东北大学学报（自然科学版）

东北大学学报:自然科学版 ›› 2019, Vol. 40 ›› Issue (10): 1474-1479.DOI: 10.12068/j.issn.1005-3026.2019.10.019

• 机械工程 • 上一篇    下一篇

基于Newmark-β的ICAA正则化参数选取方法

范玉川1， 陈晔2， 赵春雨1， 卢泽宸1   

  1. (1. 东北大学 机械工程与自动化学院， 辽宁 沈阳110819; 2. 辽宁工业大学 机械工程与自动化学院， 辽宁 锦州121001)
  • 收稿日期:2018-12-24 修回日期:2018-12-24 出版日期:2019-10-15 发布日期:2019-10-10
  • 通讯作者: 范玉川
  • 作者简介:范玉川(1988-)，男，河南新乡人，东北大学博士研究生； 赵春雨(1963-)，男，辽宁黑山人，东北大学教授，博士生导师 .
  • 基金资助:
    国家自然科学基金资助项目(51775094).

Regularization Parameter Determination Method of ICAA Based on Newmark-β

FAN Yu-chuan1， CHEN Ye2， ZHAO Chun-yu1， LU Ze-chen1   

  1. 1. School of Mechanical Engineering & Automation， Northeastern University， Shenyang 110819， China; 2. School of Mechanical Engineering and Automation， Liaoning University of Technology， Jinzhou 121001， China.
  • Received:2018-12-24 Revised:2018-12-24 Online:2019-10-15 Published:2019-10-10
  • Contact: ZHAO Chun-yu
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

743

摘要： 传统的L曲线法在使用的时候常常不容易获得准确的正则化参数，基于此，提出了一种基于Newmark-β的反算-对比-调整-逼近(inverse computation-contrast-adjustment-approach， ICAA)正则化参数选取方法.该算法相比传统的L曲线法使用起来更加直观、简便，并且计算耗费的时间更短、效率更高.通过一个四自由度系统的仿真算例和一个悬臂梁的实验验证了本算法的有效性，并把本算法的载荷识别结果与L曲线法的载荷识别结果进行了对比.结果表明:该算法相比L曲线法不仅在计算效率方面有显著优势，而且利用前者的正则化参数进行载荷识别，计算精度更高.

关键词: 正则化参数, L曲线法, 载荷识别, Newmark-β, 计算效率

Abstract: When the traditional L-curve method is applied， it is often not easy to obtain accurate regularization parameters. Therefore， a new regularization parameter determination method of ICAA(inverse computation-contrast-adjustment-approach)based on Newmark-β was proposed. Compared with the traditional L-curve method， the algorithm is more intuitive and simple to use， and the calculation time is shorter and more efficient. The validity of the algorithm was verified by a simulation example of a four-DOF system and an experiment of a cantilever beam， and the load identification results of this algorithm were compared with the load recognition results of the L-curve method. The results showed that compared with the L-curve method， the proposed algorithm not only has greater advantages in computational efficiency， but also has higher accuracy in load identification calculation using the regularization parameters obtained by the algorithm.

Key words: regularization parameter, L-curve method, load recognition identification, Newmark-β, computational efficiency

中图分类号: