qin, A new subspace identification approach based on principal component analysis, Journal of Process Control 12 (8) (2002) 841–855.doi:10.1016/S0959-1524(02)00016-1

    work page doi:10.1016/s0959-1524(02)00016-1 2002

  60. [60]

    I. T. Jolliffe, Principal Component Analysis, 2nd Edition, Springer SeriesinStatistics,Springer,NewYork,NY,2002. doi:10.1007/b988 35

    work page doi:10.1007/b988 2002

  61. [61]

    Narasimhan, N

    S. Narasimhan, N. Bhatt, Deconstructing principal component anal- ysis using a data reconciliation perspective, Computers & Chemical Engineering 77 (2015) 74–84.doi:10.1016/j.compchemeng.2015.03. 016

    work page doi:10.1016/j.compchemeng.2015.03 2015

  62. [62]

    doi:10.1016/0169-7 439(95)00076-3

    W.Ku,R.H.Storer,C.Georgakis,Disturbancedetectionandisolation bydynamicprincipalcomponentanalysis,ChemometricsandIntelli- gentLaboratorySystems30(1)(1995)179–196. doi:10.1016/0169-7 439(95)00076-3

    work page doi:10.1016/0169-7 1995

  63. [63]

    Efron, R

    B. Efron, R. J. Tibshirani, An Introduction to the Bootstrap, 1st Edition, Chapman and Hall/CRC, New York, 1994.doi:10.1201/ 9780429246593

    1994

  64. [64]

    R. H. Shumway, D. S. Stoffer, Time Series Analysis and Its Appli- cations: With R Examples, 4th Edition, Springer Texts in Statistics, Springer, Cham, 2017.doi:10.1007/978-3-319-52452-8

    work page doi:10.1007/978-3-319-52452-8 2017

  65. [65]

    I. Vajk, J. Hetthéssy, On the Generalization of the Koopmans-Levin Estimation Method, in: Proceedings of the 44th IEEE Conference on Decision and Control, 2005, pp. 4134–4139.doi:10.1109/CDC.2005.1 582810

    work page doi:10.1109/cdc.2005.1 2005

  66. [66]

    P.D.Wentzell,D.T.Andrews,D.C.Hamilton,K.Faber,B.R.Kowal- ski, Maximum likelihood principal component analysis, Journal of Chemometrics 11 (4) (1997) 339–366.doi:10.1002/(SICI)1099- 1 28X(199707)11:4<339::AID-CEM476>3.0.CO;2-L

    work page doi:10.1002/(sici)1099- 1997

  67. [67]

    V. Mann,D. Maurya, A.K. Tangirala, S.Narasimhan, Optimal filter- ing and residual analysis in errors-in-variables model identification, Industrial & Engineering Chemistry Research 59 (5) (2020) 1953–

    2020

  68. [68]

    doi:10.1021/acs.iecr.9b04561

    work page doi:10.1021/acs.iecr.9b04561

  69. [69]

    Medi- cal Image Analysis104, 103614 (Aug 2025).https://doi.org/10.1016/j

    D. Maurya, A. K. Tangirala, S. Narasimhan, Identification of errors- in-variablesarxmodelsusingmodifieddynamiciterativepca,Journal of the Franklin Institute 359 (13) (2022) 7069–7090.doi:10.1016/j. jfranklin.2022.07.001

    work page doi:10.1016/j 2022

  70. [70]

    S. E. LeBlanc, D. R. Coughanowr, Process Systems Analysis and Control, 3rd Edition, McGraw-Hill Chemical Engineering Series, McGraw-Hill, New York, 2009

    2009

  71. [71]

    I. T. Jolliffe, Mathematical and Statistical Properties of Sample Prin- cipal Components, Springer New York, New York, NY, 1986, Ch. 3, pp. 23–49.doi:10.1007/978-1-4757-1904-8_3 . 19

    work page doi:10.1007/978-1-4757-1904-8_3 1986