Structural identifiability

In the area of system identification, a dynamical system is structurally identifiable if it is possible to infer its unknown parameters by measuring its output over time. This problem arises in many branch of applied mathematics, since dynamical systems (such as the ones described by ordinary differential equations) are commonly utilized to model physical processes and these models contain unknown parameters that are typically estimated using experimental data. ^{[1] [2] [3]}

However, in certain cases, the model structure may not permit a unique solution for this estimation problem, even when the data is continuous and free from noise. To avoid potential issues, it is recommended to verify the uniqueness of the solution in advance, prior to conducting any actual experiments. ^[4] The lack of structural identifiability implies that there are multiple solutions for the problem of system identification, and the impossibility of distinguishing between these solutions suggests that the system has poor forecasting power as a model. ^[5] On the other hand, control systems have been proposed with the goal of rendering the closed-loop system unidentifiable, decreasing its susceptibility to covert attacks targeting cyber-physical systems. ^[6]

Examples

Linear time-invariant system

Source ^[2]

Consider a linear time-invariant system with the following state-space representation:

${\displaystyle {\begin{aligned}{\dot {x}}_{1}(t)&=-\theta _{1}x_{1},\\{\dot {x}}_{2}(t)&=\theta _{1}x_{1},\\y(t)&=\theta _{2}x_{2},\end{aligned}}}$

and with initial conditions given by ${\displaystyle x_{1}(0)=\theta _{3}}$ and ${\displaystyle x_{2}(0)=0}$. The solution of the output ${\displaystyle y}$ is

${\displaystyle y(t)=\theta _{2}\theta _{3}e^{-\theta _{1}t}\left(e^{\theta _{1}t}-1\right),}$

which implies that the parameters ${\displaystyle \theta _{2}}$ and ${\displaystyle \theta _{3}}$ are not structurally identifiable. For instance, the parameters ${\displaystyle \theta _{1}=1,\theta _{2}=1,\theta _{3}=1}$ generates the same output as the parameters ${\displaystyle \theta _{1}=1,\theta _{2}=2,\theta _{3}=0.5}$.

Non-linear system

Source ^[7]

A model of a possible glucose homeostasis mechanism is given by the differential equations ^[8]

${\displaystyle {\begin{aligned}&{\dot {G}}=u(0)+u-(c+s_{\mathrm {i} }\,I)G,\\&{\dot {\beta }}=\beta \left({\frac {1.4583\cdot 10^{-5}}{1+\left({\frac {8.4}{G}}\right)^{1.7}}}-{\frac {1.7361\cdot 10^{-5}}{1+\left({\frac {G}{8.4}}\right)^{8.5}}}\right),\\&{\dot {I}}=p\,\beta \,{\frac {G^{2}}{\alpha ^{2}+G^{2}}}-\gamma \,I,\end{aligned}}}$

where (c, s_i, p, α, γ) are parameters of the system, and the states are the plasma glucose concentration G, the plasma insulin concentration I, and the beta-cell functional mass β. It is possible to show that the parameters p and s_i are not structurally identifiable: any numerical choice of parameters p and s_i that have the same product ps_i are indistinguishable. ^[7]

Practical identifiability

Structural identifiability is assessed by analyzing the dynamical equations of the system, and does not take into account possible noises in the measurement of the output. In contrast, practical non-identifiability also takes noises into account. ^{[1] [9]}

Other related notions

The notion of structurally identifiable is closely related to observability, which refers to the capacity of inferring the state of the system by measuring the trajectories of the system output. It is also closely related to data informativity, which refers to the proper selection of inputs that enables the inference of the unknown parameters. ^{[10] [11]}

The (lack of) structural identifiability is also important in the context of dynamical compensation of physiological control systems. These systems should ensure a precise dynamical response despite variations in certain parameters. ^{[12] [13]} In other words, while in the field of systems identification, unidentifiability is considered a negative property, in the context of dynamical compensation, unidentifiability becomes a desirable property. ^[13]

Identifiability also appears in the context of inverse optimal control. Here, one assumes that the data comes from a solution of an optimal control problem with unknown parameters in the objective function. Here, identifiability refers to the possibility of infering the parameters present in the objective function by using the measured data. ^[14]

Software

There exist many software that can be used for analyzing the identifiability of a system, including non-linear systems: ^[15]

• PottersWheel: MATLAB toolbox that uses profile likelihood for structural and practical identifiability analysis.
• STRIKE-GOLDD: MATLAB toolbox for structural identifiability analysis. ^[16]
• StructuralIdentifiability.jl: Julia library for assessing structural parameter identifiability. ^[17]
• LikelihoodProfiler.jl: Julia library for practical identifiability analysis. ^[18]

See also

• System identification
• Observability
• Model order reduction
• Adaptive control

References

1. Miao, Hongyu; Xia, Xiaohua; Perelson, Alan S.; Wu, Hulin (2011). "On Identifiability of Nonlinear ODE Models and Applications in Viral Dynamics". SIAM Review. 53 (1): 3–39. doi:10.1137/090757009. ISSN   0036-1445. PMC   3140286 . PMID   21785515. (Erratum:  doi:10.1137/23M1568958)
2. Raue, A.; Karlsson, J.; Saccomani, M. P.; Jirstrand, M.; Timmer, J. (2014-05-15). "Comparison of approaches for parameter identifiability analysis of biological systems". Bioinformatics. 30 (10): 1440–1448. doi: 10.1093/bioinformatics/btu006 . ISSN   1367-4803. PMID   24463185. S2CID   10052322.
3. Wensing, Patrick M.; Niemeyer, Günter; Slotine, Jean-Jacques E. (2024). "A geometric characterization of observability in inertial parameter identification". The International Journal of Robotics Research. 43 (14): 2274–2302. arXiv: 1711.03896 . doi:10.1177/02783649241258215. ISSN   0278-3649.
4. Villaverde, Alejandro F; Pathirana, Dilan; Fröhlich, Fabian; Hasenauer, Jan; Banga, Julio R (2022-01-17). "A protocol for dynamic model calibration". Briefings in Bioinformatics. 23 (1) bbab387. doi:10.1093/bib/bbab387. ISSN   1467-5463. PMC   8769694 . PMID   34619769.
5. Fiacchini, Mirko; Alamir, Mazen (2021). "The Ockham's razor applied to COVID-19 model fitting French data". Annual Reviews in Control. 51: 500–510. doi:10.1016/j.arcontrol.2021.01.002. PMC   7846253 . PMID   33551664.
6. Mao, Xiangyu; He, Jianping (2023). "Unidentifiability of System Dynamics: Conditions and Controller Design". IEEE Transactions on Automatic Control. 70 (2): 1380. arXiv: 2308.15493 . Bibcode:2025ITAC...70.1380M. doi:10.1109/TAC.2024.3469392.
7. Villaverde, Alejandro F. (2019-01-01). "Observability and Structural Identifiability of Nonlinear Biological Systems". Complexity. 2019 8497093: 1–12. arXiv: 1812.04525 . doi: 10.1155/2019/8497093 . ISSN   1076-2787.
8. Karin, Omer; Swisa, Avital; Glaser, Benjamin; Dor, Yuval; Alon, Uri (2016). "Dynamical compensation in physiological circuits". Molecular Systems Biology. 12 (11): 886. doi:10.15252/msb.20167216. ISSN   1744-4292. PMC   5147051 . PMID   27875241.
9. Raue, A.; Kreutz, C.; Maiwald, T.; Bachmann, J.; Schilling, M.; Klingmüller, U.; Timmer, J. (2009-08-01). "Structural and practical identifiability analysis of partially observed dynamical models by exploiting the profile likelihood" (PDF). Bioinformatics. 25 (15): 1923–1929. doi: 10.1093/bioinformatics/btp358 . ISSN   1460-2059. PMID   19505944.
10. Van Waarde, Henk J.; Eising, Jaap; Camlibel, M. Kanat; Trentelman, Harry L. (2023). "The Informativity Approach: To Data-Driven Analysis and Control". IEEE Control Systems. 43 (6): 32–66. arXiv: 2302.10488 . Bibcode:2023ICSys..43f..32V. doi:10.1109/MCS.2023.3310305. ISSN   1066-033X. S2CID   257050367.
11. Gevers, Michel; Bazanella, Alexandre S.; Coutinho, Daniel F.; Dasgupta, Soura (2013). "Identifiability and excitation of polynomial systems". 52nd IEEE Conference on Decision and Control. Firenze: IEEE. pp. 4278–4283. doi:10.1109/CDC.2013.6760547. ISBN   978-1-4673-5717-3. S2CID   7796419.
12. Karin, Omer; Swisa, Avital; Glaser, Benjamin; Dor, Yuval; Alon, Uri (2016). "Dynamical compensation in physiological circuits". Molecular Systems Biology. 12 (11): 886. doi:10.15252/msb.20167216. ISSN   1744-4292. PMC   5147051 . PMID   27875241.
13. Sontag, Eduardo D. (2017-04-06). Komarova, Natalia L. (ed.). "Dynamic compensation, parameter identifiability, and equivariances". PLOS Computational Biology. 13 (4): e1005447. Bibcode:2017PLSCB..13E5447S. doi: 10.1371/journal.pcbi.1005447 . ISSN   1553-7358. PMC   5398758 . PMID   28384175.
14. Zhang, Han; Ringh, Axel (2024). "Inverse optimal control for averaged cost per stage linear quadratic regulators". Systems & Control Letters. 183 105658. arXiv: 2305.15332 . doi:10.1016/j.sysconle.2023.105658.
15. Barreiro, Xabier Rey; Villaverde, Alejandro F. (2023-01-31). "Benchmarking tools for a priori identifiability analysis". Bioinformatics. 39 (2): btad065. doi:10.1093/bioinformatics/btad065. ISSN   1367-4811. PMC   9913045 . PMID   36721336.
16. Díaz-Seoane, Sandra; Rey-Barreiro, Xabier; Villaverde, Alejandro F. (2023). "STRIKE-GOLDD 4.0: user-friendly, efficient analysis of structural identifiability and observability". Bioinformatics. 39 (1) btac748. arXiv: 2207.07346 . Bibcode:2023Bioin..39C.748D. doi:10.1093/bioinformatics/btac748.
17. Dong, Ruiwen; Goodbrake, Christian; Harrington, Heather A.; Pogudin, Gleb (2023-03-31). "Differential Elimination for Dynamical Models via Projections with Applications to Structural Identifiability". SIAM Journal on Applied Algebra and Geometry. 7 (1): 194–235. arXiv: 2111.00991 . doi:10.1137/22M1469067. ISSN   2470-6566. S2CID   245650629.
18. Borisov, Ivan; Metelkin, Evgeny (2020). Beard, Daniel A. (ed.). "Confidence intervals by constrained optimization—An algorithm and software package for practical identifiability analysis in systems biology". PLOS Computational Biology. 16 (12): e1008495. Bibcode:2020PLSCB..16E8495B. doi: 10.1371/journal.pcbi.1008495 . ISSN   1553-7358. PMC   7785248 . PMID   33347435.