Hybrid Karhunen-Loeve/neural modelling for a class of distributed parameter systems

Distributed parameter systems (DPS) are a class of infinite dimensional systems. However implemental control design requires low-order models. This work will focus on developing a low-order model for a class of quasi-linear parabolic distributed parameter system with unknown linear spatial operator, unknown linear boundary condition as well as unknown non-linearity. The Karhunen-Loeve (KL) Empirical Eigenfunctions (EEFs) are used as basis functions in Galerkin's method to reduce the Partial Differential Equation (PDE) system to a nonlinear low-order Ordinary Differential Equation (ODE) system. Since the states of the system are not measurable, a recurrent Radial Basis Function (RBF) Neural Network (NN) observer is designed to estimate the states and approximate unknown dynamics simultaneously. Using the estimated states, a hybrid General Regression Neural Network (GRNN) is trained to be a nonlinear offline model, which is suitable for traditional control techniques. The simulations demonstrate the effectiveness of this modeling method.

[1]  P. Christofides,et al.  Nonlinear and Robust Control of PDE Systems: Methods and Applications to Transport-Reaction Processes , 2002 .

[2]  Costas J. Spanos,et al.  Advanced process control , 1989 .

[3]  Guanrong Chen,et al.  Spectral-approximation-based intelligent modeling for distributed thermal processes , 2005, IEEE Transactions on Control Systems Technology.

[4]  Panagiotis D. Christofides,et al.  Robust output feedback control of quasi-linear parabolic PDE systems , 1998, Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207).

[5]  Karlene A. Hoo,et al.  System identification and model-based control for distributed parameter systems , 2004, Comput. Chem. Eng..

[6]  Andrew J. Newman,et al.  Model Reduction via the Karhunen-Loeve Expansion Part I: An Exposition , 1996 .

[7]  D. Schroder.  Intelligent modelling, observation and control for nonlinear systems , 2000, 6th International Workshop on Advanced Motion Control. Proceedings (Cat. No.00TH8494).

[8]  Zhihong Man,et al.  Lyapunov-theory-based radial basis function networks for adaptive filtering , 2002 .

[9]  K. Hoo,et al.  Low-Order Model Identification of Distributed Parameter Systems by a Combination of Singular Value Decomposition and the Karhunen−Loève Expansion , 2002 .

[10]  Panagiotis D. Christofides,et al.  Robust Control of Parabolic PDE Systems , 1998 .

[11]  Dierk Schröder,et al.  Intelligent modeling, observation, and control for nonlinear systems , 2001 .

[12]  Karlene A. Hoo,et al.  Low-order model identification for implementable control solutions of distributed parameter systems , 2002 .

[13]  W. Ray,et al.  Identification and control of distributed parameter systems by means of the singular value decomposition , 1995 .

[14]  Karlene A. Hoo,et al.  Low-order Control-relevant Models for A Class of Distributed Parameter Systems , 2001 .

[15]  Stephen A. Billings,et al.  Identification of finite dimensional models of infinite dimensional dynamical systems , 2002, Autom..

[16]  N. Smaoui,et al.  Modelling the dynamics of nonlinear partial differential equations using neural networks , 2004 .

[17]  R.A. Sahan,et al.  Artificial neural network-based modeling and intelligent control of transitional flows , 1997, Proceedings of the 1997 IEEE International Conference on Control Applications.

[18]  H. Park,et al.  The use of the Karhunen-Loève decomposition for the modeling of distributed parameter systems , 1996 .