非线性输入输出系统参数辨识的高阶最小二乘法  

High-order Least Squares Method for Parameter Identification of Nonlinear Input and Output Systems

作　　者：李雅会 文成林 郑锦辉 LI Yahui;WEN Chenglin;ZHENG Jinhui(School of Information and Control Engineering,Jilin Institute of Chemical Technology,Jilin132022,China;Guangdong Provincial Key Lab of Robotics and Intelligent System,Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,Shenzhen 518055,China;School of Automation,Guangdong University of Petrochemical Technology,Maoming 525000,China)

机构地区：[1]吉林化工学院信息与控制工程学院,吉林132022 [2]广东省机器人与智能系统重点实验室,中国科学院深圳先进技术研究院,广东深圳518055 [3]广东石油化工学院自动化学院,广东茂名510006

出　　处：《广东石油化工学院学报》2024年第1期70-75,共6页Journal of Guangdong University of Petrochemical Technology

基　　金：国家自然科学基金(61933013);国家自然科学基金国家杰出青年科学基金(62125307)。

摘　　要：传统的最小二乘方法在对非线性模型中的参数进行辨识时,通常采用对非线性模型的一阶线性求解,在求解过程中会引入误差,从而造成待求解系统模型参数的发散。当系统模型的非线性越强时,待辨识的模型参数发散就越严重,从而使待辨识的模型参数的估计精度不精确。文章通过采用对目标函数进行高阶展开,以减少引入误差来提高系统模型待辨识参数的精度。数值仿真结果,证实了所提方法的有效性与可行性。The traditional least squares method usually uses the first-order linear solution of the nonlinear model when identifying the parameters in the nonlinear model,which will introduce errors in the process,resulting in the divergence of the model parameters of the system to be solved.When the nonlinearity of the system model is stronger,the divergence of the model parameters to be identified will be more serious.Thus,the estimation accuracy of the model parameters to be identified is inaccurate.In this study,the high-order expansion of the objective function is adopted to reduce the introduction error and improve the accuracy of the parameters to be identified in the system model.Finally,numerical simulations confirm the effectiveness and feasibility of the proposed method.

关 键 词：最小二乘法 泰勒展开 高阶项 参数辨识 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]