哈尔滨工程大学学报2025，Vol.46Issue(10)：2032-2039,8.DOI:10.11990/jheu.202406016
时滞协方差参考独立分量分析在振源识别中的应用
Application of the time delay covariance integrated with reference independent component analysis to vibration source identification
摘要
Abstract
Numerous excitation sources are involved in mechanical systems,and determining the main vibration source is difficult.Thus,the method combining the time-delay covariance matrix and reference-independent compo-nent analysis is applied for excitation identification to effectively address the high dependency of the reference-inde-pendent component analysis algorithm on the reference signal.The fundamental period of the signal to be extracted is applied as the time delay to initially separate the vibration source signal.This result is utilized as the reference signal to further separate the vibration source.While enhancing the accuracy and stability of the algorithm,it can also accurately correspond to the sequence of the separated signal and the vibration source.Then,the contribution is estimated based on the separation result to determine the main vibration source.Finally,the accuracy of the al-gorithm is validated using numerical simulation examples and data from the vibration test bench.The results show that the error between the contribution of the vibration source gained by the algorithm and the actual contribution does not exceed 5%.Therefore,the proposed method can effectively determine the primary vibration source and provide a reliable technical approach for accurately determining the excitation characteristics of the mechanical e-quipment system and quantifying the contribution of the excitation source.关键词
载荷识别/振动/声隐身/时滞协方差/独立分量分析/参考信号/贡献量估计/盲源分离Key words
load identification/vibration/acoustic stealth/time-delay covariance/independent component analy-sis/reference signal/contribution estimation/blind source separation分类
机械工程引用本文复制引用
李文晴,叶天贵,陈玉坤,石利权,刘超,靳国永..时滞协方差参考独立分量分析在振源识别中的应用[J].哈尔滨工程大学学报,2025,46(10):2032-2039,8.基金项目
国家自然科学基金项目(52225109,52241101,52271309). （52225109,52241101,52271309）
