基于在线参数辨识的电动助力转向补偿控制

刘涛 $ ^{1} $，严周栋 $ ^{1} $，杨娜 $ ^{1} $，郭桐 $ ^{2} $

（1. 哈尔滨工业大学（威海）汽车工程学院，山东威海 264209；2. 北京中卓时代消防装备科技有限公司，北京 101300）

摘要

电动助力转向(electric power steering, EPS)系统具有非线性和时变性, 采用常系数补偿无法实现对转矩的准确跟踪, 影响驾驶员手感。文章采用滑模控制器准确跟踪电流, 并设计补偿算法, 利用带遗忘因子的递推最小二乘(recursive least squares, RLS)算法对助力装置进行在线参数辨识, 并将辨识得到的结果进行补偿控制, 在参数缓慢变化的条件下实现 EPS 对转矩的准确跟踪。

关键词

电动助力转向(EPS)；补偿控制；递推最小二乘(RLS)参数辨识；滑模控制器；遗忘因子

中图分类号：U463.4

文献标志码：A

文章编号：1003-5060（2023）02-0152-06

Compensation control of electric power steering based on online parameter identification

LIU Tao $ ^{1} $, YAN Zhoudong $ ^{1} $, YANG Na $ ^{1} $, GUO Tong $ ^{2} $

(Weihai), Weihai 264209, China; 2. Beijing Zhongzhuo Fire Fighting Equipment Co., Ltd., Beijing 101300, China)

Abstract

Electric power steering (EPS) system is nonlinear and time-varying. Constant coefficient compensation cannot track the torque accurately, which affects the driver's hand feeling. In this paper, the sliding mode controller is used to track the current accurately, and the compensation algorithm is designed. The recursive least squares (RLS) algorithm with forgetting factor is used to identify the parameters of the power assist device online, and the identified results are compensated. The EPS can accurately track the torque when the parameters change slowly.

Keywords

electric power steering(EPS); compensation control; recursive least squares(RLS) parameter identification; sliding mode controller; forgetting factor

收稿日期：2021-05-20

修回日期：2021-07-16

基金项目：国家自然科学基金青年科学基金资助项目（51905121）