Parameter Identification Methods for an Additive Nonlinear System

Jing Chen,Yunxia Ni
DOI: https://doi.org/10.1007/s00034-014-9793-6
2014-01-01
Abstract:This paper deals with parameter identification methods for an additive nonlinear system with a preload nonlinearity and a piece-wise nonlinearity. By using a switching function, we transfer the model of the additive nonlinear system into an identification model, and propose a recursive least squares algorithm and two modified stochastic gradient (SG) algorithms to estimate the parameters of the identification model. The simulation results indicate that the proposed methods converge faster than the SG algorithm.
What problem does this paper attempt to address?