基于自适应LMS算法的跨声速风洞模型系统辨识

摘要
图/表
参考文献(26)
相关文章 (15)

全文: PDF (2147 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对风洞试验模型系统辨识不准确的问题，本文利用自适应LMS（Least Mean Square）滤波器模型对跨声速风洞模型进行系统辨识。由于实测信号中存在多模态耦合，为了提高系统辨识精准度，首先对输入输出信号作了FRF(Frequency Response Analysis)分析得到试验模型俯仰方向前两阶模态，其次利用快速Fourier变换进行模态解耦，接着利用自适应LMS滤波器模型、传递函数模型、多项式模型对俯仰方向单模态进行系统辨识，最后得到了基于自适应LMS滤波器模型的俯仰方向一、二阶模态滤波器系数。通过对比不同数学模型的输出与输入之间的相关系数和均方误差及辨识结果，表明自适应LMS滤波器模型具有更高的系统辨识精准度和更简洁的数学模型结构。为后续风洞试验模型振动主动控制计算法的设计提供有力支撑。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	李斌斌1
	寇西平1
	2
	吕彬彬1
	余立1
	杨兴华1
	路波1
	曾开春1

关键词 ：系统辨识, 自适应LMS算法, 快速Fourier变换, 跨声速风洞试验, 主动振动控制

Abstract：Aiming at the low frequency and large vibration of the test model caused by the interference of shock vortex and shock boundary layer during the transonic wind tunnel test, the adaptive LMS filter algorithm developed based on Wiener filter was used to systematically test the test model. After identification, the filter coefficients of the first and second-order modes of vibration in the pitch direction of the test model are obtained, which can be used to design the digital filter of the test model and the active vibration control algorithm. The actual collected wind tunnel test data has multi-modal coupling, and the vibration energy of the model is mainly concentrated in the first two modes in the pitch direction, therefore, in order to accurately identify the first two-order modal transfer functions in the pitch direction, the fast Fourier transform is used to modally decouple the measured vibration signals of the filtered active vibration reduction ground debugging system model. The first-order modal signal containing only the pitch direction and the second-order modal signal only containing the pitch direction are obtained through the decoupling matrix. The first-order and second-order modal signals are identified by adaptive LMS filtering algorithm, and compared with the traditional transfer function model identification algorithm and polynomial model identification algorithm. The results show that the adaptive LMS filtering algorithm is not only simple in structure and easy to implement, but also has higher robustness and identification accuracy. It provides a new theoretical method for the system identification of the wind tunnel model, and provides support for the design of the subsequent active vibration control algorithm.

Key words： System identification Adaptive LMS algorithm Fast Fourier transform Active vibration control;

收稿日期: 2022-10-31 出版日期: 2024-02-15

引用本文:

李斌斌1，寇西平1，2，吕彬彬1，余立1，杨兴华1，路波1，曾开春1. 基于自适应LMS算法的跨声速风洞模型系统辨识[J]. 振动与冲击, 2024, 43(3): 164-170.
LI Binbin1, KOU Xiping1,2, L Binbin1, YU Li1, YANG Xinghua1, LU Bo1, ZENG Kaichun1. Identification of transonic wind tunnel model system based on adaptive LMS algorithm. JOURNAL OF VIBRATION AND SHOCK, 2024, 43(3): 164-170.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2024/V43/I3/164

[1] 曾开春, 寇西平, 杨兴华,等. 跨声速风洞试验模型主动减振结构优化设计[J]. 航空学报:0-0. ZENG Kaichun, KOU Xiping, YANG Xinghua, et al. Optimal Design of Active Damping Structure for Transonic Wind Tunnel Test Model[J]. Acta Aeronautica et Astronautica Sinica, 2022，43（2）：224944. [2] 余立, 杨兴华, 寇西平,等. 跨声速风洞模型主动减振系统试验研究[J]. 南京航空航天大学学报, 2019, 51(4):8. YU Li, YANG Xinghua, KOU Xiping, et al. Experimental Study on Active Damping System of Transonic Wind Tunnel Model [J]. Journal of Nanjing University of Aeronautics and Astronautics, 2019, 51(4):8. [3] 曾开春, 寇西平, 杨兴华,等. 一种控制参数计算方法及装置:, CN109657356A[P]. 2019. ZENG Kaichun, KOU Xiping, YANG Xinghua, et al. A method and device for calculating control parameters:, CN109657356A[P]. 2019. [4] 温正权，刘巍.风洞模型多模态振动主动控制方法研究[D]. 大连，大连理工大学，2020. WEN Zhengquan, LIU Wei. Research on active control method of multi-modal vibration of wind tunnel model [D] Dalian, Dalian University of Technology, 2020. [5] 曾开春, 查俊, 马涛,等. 风洞模型抑振装置及系统:, CN211013451U[P]. 2020. ZENG Kaichun, ZHA Jun, MA Tao, et al. Wind tunnel model vibration suppression device and system:, CN211013451U [P] two thousand and twenty [6] 李昱, 袁磊. 线性系统的系统辨识综述[J]. 探测与控制学报, 2021, 43(3):8. LI Yu, YUAN Lei. Overview of system identification of linear systems [J]. Journal of Detection and Control, 2021, 43 (3): 8. [7] Rivers M B , Balakrishna S . NASA Common Research Model Test Envelope Extension with Active Sting Damping at NTF[C]// AIAA Applied Aerodynamics Conference. 0. [8] Performance of an Active Sting Damper for the NASA Common Research Model[C]// AIAA Aerospace Sciences Meeting Including the New Horizons Forum & Aerospace Exposition. 2013. [9] 杨辉跃,涂亚庆,张海涛,李明.一种基于LMS的振动信号相位差自适应无偏估计方法及应用[J]. 振动与冲击,2016,35(10):55-59. YANG Yuehui, TU Yaqing, ZHANG Haitao, LI Ming. An LMS-based adaptive unbiased estimation method for phase difference of vibration signals and its application [J]. Vibration and shock, 2016,35 (10): 55-59 [10] 姜尔东.支杆式风洞模型振动主动抑制方法研究 [D]. 大连，大连理工大学，2014. JIANG Erdong. Research on active vibration suppression method of strut wind tunnel model [D] Dalian, Dalian University of Technology, 2014. [11] Meng J , Zhang H , Yan Z , et al. A reweighted l 0 -norm-constraint LMS algorithm for sparse system identification[J]. Digital Signal Processing, 2022, 123:103456-. [12] Zhang N , Ni J , Chen J , et al. Steady-State Mean-Square Error Performance Analysis of the Tensor LMS Algorithm[J]. Circuits and Systems II: Express Briefs, IEEE Transactions on, 2020, PP(99):1-1. [13] 王诗彬，朱忠奎等.基于瞬态冲击响应参数辨识的轴承故障特征检测[J].振动工程学报，2010.23（4）：445-449. WANG Shibin, ZHU Zhongkui, et al. Bearing fault feature detection based on transient impact response parameter identification [J]. Journal of Vibration Engineering, 2010.23 (4): 445-449. [14] 李盈颖,万建伟,周良柱.一种改进的变步长归一化LMS算法[J].国防科技大学学报,1999(01):97-99. LI Yingying, WAN Jianwei, ZHOU Liangzhu. An improved variable step-size normalized LMS algorithm [J]. Journal of National Defense University of Science and Technology, 1999 (01): 97-99. [15] Östring M, Gunnarsson S, Norrlöf M. Closed-loop identification of an industrial robot containing flexibilities [J]. Control Engineering Practice, 2003, 11(3): 291-300. [16] Saarakkala S E, Hinkkanen M. Identification of two-mass mechanical systems using torque excitation: Design and experimental evaluation[J]. IEEE Transactions on Industry Applications, 2015, 51(5): 4180-4189. [17] Gui G , Peng W , Adachi F . Adaptive system identification using robust LMS/F algorithm[J]. International Journal of Communication Systems, 2014. [18] Saarakkala S E, Hinkkanen M. Identification of two-mass mechanical systems in closed-loop speed control[C]. Proceedings of the 39th Annual Conference of the IEEE Industrial Electronics Society, 2013: 2905-2910. [19] 张春路，丁国良.电冰箱动态负荷传递函数模型[J].上海交通大学学报，2002,36（2）：188-192. ZHANG Chunlu, DING Guoliang. Refrigerator dynamic load transfer function model [J]. Journal of Shanghai Jiaotong University, 2002,36 (2): 188-192. [20] Yu R , Ying S , Nambiar M . Fast system identification using prominent subspace LMS[J]. Digital Signal Processing, 2014, 27(2):44-56. [21] 李自强,等. 基于非线性自适应滤波算法的齿轮传动系统振动主动控制[C]//中国振动工程学会, 2015. LI Ziqiang, et al. Active vibration control of gear transmission system based on nonlinear adaptive filtering algorithm [C]// China Vibration Engineering Society, 2015 [22] 李竹, 杨培林, 行小帅. 一种改进变步长LMS算法及其在系统辨识中的应用[J]. 仪器仪表学报, 2007, 28(7):5. LI Zhu, YANG Peilin, XING Xiaoshuai. An improved variable step LMS algorithm and its application in system identification [J]. Journal of Instrumentation, 2007, 28 (7): 5. [23] Liu W，Pokharel P P，Principe J C.The kernel Least-Mean-Square algorithm[J]. IEEE Transactions on Signal ProcessingVolume，2008，56（2）：543–554. [24] 朱陈良.基于核方法的自适应滤波的算法的研究[D].西华大学，2011. ZHU Chenliang. Research on adaptive filtering algorithm based on kernel method [D]. Xihua University, 2011. [25] Jian J , Gu Y , Mei S . Adaptive algorithm for sparse system identification: Zero-attracting LMS[J]. Journal of Tsinghua University, 2010, 50(10). [26] 张炳婷, 赵建平, 马淑丽. 新的变步长LMS算法在系统辨识中的应用[J]. 通信技术, 2015, 48(6):4. ZHANG Binting, ZHAO Jianping, MA Shuli. Application of a new variable step LMS algorithm in system identification [J]. Communication Technology, 2015, 48 (6): 4.