Abstract
In this paper, we provide a complete answer to the question of characteristic cones for discrete autonomous nD systems, with arbitrary \(n\geqslant 2\), described by linear partial difference equations with real constant coefficients. A characteristic cone is a special subset (having the structure of a cone) of the domain (here \(\mathbb {Z}^n\)) such that the knowledge of the trajectories on this set uniquely determines them over the whole domain. Despite its importance in numerous system-theoretic issues, the question of characteristic sets for multidimensional systems has not been answered in its full generality except for Valcher’s seminal work for the special case of 2D systems (Valcher in IEEE Trans Circuits and Syst Part I Fundam Theory Appl 47(3):290–302, 2000). This apparent lack of progress is perhaps due to inapplicability of a crucial intermediate result by Valcher to cases with \(n\geqslant 3\). We illustrate this inapplicability of the above-mentioned result in Sect. 3 with the help of an example. We then provide an answer to this open problem of characterizing characteristic cones for discrete nD autonomous systems with general n; we prove an algebraic condition that is necessary and sufficient for a given cone to be a characteristic cone for a given system of linear partial difference equations with real constant coefficients. In the second part of the paper, we convert this necessary and sufficient condition to another equivalent algebraic condition, which is more suited from algorithmic perspective. Using this result, we provide an algorithm, based on Gröbner bases, that is implementable using standard computer algebra packages, for testing whether a given cone is a characteristic cone for a given discrete autonomous nD system.
Similar content being viewed by others
Notes
A matrix \(P(\varvec{\xi },\varvec{\xi }^{-1})\) is called left-factor-prime if any decomposition \(P = E P_1\), where E is square, implies that E is invertible as a Laurent polynomial matrix, i.e., \(\mathrm{det~}E\) is a unit in \(\mathcal {A}\). A matrix \(R(\varvec{\xi },\varvec{\xi }^{-1})\) is said to be right-factor-prime if \(R^T(\varvec{\xi },\varvec{\xi }^{-1})\) is left-factor-prime. See Youla and Gnavi (1979) for more details.
A closed, pointed, solid, convex cone is called a proper cone; we elaborate more on this in Sect. 4.
A Hamel basis of a possibly infinite dimensional vector space \(\mathcal {V}\) over a field \(\mathbb {K}\) is a subset \(\mathcal {E}\) of \(\mathcal {V}\) that satisfies:
-
1.
elements in \(\mathcal {E}\) are linearly independent over \(\mathbb {K}\), that is, no finite non-zero linear combination of elements in \(\mathcal {E}\) equals zero, and
-
2.
every element of \(\mathcal {V}\) can be written as a finite linear combination of elements from \(\mathcal {E}\).
See Limaye (1996, Section 2). Note also that \(\mathcal {M}\) admits a countable Hamel basis. This justifies writing down a basis of \(\mathcal {M}\) as a set \(\mathcal {E}\) which is indexed by the natural numbers.
-
1.
The ordering requires to satisfy the following properties: 1) it must be a total ordering on \(\mathbb {Z}^n_{\geqslant 0}\), 2) it must respect the additive structure of \(\mathbb {Z}^n_{\geqslant 0}\), 3) it must be a well-ordering. See (Cox et al. 2007, Chapter 2, Section 2)
It is important to note that division here refers to division in \(\mathbb {R}[\varvec{\xi }]^q\). An algorithm for this can be found in Adams and Loustaunau (2012, Algorithm 3.5.1)
The scalar multiplication property of the homomorphism obeys the following relation: for \(\alpha \in \mathbb {R}[\varvec{\xi },\varvec{\eta }]\), \(\varvec{t} \in \mathbb {R}[\varvec{\xi },\varvec{\eta }]^{1\times q}\)
Refer to Eq. (27).
Refer to Eq. (27).
An algorithm for calculating the Gröbner basis of a module can be found in Adams and Loustaunau (2012, Algorithm 3.5.2).
Note that, the symbols \(z_1,z_2\) in Valcher (2000) carry the meaning of \(\sigma _1^{-1},\sigma _2^{-1}\) of this paper, respectively.
References
Adams, W. W., & Loustaunau, P. (2012). An introduction to Gröbner Bases (Vol. 3). New Delhi: American Mathematical Society.
Atiyah, M., & MacDonald, I. (1969). Introduction to commutative algebra. Britain: Addison-Wesley Publishing Company.
Avelli, D. N., Rapisarda, P., & Rocha, P. (2011a). Lyapunov stability of \(2\)D finite-dimensional behaviours. International Journal of Control, 84(4), 737–745.
Avelli, D. N., Rapisarda, P., & Rocha, P. (2011b). Time-relevant \(2\)D behaviors. Automatica, 47(11), 2373–2382.
Avelli, D. N., Rapisarda, P., & Rocha, P. (2012). Lyapunov functions for time-relevant \(2\)D systems, with application to first-orthant stable systems. Automatica, 48(9), 1998–2006.
Björk, J. E. (1979). Rings of differential operators. New York: North Holland Publishing Company.
Cox, D., Little, J., & O’Shea, D. (2007). Ideals, varieties and algorithms. New York: Springer.
Fornasini, E., Rocha, P., & Zampieri, S. (1993). State space realizations of 2-D finite-dimensional behaviours. SIAM Journal of Control and Optimization, 31(6), 1502–1517.
Fornasini, E., & Valcher, M. E. (1997). nD polynomial matrices with applications to multidimensional signal analysis. Multidimensional Systems and Signal Processing, 8, 387–407.
Kreuzer, M., & Robbiano, L. (2000). Computational Commutative Algebra 1. Berlin: Springer.
Limaye, B. V. (1996). Functional analysis. New Delhi: New Age International (P) Ltd., Publishers.
Miller, E., & Sturmfels, B. (2004). Combinatorial commutative algebra. New York: Springer.
Mukherjee, M., & Pal, D. (2016). On characteristic cones of scalar autonomous \(n\)D systems, with general \(n\). In 22nd International symposium on mathematical theory of networks and systems, Minneapolis, USA (pp. 839–845).
Mukherjee, M., & Pal, D. (2017). Algorithms for verification of characteristic sets of discrete autonomous \(n\)D systems with \(n\geqslant 2\). IFAC PapersOnline, 50(1), 1840–1846.
Oberst, U. (1990). Multidimensional constant linear systems. Acta Applicandae Mathematicae, 20, 1–175.
Pal, D. (2015). Every discrete 2D autonomous system admits a finite union of parallel lines as a characteristic set. Multidimensional Systems and Signal Processing,. https://doi.org/10.1007/s11045-015-0330-y.
Pal, D., & Pillai, H. K. (2011). Lyapunov stability of \(n\)-D strongly autonomous systems. International Journal of Control, 84(11), 1759–1768.
Pal, D., & Pillai, H. K. (2014). On restrictions of \(n\)-d systems to 1-d subspaces. Multidimensional Systems and Signal Processing, 25, 115–144.
Pal, D., & Pillai, H. K. (2016). Multidimensional behaviors: The state-space paradigm. Systems and Control Letters, 95, 27–34.
Pauer, F., & Unterkircher, A. (1999). Gröbner bases for ideals in Laurent polynomial rings and their application to systems of difference equations. Applicable Algebra in Engineering, Communication and Computing, 9, 271–291.
Pillai, H. K., & Shankar, S. (1998). A behavioral approach to control of distributed systems. SIAM Journal on Control and Optimization, 37(2), 388–408.
Pommaret, J. F., & Quadrat, A. (1999). Algebraic analysis of linear multidimensional control systems. IMA Journal of Mathematical Control & Information, 16, 275–297.
Rocha, P., & Willems, J. C. (1989). State for 2-D systems. Linear Algebra and Its Applications, 122–124, 1003–1038.
Rocha, P., & Willems, J. C. (2006). Markov properties for systems described by PDEs and first-order representations. Systems and Control Letters, 55, 538–542.
Rogers, E., Galkowski, K., Paszke, W., Moore, K. L., Bauer, P. H., Hladowski, L., et al. (2015). Multidimensional control systems: case studies in design and evaluation. Multidimensional Systems and Signal Processing, 26, 895–939.
Valcher, M. E. (2000). Characteristic cones and stability properties of two-dimensional autonomous behaviors. IEEE Transactions On Circuits and Systems Part I Fundamental Theory and Applications, 47(3), 290–302.
Willems, J. C. (1989). Models for dynamics. Dynamics Reported, 2, 171–269.
Willems, J. C. (1991). Paradigms and puzzles in theory of dynamical systems. IEEE Transactions On Automatic Control, 36(6), 259–294.
Wood, J. (2000). Modules and behaviours in \(n\)D systems theory. Multidimensional Systems and Signal Processing, 11, 11–48.
Wood, J., Rogers, E., & Owens, D. H. (1999). Controllable and autonomous \(n\)D linear systems. Multidimensional Systems and Signal Processing, 10, 33–69.
Wood, J., Sule, V. R., & Rogers, E. (2005). Causal and stable input/output structures on multidimensional behaviors. SIAM Journal on Control and Optimization, 43(4), 1493–1520.
Youla, D. C., & Gnavi, G. (1979). Notes on n-dimensional system theory. IEEE Transactions on Circuits and Systems, Cas–26(2), 105–111.
Zerz, E., & Oberst, U. (1993). The canonical Cauchy problem for linear systems of partial difference equations with constant coefficients over the complete \(r\)-dimensional integral lattice \(\mathbb{Z}^{r}\). Acta Applicandae Mathematicae, 31, 249–273.
Acknowledgements
This work has been supported in parts by DST-INSPIRE Faculty Grant, the Department of Science and Technology (DST), Govt. of India (Grant Code: IFA14-ENG-99); and the Industrial Research and Consultancy Centre (IRCC) IIT Bombay (Project ID: 15IRCCSG012).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mukherjee, M., Pal, D. On characteristic cones of discrete nD autonomous systems: theory and an algorithm. Multidim Syst Sign Process 30, 611–640 (2019). https://doi.org/10.1007/s11045-018-0571-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11045-018-0571-7