A hybrid clustering and gradient descent approach for fuzzymodeling

摘要

In this paper, a hybrid clustering and gradient descent approachnis proposed for automatically constructing a multi-input fuzzy modelnwhere only the input-output data of the identified system are available.nThe proposed approach is composed of two steps: structure identificationnand parameter identification. In the process of structurenidentification, a clustering method is proposed to provide a systematicnprocedure to determine the number of fuzzy rules and construct anninitial fuzzy model from the given input-output data. In the process ofnparameter identification, the gradient descent method is used to tunenthe parameters of the constructed fuzzy model to obtain a more precisenfuzzy model from the given input-output data. Finally, two examples ofnnonlinear system are given to illustrate the effectiveness of thenproposed approach