| 注册
首页|期刊导航|湖南大学学报(自然科学版)|最优过程噪声协方差Q估计下的结构响应重构

最优过程噪声协方差Q估计下的结构响应重构

彭珍瑞 车佳亮

湖南大学学报(自然科学版)2025,Vol.52Issue(10):31-42,12.
湖南大学学报(自然科学版)2025,Vol.52Issue(10):31-42,12.DOI:10.16339/j.cnki.hdxbzkb.2025203

最优过程噪声协方差Q估计下的结构响应重构

Structural Response Reconstruction under Optimal Process Noise Covariance Q Estimation

彭珍瑞 1车佳亮1

作者信息

  • 1. 兰州交通大学 机电工程学院,甘肃 兰州 730070
  • 折叠

摘要

Abstract

To address the challenge of obtaining process noise covariance in Kalman filters for structural response reconstruction,a two-stage optimal Q estimation(TOQE)algorithm was proposed.Under a fixed window length,a minimization error function concerning Q was established,and the search interval was derived based on the innovation theory.A one-dimensional two-stage search strategy was employed to accelerate the search for the process noise covariance Q.The Savitzky-Golay(SG)smoothing algorithm was used to smooth the estimated process noise covariance between windows,achieving the optimal process noise covariance estimation.Finally,the Kalman filters was used to reconstruct the structural response.The effectiveness of the algorithm was validated through simulations of a vertical axis wind turbine and experiments on a cantilever beam.The reconstructed acceleration response was applied to identify the excitation of the cantilever beam structure under L1/L2 regularization.The results show that compared to the traditional optimal Q estimation(OQE)algorithm for response reconstruction,the TOQE algorithm can determine the search interval for one-dimensional search under 5%,10%,and 15%measurement noise,respectively,avoiding the subjective selection of OQE search interval and achieving optimal process noise covariance estimation.The search efficiency was improved by 50%~60%.Under different accuracy evaluation standards,the response reconstruction error of the TOQE algorithm is reduced,and the reconstructed acceleration response can better identify the external excitation of the cantilever beam structure.The TOQE algorithm demonstrates good real-time performance and robustness.

关键词

过程噪声协方差/响应重构/Savitzky-Golay/卡尔曼滤波/激励识别

Key words

process noise covariance/response reconstruction/Savitzky-Golay/Kalman filters/excitation iden-tification

分类

机械工程

引用本文复制引用

彭珍瑞,车佳亮..最优过程噪声协方差Q估计下的结构响应重构[J].湖南大学学报(自然科学版),2025,52(10):31-42,12.

基金项目

国家自然科学基金资助项目(62161018),National Natural Science Foundation of China(62161018) (62161018)

湖南大学学报(自然科学版)

OA北大核心

1674-2974

访问量0
|
下载量0
段落导航相关论文