A Kohonen Neural Network based method for PWARX identification

摘要

In this paper, we address the problem of identifying PWARX models. The identification of PWARX models involves both the estimation of the parameters of the affine sub-models and the hyperplanes defining the partition of the state-input regression. Only the clustering based methods are considered. The performance of these methods depends on the used clustering technique that the majority of them are sensible to poor initialization and suffer from the presence of outliers. To overcome these problems, we propose to exploit the Kohonen neural network approach. Simulation results are presented to illustrate the performance of the proposed method.