Parameter Estimation in Nonlinear Systems Using Hopfield Neural Networks

A method using the Hopfield neural network is developed for estimating the parameters of a nonlinear system whose theoretical model is assumed to exist. A linearization procedure is presented, and the errors between the dynamics of the plant and its model are minimized through a cost function that is equated to the energy function of a Hopfield neural network. The minimization process yields the weights and biases of the neural network. Proof of convergence of the modeled parameters to their true values and boundedness of parameter estimates at each step are provided. Numerical results from a scalar time-varying problem and a complex nine-state aircraft problem are presented to demonstrate the potential of this method

[1]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[2]  J. R. Raol,et al.  Neural network based parameter estimation of unstable aerospace dynamic systems , 1994 .

[3]  Sergey Edward Lyshevski,et al.  Nonlinear identification of control systems , 1998, Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207).

[4]  J. R. Raol.  Parameter estimation of state space models by recurrent neural networks , 1995 .

[5]  A. Cichocki,et al.  Neural networks for solving systems of linear equations and related problems , 1992 .

[6]  Frank L. Lewis,et al.  A dynamic recurrent neural-network-based adaptive observer for a class of nonlinear systems , 1997, Autom..

[7]  F. De Bruyne.  Closed-loop relevant identification of input-output and noise dynamics , 1998 .

[8]  Kevin Warwick,et al.  Inverting recurrent neural networks for internal model control of nonlinear systems , 1998, Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207).

[9]  Allen R. Stubberud,et al.  A neural-network-based system identification technique , 1991, [1991] Proceedings of the 30th IEEE Conference on Decision and Control.

[10]  S. Lyashevskiy.  Nonlinear error mapping technology in system identification , 1997, Proceedings of the 1997 American Control Conference (Cat. No.97CH36041).

[11]  J J Hopfield,et al.  Neurons with graded response have collective computational properties like those of two-state neurons. , 1984, Proceedings of the National Academy of Sciences of the United States of America.

[12]  J. R. Raol,et al.  Neural network architectures for parameter estimation of dynamical systems , 1996 .

[13]  J. J. Hopfield,et al.  “Neural” computation of decisions in optimization problems , 1985, Biological Cybernetics.

[14]  Sabri Cetinkunt,et al.  Estimation of model parameters of linear structural systems using Hopfield neural networks , 1994 .