Identifiabilities and nonlinearities

  • E. Walter
  • L. Pronzato

Abstract

A parametric model structure is (globally) identifiable if the parameter vector associated with a given input-output behavior is unique. If this is not the case, careless estimation of this vector from experimental data may lead to completely erroneous results. (The situation is similar when unobserved state variables have to be estimated from the outputs.) The effect of two types of nonlinearity of the output (with respect to the parameters and to the inputs) on structural identifiability is described. Various methods to test state space models for identifiability are presented. They apply to models that are nonlinear with respect to the parameters and may be linear or not with respect to the inputs. Their use is illustrated on simple examples and an actual problem in chemical engineering. Advantages and limitations of the existing techniques are evidenced. Relationships between structural identifiability and practical identifiability, i.e. the ability to actually estimate the parameters of the model from experimental data, are considered, as well as techniques available to design an experiment so as to make the parameters of interest as identifiable as possible.

Keywords

Methane Manifold Covariance Catalysis Helium 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Bibliography

  1. [1]
    Anderson, D. H. (1982). Compartmental Modeling and Tracer Kinetics, Springer, Berlin.Google Scholar
  2. [2]
    ÅSTRÖM, K. J. (1972). System Identification. Lecture Notes Math. Germ., 294, 35–55.CrossRefGoogle Scholar
  3. [3]
    Atkinson, A. C. and D. R. Cox (1974). Planning experiments for discriminating between models (with discussion), J. Royal Statist. Soc., B 36, 321–348.MathSciNetMATHGoogle Scholar
  4. [4]
    Atkinson, A. C. and A. N. Donev (1992). Optimal Experimental Design, Oxford University Press, Oxford.Google Scholar
  5. [5]
    Audoly, S. and L. D’Angio (1983). On the identifiability of linear compartmental systems: a revisited transfer function approach based on topological properties. Math. Biosci., 66, 201–228.MathSciNetMATHCrossRefGoogle Scholar
  6. [6]
    Bellman, R. and K. J. Åström (1970). On structural identifiability. Math. Biosci., 7, 329–339.CrossRefGoogle Scholar
  7. [7]
    Berman, M. and R. Schoenfeld (1956). Invariants in experimental data on linear kinetics and the formulation of models. J. of Applied Physics, 27,1361–1370.CrossRefGoogle Scholar
  8. [8]
    Berntsen, H. E. and J. G. Balchen (1973). Identifiability of linear dynamical systems. Proc 3rd IFAC Symp. on Identification and System Parameter Estimation, The Hague, 871–874.Google Scholar
  9. [9]
    Buchberger, B. (1970). Ein algorithmisches Kriterium für die Lösbarkeit eines algebraischen Gleichungssystems. Aeq. Math., 4, 374–383.MathSciNetMATHCrossRefGoogle Scholar
  10. [10]
    Chappell, M. J., K. R. Godfrey and S. Vajda (1990). Global identifiability of the parameters of nonlinear systems with specified inputs: a comparison of methods. Math. Biosci., 102, 41–73.MATHCrossRefGoogle Scholar
  11. [11]
    Cobelli, C., A. Lepschy and G. Romanin-Jacur (1979a). Identifiability results on some constrained compartmental systems. Math. Biosci., 47, 173–196.MathSciNetMATHCrossRefGoogle Scholar
  12. [12]
    Cobelli, C., A. Lepschy and G. Romanin-Jacur (1979b). Identifiability of compartmental models and related structural properties. Math. Biosci., 48, 1–18.CrossRefGoogle Scholar
  13. [13]
    Diop, S. (1989). Théorie de l’élimination et principe du modèle interne en automatique. Doctorate thesis, Université Paris Sud, Orsay, France.Google Scholar
  14. [14]
    Diop, S. and M. Fliess (1991). Nonlinear observability, identifiability and persistent trajectories. Proc. 30th IEEE CDC, Brighton, 714–719.Google Scholar
  15. [15]
    Fedorov, V. V. (1972). Theory of Optimal Experiments. Academic Press, New York.Google Scholar
  16. [16]
    Fisher, F. M. (1966). The Identification Problem in Economics. McGraw-Hill, New York.Google Scholar
  17. [17]
    Fliess, M. (1981). Fonctionnelles causales non linéaires et indéterminées non commutatives. Bull. Soc. Math, de France. 109, 3–40.MathSciNetMATHGoogle Scholar
  18. [18]
    Fliess, M. (1989). Automatique et corps différentiels. Forum Math., 1, 227–238.MathSciNetMATHCrossRefGoogle Scholar
  19. [19]
    Glad, S. T. and L. Ljung (1990). Parametrization of nonlinear model structures as linear regressions. Prep, 11th IFAC World Congress, Tallinn, 6, 67–71.Google Scholar
  20. [20]
    Glover, K. and J. C. Willems (1974). Parametrizations of linear dynamical systems: canonical forms and identifiability. IEEE Trans. Autom. Control, AC-19, 640–644.MathSciNetMATHCrossRefGoogle Scholar
  21. [21]
    Grewal, M. S. and K. Glover (1976). Identifiability of linear and nonlinear dynamical systems. IEEE Trans. Autom. Control, AC-21, 833–837.MathSciNetMATHCrossRefGoogle Scholar
  22. [22]
    Hermann, R. and A. J. Krener (1977). Nonlinear controllability and observability. IEEE Trans. Autom. Control, AC-22, 728–740.MathSciNetMATHCrossRefGoogle Scholar
  23. [23]
    Hill, P. D. H. (1978). A review of experimental design procedures for regression model discrimination, Technometrics, 20, 15–21.MATHCrossRefGoogle Scholar
  24. [24]
    Horst, R. and H. Tuy (1990). Global Optimization, Deterministic Approaches. Springer-Verlag, Berlin.MATHGoogle Scholar
  25. [25]
    Huang, C.-Y. (1991). Planification d’expériences pour la discrimination entre modèles. Doctorate thesis, Université de Paris Sud, Orsay, France.Google Scholar
  26. [26]
    Launay, G. (1989). Modélisation du stockage de la sérotonine dans les plaquettes humaines. Doctorate thesis, Université Paris Dauphine, Paris, France.Google Scholar
  27. [27]
    Lecourtier, Y., F. Lamnabhi-Lagarrigue and E. Walter (1987). Volterra and generating power series approaches to identifiability testing. In E. Walter (ed.), Identifiability of Parametric Models. Pergamon, Oxford, pp. 5066.Google Scholar
  28. [28]
    Lecourtier, Y. and A. Raksanyi (1987). The testing of structural properties through symbolic computation. In E. Walter (ed.), Identifiability of Parametric Models. Pergamon, Oxford, pp. 75–84.Google Scholar
  29. [29]
    Ljung, L. (1987). System Identification, Theory for the User. Prentice-Hall, En-glewood Cliffs.MATHGoogle Scholar
  30. [30]
    Ljung, L. and T. Glad (1994). On global identifiability for arbitrary model parametrizations. Automatica, 30 (2), 265–276.MathSciNetMATHCrossRefGoogle Scholar
  31. [31]
    Norton, J. P. (1986). An Introduction to Identification. Academic Press, London.MATHGoogle Scholar
  32. [32]
    Ollivier, F. (1990). Le problème de l’identifiabilité structurelle globale: approche théorique, méthodes effectives et bornes de complexité. Doctorate thesis, Ecole Polytechnique, Palaiseau, France.Google Scholar
  33. [33]
    Pazman, A.(1986). Foundations of Optimum Experimental Design. VEDA, Bratislava.MATHGoogle Scholar
  34. [34]
    Pohjanpalo, H. (1978). System identifiability based on the power series expansion of the solution, Math. Biosci., 41, 21–33.MathSciNetMATHCrossRefGoogle Scholar
  35. [35]
    Ratschek, H. and J. Rokne (1988). New Computer Methods for Global Optimization. Ellis Horwood, Chichester. (Distributed by John Wiley).MATHGoogle Scholar
  36. [36]
    Ritt, J. F. (1950). Differential Algebra, American Mathematical Society, Providence.MATHGoogle Scholar
  37. [37]
    Silvey, S. D. (1980). Optimal Design. Chapman & Hall, London.MATHCrossRefGoogle Scholar
  38. [38]
    Steinberg, D. M. and W. G. Hunter (1984). Experimental design: review and comment (with discussion). Technometrics, 26, 71–130.MathSciNetMATHCrossRefGoogle Scholar
  39. [39]
    Sussman, H. J. (1977). Existence and uniqueness of minimal realizations of nonlinear systems. Math. Syst. Theory, 10, 263–284.MATHCrossRefGoogle Scholar
  40. [40]
    Titterington, D. M. (1980). Aspects of optimal design in dynamic systems. Technometrics, 22, 287–299.MathSciNetMATHCrossRefGoogle Scholar
  41. [41]
    Tunali, E. T. and T. J. Tarn (1987). New results for identifiability of nonlinear systems. IEEE Trans. Autom. Control, AC-32, 146–154.MathSciNetMATHCrossRefGoogle Scholar
  42. [42]
    Vajda, S., K. R. Godfrey and H. Rabitz (1989). Similarity transformation approach to identifiability analysis of nonlinear compartmental models. Math. Biosci., 93, 217–248.MathSciNetMATHCrossRefGoogle Scholar
  43. [43]
    Vajda, S. and H. Rabitz (1989). State isomorphism approach to global identifiability of nonlinear systems. IEEE Trans. Autom. Control, AC-34, 220–223.MathSciNetMATHCrossRefGoogle Scholar
  44. [44]
    Vajda, S., H. Rabitz, E. Walter and Y. Lecourtier (1989). Qualitative and quantitative identifiability analysis of nonlinear chemical models. Chem. Eng. Comm., 83, 191–219.CrossRefGoogle Scholar
  45. [45]
    Walter, E. (1982). Identifiability of State Space Models. Springer, Berlin.MATHCrossRefGoogle Scholar
  46. [46]
    Walter, E. (Ed.) (1987). Identifiability of Parametric Models. Pergamon, Oxford.MATHGoogle Scholar
  47. [47]
    Walter, E. and Y. Lecourtier (1981). Unidentifiable compartmental models: what to do? Math. Biosci., 56, 1–25.MathSciNetMATHCrossRefGoogle Scholar
  48. [48]
    Walter, E. and Y. Lecourtier (1982). Global approaches to identifiability testing for linear and nonlinear state space models. Math. and Comput. in Simul., 24, 472–482.MathSciNetMATHCrossRefGoogle Scholar
  49. [49]
    Walter, E., Y. Lecourtier and J. Happel (1984). On the structural output distinguishability of parametric models, and its relations with structural identifiability. IEEE Trans. Autom. Control, AC- 29, 56–57.MATHCrossRefGoogle Scholar
  50. [50]
    Walter, E., Y. Lecourtier,A. Raksanyi and J. Happel (1985). On the distinguishability of parametric models with different structures. In J. Eisenfeld and C. De Lisi (eds), Mathematics and Computers in Biomedical Applications, North-Holland, Amsterdam, pp. 145–160.Google Scholar
  51. [51]
    Walter, E., H. Piet-Lahanier and J. Happel (1986). Estimation of nonuniquely identifiable parameters via exhaustive modeling and membership set theory, Math. and Comput. in Simul., 28, 479–490.CrossRefGoogle Scholar
  52. [52]
    Walter, E. and L. Pronzato (1990). Qualitative and quantitative experiment design for phenomenological models — a survey. Automatica, 26, 195–213.MathSciNetMATHCrossRefGoogle Scholar
  53. [53]
    Walter, E. and L. Pronzato (1994). Identification de modèles paramétriques à partir de données expérimentales. Masson, Paris.Google Scholar
  54. [54]
    Zarrop, M. B. (1979). Optimal Experiment Design for Dynamic System Identification. Springer, Heidelberg.MATHCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 1995

Authors and Affiliations

  • E. Walter
  • L. Pronzato

There are no affiliations available

Personalised recommendations