Modelling of CSTR by fuzzy clustering

Abstract

A continuous stirred tank reactor (CSTR) characterized by three non-linear ordinary equations is modelled using fuzzy clustering. The construction of Takagi-Sugeno fuzzy model by means of Gustafson-Kessel clustering is discussed in this paper to decompose the input space. The model obtained is readily absorbed by MATLAB for analysis and design. The simulation results proved the feasibility of using fuzzy clustering for modelling CSTR. Simulation modelling provides an effective and powerful approach for capturing and analysing the complex CSTR plant. The validation of the simulation is acknowledged as an integral part of this work.