We have applied tailor-made neural networks to the post-linearization of nonlinear systems containing memory. The problems we address involve tracking systems that contain linear dynamics together with memoryless, nonlinear sensors and amplifiers. In general, the goal is to accurately infer a system’s inputs based only on the system’s outputs, which have been corrupted by nonlinear components. The linearizing neural network is trained to emulate the inverse of the Volterra operator which describes the nonlinear system. In implementation, the network estimates the original input signal from the system’s output sequence. The post-linearizing network architecture is determined from an approximate model of the system to be linearized. The network is trained with test signals that excite the tracking system over its domain of operation and expose much of its nonlinear behavior. Network weights and biases are adjusted using a novel algorithm, batch backpropagation-through-time (BBTT). This paper presents a test case involving a sensor with an input-output relation similar to that of a scaled dc SQUID. The sensor and amplifier nonlinearities are embedded within a fourth-order dynamic system with negative feedback. The problem is generally formulated and we discuss the application of our methodology to a variety of nonlinear sensing and amplification systems.

Issue Section:
Technical Papers
Topics:
Artificial neural networks, Nonlinear systems, Sensors, Signals, Algorithms, Dynamic systems, Dynamics (Mechanics), Feedback, Superconducting quantum interference devices
1.
Kuznik
F.
, “
The Einstein Test
,”
Air & Space
, Vol.
7
, No.
5
, pp.
70
–
76
, Dec.
1992
.
2.
Gutt
G. M.
,
Kasdin
N. J.
,
Condron
M. R.
,
Muhfelder
B.
,
Lockhart
J. M.
, and
Cromar
M. W.
, “
A Method for Simulating a Flux-Locked DC SQUID
,”
IEEE Transactions on Applied Superconductivity
, Vol.
3
, No.
1
, pt. 4, pp.
1837
–
40
, Mar.
1993
.
3.
Wellstood
F.
,
Heidan
C.
, and
Clarke
J.
, “
Integrated DC SQUID Magnetometer With a High Slew Rate
,”
Rev. Sci. Instrum.
, Vol.
55
, No.
6
, pp.
952
–
957
, June
1984
.
4.
A. M. Liapunov, “Probleme General de la Stabilite du Mouvement,” Annals of Mathematical Studies, Vol. 17, Princeton University Press, Princeton, NJ, 1949.
5.
S. Boyd, “Volterra Series: Engineering Fundamentals,” Ph.D. dissertation, University of California at Berkeley, Berkeley, 1985.
6.
S. Boyd, “Input/Output Linearization of Systems Described by Volterra Series,” Information Systems Laboratory, Dept. of Electrical Engineering, Stanford University, Apr. 1988.
7.
deFigueiredo
R. J. P.
, and
Dwyer
T. A.
, “
A Best Approximation Framework and Implementation for Simulation of Large-Scale Nonlinear Systems
,”
IEEE Trans. Circuits and Syst.
, Vol.
CAS-2
, pp.
1005
–
1014
, Nov.
1980
.
8.
Derrick H. Nguyen, “Applications of Neural Networks in Adaptive Control,” Ph.D. dissertation, Dept. of Electrical Engineering, Stanford University, Stanford, CA, June 1991.
9.
Lippmann
R. P.
, “
An Introduction to Computing with Neural Nets
,”
IEEE ASSP Magazine
, Vol.
4
, pp.
4
–
22
, Apr.
1987
.
10.
A. Lapedes, and R. Farber, “Nonlinear Signal Processing Using Neural Networks: Predication and System Modeling,” Tech. Rep., Los Alamos National Laboratory, LA-UR-87-2662, 1987.
11.
David E. Rumelhart and James L. McLelland, and the PDP Research Group, Parallel Distributed Processing, Explorations in the Microstructures of Cognition, Vol. 1, MIT Press, Cambridge, Mass. ISBN 0-262-18120-7, 1987.
12.
L. C. Ludeman, Fundamentals of Digital Signal Processing, Wiley, New York, pp. 57–67, 1986.
13.
Girish Govind, and P. A. Ramapoorthy, “Multi-Layered Neural Networks and Volterra Series: The Missing Link,” IEEE International Conference on Systems Engineering, pp. 633–636, Aug. 1990.
14.
A. Lapedes, and R. Farber, “How Neural Nets Work,” Neural Net Information Processing Systems, Conference Publication, Denver, CO, American Institute of Physics, NY, pp. 871, ISBN 0-88318-569-5, 1988.
15.
Gerald W. Davis, and Michael L, Gasperi, “ANN Modeling of Volterra Systems,” IEEE, pp. II 727–732, 1991.
16.
E. A. Wan, “Finite Impulse Response Neural Networks with Application in Time-Series Prediction,” Ph.D. dissertation, Dept. of Electrical Engineering, Stanford University, Stanford, CA, Nov. 1993.
17.
J. M. Ortega, Numerical Analyses: A Second Course, Philadelphia: Society for Industrial and Applied Mathematics, pp. 72, 1990.
18.
Vogl
T. P.
,
Mangis
J. K.
,
Rigler
A. K.
,
Zink
W. T.
, and
Alkon
D. L.
, “
Accelerating the Convergence of the Backpropagation Method
,”
Biological Cybernetics
, Vol.
59
, pp.
257
–
263
,
1988
.
19.
van der Ouderaa
E.
,
Schoukens
J.
, and
Renneboog
J.
, “
Peak Factor Minimization, Using Time-Frequency Domain Sapping Algorithm
,”
IEEE Trans. Intrum. and Measurement
, Vol.
37
, No.
1
, pp.
144
–
147
,
1988
.
20.
A. I. Lur’e, “Some Nonlinear Problems in the Theory of Automatic Control,” H. M. Stantionary Office, London, 1957, In Russian 1951.
21.
S. Boyd et al., “Linear Matrix Inequalities in System and Control Theory,” SIAM, 1994, pp. 119–126.
22.
M. Rabinowitz et al., “Training Dynamic Multi-Layer Adaptive Filters,” S96-132. Provisional Filing. U.S. Patent and Trademark Office, 1997.
This content is only available via PDF.
Copyright © 1999
 by The American Society of Mechanical Engineers
You do not currently have access to this content.