Identification of partially known non-linear stochastic spatio-temporal dynamical systems by using a novel partially linear Kernel method
- Author(s): Hanwen Ning 1 and Xingjian Jing 2, 3
-
-
View affiliations
-
Affiliations:
1:
School of Statistics and Mathematics, Zhongnan University of Economics and Law, Wuhan, Hubei, People's Republic of China;
2: Department of Mechanical Engineering, Hong Kong Polytechnic University, HungHom, Kowloon, Hong Kong, People's Republic of China;
3: Hong Kong Polytechnic University, Shenzhen Research Institute, Shenzhen, People's Republic of China
-
Affiliations:
1:
School of Statistics and Mathematics, Zhongnan University of Economics and Law, Wuhan, Hubei, People's Republic of China;
- Source:
Volume 9, Issue 1,
02 January 2015,
p.
21 – 33
DOI: 10.1049/iet-cta.2014.0242 , Print ISSN 1751-8644, Online ISSN 1751-8652
The identification of non-linear stochastic spatio-temporal dynamical systems given by stochastic partial differential equations is of great significance to engineering practice, since it can always provide useful insight into the mechanism and physical characteristics of the underlying dynamics. In this study, based on the difference method for stochastic partial differential equations, a novel state-space model named multi-input–multi-output extended partially linear model for stochastic spatio-temporal dynamical system is proposed. A new Reproducing Kernel Hilbert Space-based algorithm named extended partially linear least square ridge regression is thus particularly developed for the identification of the extended partially linear model. Compared with existing identification methods available for spatio-temporal dynamics, the advantages of the proposed identification method include that (i) it can make full use of the partially linear structural information of physical models, (ii) it can achieve more accurate estimation results for system non-linear dynamics and (iii) the resulting estimated model parameters have clear physical meaning or properties closely related to the underlying dynamical system. Moreover, the proposed extended partially linear model also provide a convenient state-space model for system analysis and design (e.g. controller or filter design) of the class of non-linear stochastic partial differential dynamical systems.
Inspec keywords: Hilbert spaces; spatiotemporal phenomena; nonlinear estimation; MIMO systems; regression analysis; nonlinear dynamical systems; stochastic systems; control system synthesis; partial differential equations; least mean squares methods; state-space methods
Other keywords: nonlinear system analysis; partially linear kernel method; state-space model; stochastic partial differential equations; nonlinear stochastic spatiotemporal dynamical system identification; model parameter estimation; partially linear structural information; extended partially linear least square ridge regression model; physical model; nonlinear system design; identification method; nonlinear stochastic partial differential dynamical system; multiple input multiple output extended partially linear model; kernel Hilbert space-based algorithm
Subjects: Differential equations (numerical analysis); Interpolation and function approximation (numerical analysis); Nonlinear control systems; Control system analysis and synthesis methods; Simulation, modelling and identification; Multivariable control systems; Time-varying control systems; Other topics in statistics
References
-
-
1)
-
28. Walsh, J.B.: ‘An introduction to Stochastic partial differential equations(Lecturer Notes in Mathematics, 1180), Springer, Berlin, 1986, pp. 265–439.
-
-
2)
-
47. Ai, L., Ye, S.: ‘Model predictive control for nonlinear distributed parameter systems based on LSSVM’, Asian J. Control, 2013, 15, pp. 1407–1416.
-
-
3)
-
7. Yoo, H.: ‘Semi-discretization of stochastic partial differential equations on R1 by a finite-difference method’, Math. Comput., 1999, 69, pp. 653–666 (doi: 10.1090/S0025-5718-99-01150-3).
-
-
4)
-
56. Bashirov, A.E., Etikan, H., Semi, N.: ‘Filtering, smoothing and prediction for wide band noise driven systems’, J. Franklin Inst. Eng. Appl. Math., 1997, 334, pp. 667–683 (doi: 10.1016/S0016-0032(96)00052-X).
-
-
5)
-
9. Bishop, C.M.: ‘Pattern recognition and machine learning’ (Springer, Singapore, 2006).
-
-
6)
-
48. Chenkun, Q., Li, H., Zhang, X., Zhao, X., Li, S., Gao, F.: ‘Time-space separation based SVM modeling for nonlinear distributed parameter processes’, Ind. Eng. Chem. Res., 2011, 50, pp. 332–342 (doi: 10.1021/ie1002075).
-
-
7)
-
39. Laurain, V., Xing, W., Toth, R.: ‘Introducing instrumental variables in the LSSVM based identification framework’. 50th IEEE Conf. on Decision and Control and European Control Conf., 2011.
-
-
8)
-
22. Scholkopf, B., Smola, A.J.: ‘Learning with Kernels: support vector machines, regularization, optimization and beyond’ (MIT Press, Cambridge, 2002).
-
-
9)
-
33. Sardanyons, L.Q., Sole, M.S.: ‘Space semi-discretisations for a stochastic wave equation’, Potential Anal., 2006, 24, pp. 303–332 (doi: 10.1007/s11118-005-9002-0).
-
-
10)
-
21. Guo, L., Billings, S.A.: ‘State-space reconstruction and spatiotemporal prediction of lattice dynamical systems’, IEEE Trans. Autom. Control, 2007, 4, pp. 622–632 (doi: 10.1109/TAC.2007.894513).
-
-
11)
- A. Hagenblad , L. Ljung , A. Wills . Maximum likelihood identification of Wiener models. Automatica , 11 , 2697 - 2705
-
12)
-
35. Wu, Q., Ying, Y.M., Zhou, D.X.: ‘Learning rates of least square regularized regression’, Found. Comput. Math., 2006, 6, pp. 171–192 (doi: 10.1007/s10208-004-0155-9).
-
-
13)
-
17. Billings, S.A., Guo, L., Wei, H.: ‘Identification of coupled map lattice models for spatiotemporal patterns using wavelets’, Int. J. Syst. Sci., 2006, 37, pp. 1021–1038 (doi: 10.1080/00207720600963015).
-
-
14)
-
15. Jing, X.J.: ‘Robust adaptive learning of feedforward neural networks via LMI optimizations’, Neural Netw., 2012, 31, pp. 33–45 (doi: 10.1016/j.neunet.2012.03.003).
-
-
15)
-
58. Real, J.: ‘Stochastic partial differential equations with delays’, Stochastics, 1982, 8, pp. 81–102 (doi: 10.1080/17442508208833230).
-
-
16)
-
19. Li, H., Qi, C.: ‘Spatiotemporal modeling of nonlinear distributed parameter systems’, (Springer, Berlin, 2011).
-
-
17)
-
27. Duffy, D.J.: ‘Finite difference methods in financial engineering: a partial differential approach’ (John Wiley and Sons, England, 2006).
-
-
18)
-
26. Xu, Y., Chen, D.: ‘Partially linear least squares regularized regression for system identification’, IEEE Trans. Autom. Control, 2009, 54, (11), pp. 2637–2641 (doi: 10.1109/TAC.2009.2031566).
-
-
19)
-
20. Coca, D., Billings, S.A.: ‘Identification of finite dimensional models of infinite dimensional dynamical systems’, Automatica, 2002, 38, pp. 1851–1865 (doi: 10.1016/S0005-1098(02)00099-7).
-
-
20)
-
45. Suykens, J.A.K., Vandewalle, J., Moor, B.D.: ‘Optimal control by least squares support vector machines’, Neural Netw., 2011, 24, pp. 23–35.
-
-
21)
-
54. Hairer, M.: ‘Rough stochastic partial differential equations’, Commun. Pure Appl. Math., 2011, 64, pp. 1547–1585.
-
-
22)
-
2. Chow, P.: ‘Stochastic partial differential equations’ (Chapman and Hall, BocaRaton, FL, 2007).
-
-
23)
-
6. Allen, E.J., Novosel, S.J., Zhang, Z.: ‘Finite element and difference approximation of some linear stochastic partial differential equations’, Stoch. Stoch. Rep., 1998, 64, pp. 117–142 (doi: 10.1080/17442509808834159).
-
-
24)
-
3. Peszat, S., Zabczyk, J.: ‘Stochastic partial differential equations with levy noise: an evolution equation approach’ (Cambridge University Press, Cambridge, UK, 2007).
-
-
25)
-
36. Xu, Y.L., Chen, D.R., Li, H.X., Liu, L.: ‘Least square regularized regression in sum space’, IEEE Trans. Neural Netw., 2013, 24, pp. 635–646 (doi: 10.1109/TNNLS.2013.2242091).
-
-
26)
-
23. Ning, H., Jing, X., Chen, L.: ‘Online identification of nonlinear spatiotemporal systems using kernel learning approach’, IEEE Trans. Neural Netw., 2011, 22, (9), pp. 1381–1394 (doi: 10.1109/TNN.2011.2161331).
-
-
27)
-
32. Jentzen, A., Kloeden, P.E.: ‘The numerical approximation of stochastic partial differential equations’, Milan J. Math., 2009, 77, pp. 205–244 (doi: 10.1007/s00032-009-0100-0).
-
-
28)
- N.I. Mahmudov . Controllability of linear stochastic systems. IEEE Trans. Autom. Control , 5 , 724 - 731
-
29)
-
8. Kovacs, M., Larsson, S., Saedpanah, F.: ‘Finite element approximation of the linear stochastic wave equation with additive noise’, SIAM J. Numer. Anal., 2010, 2, pp. 408–427 (doi: 10.1137/090772241).
-
-
30)
-
44. Steinke, F., Scholkopf, B.: ‘Kernels regularization and differential equations’, Pattern Recognit., 2008, 41, pp. 3271–3286 (doi: 10.1016/j.patcog.2008.06.011).
-
-
31)
-
31. Gyongy, I.: ‘Lattice approximations for stochastic quasi-linear parabolic partial equations driven by space-time white noise II’, Potential Anal., 1999, 11, pp. 1–37 (doi: 10.1023/A:1008699504438).
-
-
32)
-
41. Tsybakov, A.B.: ‘Introduction to nonparametric estimation’ (Springer-Verlag, New York, 2009).
-
-
33)
-
16. Wei, H., Billings, S.A., Zhao, Y., Guo, L.: ‘Lattice dynamical wavelet neural networks implemented using particle swarm optimization for spatio-temporal system identification’, IEEE Trans. Neural Netw., 2009, 20, pp. 181–185 (doi: 10.1109/TNN.2008.2009639).
-
-
34)
-
15. Jing, X., Cheng, L.: ‘An optimal PID control algorithm for training feedforward neural networks’, IEEE Trans. Ind. Electron., 2013, 60, (6), pp. 2273–2283 (doi: 10.1109/TIE.2012.2194973).
-
-
35)
-
50. Rencher, A.C., Christensen, W.F.: ‘Methods of multivariate analysis’ (John Wiley and Sons, New York, 2012, 3rd edn.).
-
-
36)
-
55. Carreras, D.M., Mellou, M., Sarra, M.: ‘On stochastic partial differential equations with spatially correlated noise: smoothness of the law’, Stoch. Process. Appl., 2001, 93, pp. 269–284 (doi: 10.1016/S0304-4149(00)00099-5).
-
-
37)
-
59. Li, H.X., Qi, C.K.: ‘Modeling of distributed parameter systems for applications – a synthesized review from time-space separation’, J. Process. Control, 2010, 20, (8), pp. 891–901 (doi: 10.1016/j.jprocont.2010.06.016).
-
-
38)
-
43. Chan, W.C., Chan, C.W., Cheung, K.C., Harris, C.J.: ‘On the modelling of nonlinear dynamic systems using support vector neural networks’, Eng. Appl. Artif. Intell., 2001, 14, pp. 105–113 (doi: 10.1016/S0952-1976(00)00069-5).
-
-
39)
-
46. Cialenco, I., Fasshauer, G.E., Ye, Q.: ‘Approximation of stochastic partial differential equations by a Kernel-based collocation method’, Int. J. Comput. Math., 2012, 89, pp. 2543–2561 (doi: 10.1080/00207160.2012.688111).
-
-
40)
-
24. Ning, H., Jing, X., Chen, L.: ‘Identification of non-linear stochastic spatiotemporal dynamical systems’, IET Control Theory Appl., 2013, 7, pp. 2069–2083 (doi: 10.1049/iet-cta.2013.0150).
-
-
41)
-
4. Friedman, A.: ‘Stochastic differential equations and applications’ (Academic Press, New York, 1975).
-
-
42)
-
11. Chen, S., Xia, H., Harris, J.: ‘Particle swarm optimization aided orthogonal forward regression for unified data modeling’, IEEE Trans. Evol. Comput., 2010, 4, pp. 477–499 (doi: 10.1109/TEVC.2009.2035921).
-
-
43)
-
12. Vapnik, V.: ‘The nature of statistical learning theory’ (Springer-Verlag, New York, 1995).
-
-
44)
-
49. Farlow, S.J.: ‘Partial differential equations for scientists and engineers’ (John Wiley and Sons Inc., New York, 1982).
-
-
45)
-
13. Suykens, J.A.K., Gestel, V.T., Brabanter, J.D., Moor, B.D., Vandewalle, J.: ‘Least squares support vector machines’ (World Scientific, Singapore, 2002).
-
-
46)
-
37. Wasserman, L.: ‘All of nonparametric statistics’ (Springer, 2006).
-
-
47)
- M. Espinoza , J.A.K. Suykens , B. De Moor . Kernel based partially linear models and nonlinear identification. IEEE Trans. Autom. Control , 10 , 1602 - 1606
-
48)
-
1. Da Prato, G., Zabczyk, J.: ‘Stochastic equations in infinite dimensions’ (Cambridge University Press, Cambridge, UK, 1992).
-
-
49)
-
18. Guo, L., Billings, S.A.: ‘Identification of partial differential equation models for continuous spatiotemporal dynamical systems’, IEEE Trans. Circuits Syst. II, Express Briefs, 2006, 53, (8), pp. 657–661 (doi: 10.1109/TCSII.2006.876464).
-
-
50)
-
38. Brabanter, D.D., Brabanter, J.D., Suykens, J.A.K.: ‘Approximate confidence and prediction intervals for least squares support vector regression’, IEEE Trans. Neural Netw., 2011, 2, pp. 110–120 (doi: 10.1109/TNN.2010.2087769).
-
-
51)
-
28. Walsh, J.B.: ‘An introduction to Stochastic partial differential equations (Lecturer Notes in Mathematics, 1180), Springer, Berlin, 1986, pp. 265–439.
-
-
52)
- S. Chen , S.A. Billings , W. Luo . Orthogonal least squares methods and their application to non-linear system identification. Int. J. Control , 5 , 1873 - 1896
-
53)
-
57. Ferrante, M., Martaa, S.: ‘SPDEs with coloured noise: analytic and stochastic approaches’, ESAIM: Prob. Stat., 2006, 10, pp. 380–405 (doi: 10.1051/ps:2006016).
-
-
54)
- H.T. Toivonen , S. Tötterman , B. Åkesson . Identification of state-dependent parameter models with support vector regression. Int. J. Control , 9 , 1454 - 1470
-
55)
- G.C. Cawley , N.L.C. Talbot . Fast exact leave-one-out cross-validation of sparse least-squares support vector machines. Neural Netw. , 1467 - 1475
-
56)
-
34. Smale, S., Zhou, D.X.: ‘Estimating the approximation error in learning theory’, Anal. Appl., 2003, 1, pp. 1–25 (doi: 10.1142/S0219530503000089).
-
-
57)
-
29. McDonald, S.: ‘Finite difference approximation for linear Stochastic partial differential equations with method of lines’, MPRA Paper, 2006.
-
-
58)
-
1. Ljung, L.: ‘Perspectives on system identification’, Annu. Rev. Control, 2010, 34, pp. 1–12 (doi: 10.1016/j.arcontrol.2009.12.001).
-
-
59)
-
53. Hairer, M., Voss, J.: ‘Approximations to the stochastic Burgers equations’, J. Nonlinear Sci., 2011, 21, pp. 897–920 (doi: 10.1007/s00332-011-9104-3).
-
-
60)
-
30. Gyongy, I.: ‘Lattice approximations for stochastic quasi-linear parabolic partial equations driven by space-time white noise I’, Potential Anal., 1998, 9, pp. 1–25 (doi: 10.1023/A:1008615012377).
-
-
1)