A Novel Sequential Learning Algorithm for RBF Networks and Its Application to Dynamic System Identification

This paper presents a novel sequential learning algorithm for radial basis function (RBF) networks referred to as dynamic orthogonal structure adaptation (DOSA) algorithm. The algorithm enables the RBF network to on-line adjust its structure and weights to the identified dynamics with a compact network structure. It makes use of the well-known idea of error reduction ratio in orthogonal least squares (OLS) method for network pruning, and lakes advantage of a sliding data window for monitoring system dynamics. Simulation results of nonlinear dynamic system identification demonstrate the adaptive tracking ability and high learning speed of the proposed algorithm.

[1]  Visakan Kadirkamanathan,et al.  A Function Estimation Approach to Sequential Learning with Neural Networks , 1993, Neural Computation.

[2]  Shang-Liang Chen,et al.  Orthogonal least squares learning algorithm for radial basis function networks , 1991, IEEE Trans. Neural Networks.

[3]  Yan Li,et al.  Analysis of minimal radial basis function network algorithm for real-time identification of nonlinear dynamic systems , 2000 .

[4]  Cheng-Liang Chen,et al.  Hybrid learning algorithm for Gaussian potential function networks , 1993 .

[5]  John C. Platt.  A Resource-Allocating Network for Function Interpolation , 1991, Neural Computation.

[6]  Narasimhan Sundararajan,et al.  An efficient sequential learning algorithm for growing and pruning RBF (GAP-RBF) networks , 2004, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[7]  D. Broomhead,et al.  Radial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks , 1988 .

[8]  Sheng Chen,et al.  Recursive hybrid algorithm for non-linear system identification using radial basis function networks , 1992 .

[9]  Narasimhan Sundararajan,et al.  Identification of time-varying nonlinear systems using minimal radial basis function neural networks , 1997 .

[10]  Y Lu,et al.  A Sequential Learning Scheme for Function Approximation Using Minimal Radial Basis Function Neural Networks , 1997, Neural Computation.

[11]  Narasimhan Sundararajan,et al.  A generalized growing and pruning RBF (GGAP-RBF) neural network for function approximation , 2005, IEEE Transactions on Neural Networks.

[12]  Rosalind W. Picard,et al.  On the efficiency of the orthogonal least squares training method for radial basis function networks , 1996, IEEE Trans. Neural Networks.

[13]  James D. Keeler,et al.  Layered Neural Networks with Gaussian Hidden Units as Universal Approximations , 1990, Neural Computation.