Relay feedback approaches for the identification of Hammerstein-type nonlinear processes

Abstract

For identification of a Hammerstein-type nonlinear process that consists of a nonlinear static function and a linear dynamic subsystem, the use of a relay feedback test and subsequent triangular input test is investigated. The proposed approach can separate the identification problem of the nonlinear static function from that of the linear dynamic subsystem, which makes it possible to estimate the model parameters without any iterative procedure. From the relay feedback test, the frequency response data of the linear dynamic subsystem are estimated. Then, the model parameters of the nonlinear static function are estimated from the subsequent triangular-type input test. The advantages of the proposed experimental approach are demonstrated with two numerical examples and a simple linearizing control system based on the estimated Hammerstein model.