The paper addresses data-driven statistical pattern identification in complex dynamical systems, where the concept is built upon thermodynamic formalism of symbolic data sequences in the setting of lattice spin systems. The transfer matrix approach has been used for generation of pattern vectors from time series data of observed parameters. Efficacy of pattern identification is demonstrated for early detection of anomalies (i.e., deviations from the nominal pattern) on an experimental apparatus of nonlinear active electronic circuits.

1.
R.
Badii
and
A.
Politi
,
Complexity Hierarchical Structures and Scaling in Physics
(
Cambridge University Press
,
Cambridge, UK
,
1997
), Chap. 6, pp.
119
123
.
2.
R. K.
Pathria
,
Statistical Mechanics
(
Butterworth-Heinemann
,
Woburn, MA, USA
,
1996
), Chap. 3, pp.
43
56
.
3.
D.
Ruelle
,
Thermodynamic Formalism
(
Cambridge University Press
,
Cambridge, UK
,
2004
), Chap. 1, pp.
1
9
.
4.
C.
Beck
and
F.
Schlögl
,
Thermodynamics of Chaotic Systems: An Introduction
(
Cambridge University Press
,
Cambridge, UK
,
1993
), Chap. 8, pp.
78
87
.
5.
A.
Ray
,
Signal Processing
,
84
,
1115
(
2004
).
6.
D. P.
Feldman
, Ph.D. thesis,
University of California Davis
,
1998
.
7.
D.
Lind
and
M.
Marcus
,
An Introduction to Symbolic Dynamics and Coding
(
Cambridge University Press
,
Cambridge, UK
,
1995
), Chap. 2, pp.
28
48
.
8.
V.
Rajagopalan
and
A.
Ray
,
Signal Processing
,
86
,
3309
(
2006
).
You do not currently have access to this content.