Iterative Parameter Estimation For Nonlinear Systems With Input Saturation Nonlinearities

In industrial applications and control,it is difficult for linear systems to accurately describe the characteristics of industrial processes due to factors such as equipment characteristics and production loss.The saturation nonlinearities and the deadzone nonlinearities are widely encountered in engineering applications.The identification of these nonlinear systems is a research hotspot in the field of the industrial control and has important theoretical significance.This paper investigates the identification problem of an equationerror nonlinear system with saturation nonlinearity and an outputerror nonlinear system with saturation and deadzone nonlinearity.The main research contents are as follows.1.Introducing a switching function,the identification model of the equationerror nonlinear system with saturation nonlinearity is established.Using the least squares identification principle and the negative gradient search identification principle,an overparameterization leastsquaresbased iterative(OLSI)algorithm and an overparameterization gradientbased iterative(OGI)algorithm are presented to estimate the parameter products of the nonlinear system based on the overparameterization identification model of the nonlinear system with saturation nonlinearity.2.Introducing the negative gradient search identification principle,a gradientbased iterative(GI)algorithm is developed to identify directly the parameters of the nonlinear system for reducing computational loads.In order to enhance the computational efficiency,the hierarchical identification model of the nonlinear system with saturation nonlinearity is established.Based on the hierarchical identification model,a hierarchical gradientbased iterative(HGI)algorithm and a hierarchical leastsquaresbased iterative(HLSI)algorithm are proposed by using the least squares identification principle and the negative gradient search identification principle.3.Introducing a symbolic function,the identification model of the outputerror nonlinear system with saturation and deadzone nonlinearity is established.Based on the overparameterization identification model and the auxiliary model identification idea,some overparameterization auxiliary model identification methods are developed to estimate the parameters of the nonlinear system with the unknown intermediate variables.Using the hierarchical identification principle,some hierarchical auxiliary model identification methods are derived for reducing computational loads.This paper gives the derivation processes and calculation steps of the iterative algorithms proposed above.The numerical simulation results illustrate the effectiveness of the proposed algorithms,and the computational loads of these algorithms are analyzed and compared by giving some examples.At the end of the paper,some concluding remarks of this project are made.