Gradient-based iterative parameter estimation for bilinear-in-parameter systems using the model decomposition technique

作     者：Chen, Mengting Ding, Feng Yang, Erfu 

作者机构：Jiangnan Univ Lab Adv Proc Control Light Ind Minist Educ Sch Internet Things Engn Wuxi 214122 Peoples R China Qingdao Univ Sci & Technol Coll Automat & Elect Engn Qingdao 266061 Peoples R China Univ Strathclyde Space Mechatron Syst Technol Lab Glasgow G1 1XJ Lanark Scotland 

出 版 物：《IET CONTROL THEORY AND APPLICATIONS》 (IET控制论与应用)

年 卷 期：2018年第12卷第17期

页      面：2380-2389页

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 0804[工学-仪器科学与技术] 0811[工学-控制科学与工程] 

基　　金：National Natural Science Foundation of China Taishan Scholar Project Fund of Shandong Province of China [ts20130939] 111 Project [B12018] National First-Class Discipline Program of Light Industry Technology and Engineering [LITE2018-26] 

主　　题：parameter estimation gradient methods nonlinear control systems bilinear systems stochastic gradient algorithms iterative algorithm bilinear-in-parameter system model decomposition technique parameter estimation issues block-oriented nonlinear system bilinear-in-parameter model nonlinear block noise model gradient-based iterative parameter estimation 

摘      要：The parameter estimation issues of a block-oriented non-linear system that is bilinear in the parameters are studied, i.e. the bilinear-in-parameter system. Using the model decomposition technique, the bilinear-in-parameter model is decomposed into two fictitious submodels: one containing the unknown parameters in the non-linear block and the other containing the unknown parameters in the linear dynamic one and the noise model. Then a gradient-based iterative algorithm is proposed to estimate all the unknown parameters by formulating and minimising two criterion functions. The stochastic gradient algorithms are provided for comparison. The simulation results indicate that the proposed iterative algorithm can give higher parameter estimation accuracy than the stochastic gradient algorithms.