Local Model Architectures for Nonlinear Modelling and Control
Local Model Networks are learning systems which are able to model and control unknown nonlinear dynamic processes from their observed input-output behaviour. Simple, locally accurate models are used to represent a globally complex process. The framework supports the modelling process in real applications better than most artificial neural network architectures. This paper shows how their structure also allows them to more easily integrate knowledge, methods and a priori models from other paradigms such as fuzzy logic, system identification and statistics. Algorithms for automatic parameter estimation and model structure identification are given.
Local Models intuitively lend themselves to the use of Local Controllers, where the global controller is composed of a combination of simple locally accurate control laws. A Local Controller Network (LCN) for controlling the lateral deviation of a car on a straight road is demonstrated.
KeywordsBasis Function Local Model Input Space Multivariate Adaptive Regression Spline Local Controller
Unable to display preview. Download preview PDF.
- 3.R. Shorten and R. Murray-Smith, “On Normalising Basis Function networks,” in 4th Irish Neural Networks Conf., Univ. College Dublin, Sept. 1994.Google Scholar
- 8.M. Brown and C. Harris, Neurofuzzy Adaptive Modelling and Control. Hemel-Hempstead, UK: Prentice Hall, 1994.Google Scholar
- 9.R. Murray-Smith, A Local Model Network Approach to Nonlinear Modelling. Ph.D. Thesis, Department of Computer Science, University of Strathclyde, Glasgow, Scotland, Nov. 1994. E-mail:murray@DBresearch-berlin.de.Google Scholar
- 11.R. Murray-Smith and H. Gollee, “A constructive learning algorithm for local model networks,” in Proc. IEEE Workshop on Computer-intensive methods in control and signal processing, Prague, Czech Republic, pp. 21–29, 1994. E-mail:murray@DBresearch-berlin.de.Google Scholar
- 12.R. Murray-Smith, “A Fractal Radial Basis Function network for modelling,” in Inter. Conf. on Automation, Robotics and Computer Vision, Singapore, vol. 1, pp. NW-2.6.1–NW-2.6.5, 1992. E-mail:murray@DBresearch-berlin.de.Google Scholar
- 14.R. Żbikowski, K. J. Hunt, A. Dzieliński, R. Murray-Smith, and P. J. Gawthrop, “A review of advances in neural adaptive control systems,” Technical Report of the ESPRIT NACT Project TP-1, Glasgow University and Daimler-Benz Research, 1994.Google Scholar
- 15.J. S. Shamma and M. Athans, “Gain scheduling: Potential hazards and possible remedies,” IEEE Control Systems Magazine, pp. 101–107, June 1992.Google Scholar
- 19.K. J. Hunt, R. Haas, and M. Brown, “On the functional equivalence of fuzzy inference systems and spline-based networks,” International Journal of Neural Systems, 1995. To appear — March issue.Google Scholar
- 21.M. Brown and C. Harris, Neurofuzzy Adaptive Modelling and Control. Hemel-Hempstead, UK: Prentice Hall, 1994.Google Scholar
- 23.T. A. Johansen, “Adaptive control of MIMO non-linear systems using local ARX models and interpolation,” in IFAC ADCHEM 94, (Kyoto, Japan), May 1994.Google Scholar
- 24.H. Wang, M. Brown, and C. Harris, “Modelling and control of nonlinear, operating point dependent systems via associative memory networks,” J. of Dynamics and Control, accepted for publication 1994.Google Scholar
- 25.J. Ackermann, Robuste Regelung. Berlin: Springer-Verlag, 1993.Google Scholar
- 26.K. Mecklenburg, T. Hrycej, U. Franke, and H. Fritz, “Neural control of autonomous vehicles,” in Proc. IEEE Vehicular Technology Conference, Denver, USA, 1992.Google Scholar