An experimental validation of the Kohonen clustering approach for PWARX model identification on a semi-batch reactor

Among hybrid systems, the piecewise affine systems are a common class to be identified from input/output data. The work presented in this paper is concerned with the identification of piecewise affine systems using clustering based procedures. In fact, the Kohonen´s Self Organizing Map is used to identify both the parameters of the affine sub-models and the hyperplanes defining the partitions of the state-input regression. The purpose of this paper is to validate the proposed method using real data measured from an olive oil esterification reactor.