Online fuzzy identification for an intelligent controller based on a simple platform

  • Authors:
  • Sašo Blaič;Igor Škrjanc;Samo Gerkšič;Gregor Dolanc;Stanko Strmčnik;Mincho B. Hadjiski;Anna Stathaki

  • Affiliations:
  • Faculty of Electrical Engineering, University of Ljubljana, Traška 25, 1000 Ljubljana, Slovenia;Faculty of Electrical Engineering, University of Ljubljana, Traška 25, 1000 Ljubljana, Slovenia;Joef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia;Joef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia;Joef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia;UCTM Sofia, Kliment Ohridski Blvd. 8, Sofia, Bulgaria;Computer Technology Institute, Akteou and Poulopoulou St. 11, Athens, Greece

  • Venue:
  • Engineering Applications of Artificial Intelligence
  • Year:
  • 2009

Quantified Score

Hi-index 0.01

Visualization

Abstract

The paper presents the identification issues of the self-tuning nonlinear controller ASPECT (Advanced control algorithmS for ProgrammablE logiC conTrollers). The controller is implemented on a simple PLC platform with an extra mathematical coprocessor, but is intended for the advanced control of complex processes. The model of the controlled plant is obtained by means of experimental modelling. A special batch-wise algorithm that is based on the Takagi-Sugeno model and uses ''fuzzy instrumental variables'' technique is described in the paper. Many robustness problems of the classical adaptive approaches can be circumvented to some extent by the proposed batch-wise approach combined with a supervisory mechanism. The paper also includes some experimental results on the hydraulic pilot plant and some simulation case studies.