A Localized Forgetting Method for Gaussian RBFN Model Adaptation
- 42 Downloads
In this paper a localized forgetting method is proposed for on-line adaptation of Gaussian radial basis function network models. It is realised that the commonly used exponential forgetting applies to the past data from the entire operating space uniformly and therefore, is not correct for nonlinear systems where dynamics are different in different operating regions. The new method proposed in this paper sets different forgetting factors in different regions according to the response of the local centre to the current measurement data. The method is applied in conjunction with the recursive orthogonal Least Squares algorithm and the computing is consequently very efficient. The developed method is applied to modelling of dissolved oxygen in a chemical reactor rig. It shows a smaller mean squared error for one-step-ahead prediction than using the uniform forgetting.
Unable to display preview. Download preview PDF.
- 1.Chen, S., Cowan, C. F. N. and Grant, P. M.: Orthogonal least squares learning algorithm for radial basis function networks, IEEE Transaction on Neural Networks, 2 (1991), 302-309.Google Scholar
- 2.Karayiannis, N. B. and Mi, G. W.: Growing radial basis networks: merging supervised and unsupervised learning with network growth techniques, IEEE Transaction on Neural Networks, 8(6) (1997), 1492-1506.Google Scholar
- 3.Platt, J.: A resource allocating network for function interpolation, Neural Computation, 4(2) (1991), 213-225.Google Scholar
- 4.Liu, G. P., Kadirkamanathan, V. and Billings, S. A.: Variable neural networks for adaptive control of nonlinear systems, IEEE Trans. on Systems, Man, and Cybernetics_Part C: Applications and Reviews, 29(1), (1999) 34-43.Google Scholar
- 5.Yu, D. L., Gomm J. B.and Williams D.: A recursive orthogonal least squares algorithm for training RBF networks, Neural Processing Letters, 5(3) (1997), 167-176.Google Scholar
- 6.Gomm, J. B. and Yu D. L.: Selecting radial basis function network centres with recursive orthogonal least squares training, IEEE Transaction on Neural Networks, 11(3) (2000), 306-314.Google Scholar
- 7.Yu, D. W. and Yu, D. L.: Development of an Adaptive RBF Model for Process Modelling, Technical report, School of Engineering, Liverpool John Moores University, (2002).Google Scholar