Title :
Fuzzy Model Identification of a Biological Process Based on Input-Output Data Clustering
Author :
Grisales, Victor Hugo ; Gauthier, Alain ; Roux, Gilles
Author_Institution :
Lab. LAAS-CNRS, Univ. Paul Sabatier, Toulouse
Abstract :
This paper deals with the application of fuzzy clustering techniques for the model identification of an enhanced FAMIMO biological wastewater treatment process from input-output data. This MIMO system is represented as a set of coupled MISO models of the Takagi-Sugeno type. A comparative study with two clustering algorithms for the construction of the fuzzy model is carried out. The Gustafson-Kessel (GK) algorithm and the so-called robust parallel competitive agglomerative (RPCA) algorithm are considered. From a biotechnological point of view, different simulation experiences were conducted integrating both continuous and batch modes in order to validate the obtained models. Results are reported and discussed
Keywords :
MIMO systems; fuzzy control; fuzzy set theory; nonlinear systems; parallel algorithms; pattern clustering; robust control; wastewater treatment; FAMIMO biological wastewater treatment process; Gustafson-Kessel algorithm; MIMO system; MISO models; Takagi-Sugeno type models; batch mode; biological process; biotechnological processes; continuous mode; fuzzy clustering technique; fuzzy model identification; input-output data clustering; nonlinear identification; robust parallel competitive agglomerative algorithm; Biological control systems; Biological processes; Biological system modeling; Clustering algorithms; Laboratories; MIMO; Mathematical model; Robustness; Takagi-Sugeno model; Wastewater treatment;
Conference_Titel :
Fuzzy Systems, 2005. FUZZ '05. The 14th IEEE International Conference on
Conference_Location :
Reno, NV
Print_ISBN :
0-7803-9159-4
DOI :
10.1109/FUZZY.2005.1452518
