Back
A case based comparison of identification with neural network and Gaussian process models.
In this paper an alternative approach to black-box identification of non-linear dynamic systems is compared with the more established approach of using artificial neural networks. The Gaussian process prior approach is a representative of non-parametric modelling approaches. It was compared on a pH process modelling case study. The purpose of modelling was to use the model for control design. The comparison revealed that even though Gaussian process models can be effectively used for modelling dynamic systems caution has to be axercised when signals are selected.
@inproceedings{2314, title = {A case based comparison of identification with neural network and Gaussian process models.}, journal = {Proceedings of the International Conference on Intelligent Control Systems and Signal Processing ICONS 2003}, abstract = {In this paper an alternative approach to black-box identification of non-linear dynamic systems is compared with the more established approach of using artificial neural networks. The Gaussian process prior approach is a representative of non-parametric modelling approaches. It was compared on a pH process modelling case study. The purpose of modelling was to use the model for control design. The comparison revealed that even though Gaussian process models can be effectively used for modelling dynamic systems caution has to be axercised when signals are selected.}, volume = {1}, pages = {137-142}, editors = {Ruano, E.A.}, organization = {Max-Planck-Gesellschaft}, school = {Biologische Kybernetik}, month = apr, year = {2003}, slug = {2314}, author = {Kocijan, J. and Banko, B. and Likar, B. and Girard, A. and Murray-Smith, R. and Rasmussen, CE.}, month_numeric = {4} }