摘要
针对传统方法难以有效将非线性振动信号从复杂强干扰中提取的难题,提出了奇异值分解(SVD)和集合经验模态分解(EEMD)的降噪方法.该方法利用EEMD叠加白噪声预处理的过程,抑制脉冲噪声的影响并克服了EMD模式混叠效应,然后提取信号的趋势项,克服了信号趋势项对SVD选择奇异值的影响,最后将SVD方法降噪后的信号与趋势项叠加达到降噪目的,实现SVD和EEMD的优势互补,提高降噪效果.对模拟信号和实测非线性振动信号进行了仿真试验研究,结果表明,该方法可以同时有效地抑制非线性振动信号中的白噪声和脉冲噪声,对工程实际信号的进一步分析处理提供有效的预处理手段.
In order to extract nonlinear vibration signals from complex strong interference effectively,a de-noising method based on singular value decomposition and ensemble empirical mode decomposition was proposed.Firstly,the superposing white noise process of EEMD was used to suppress the impulse noise and pattern aliasing effect of EMD.Then the trend of the signal was extracted to overcome the problem of hard to select the singular value.Finally,the de-noised signal based on SVD and the trend were rebuilt to realize the purpose of noise reduction.The proposed method realized the complementary advantages of SVD and EEMD and improve the effect of noise reduction.The simulation and measured nonlinear vibration signals were simulated and studied.The results show that the proposed method could suppress the white noise and impulse noise in nonlinear vibration signals effectively,and lay a foundation for further analysis and processing of actual engineering signals.
作者
刘树聃
陈知行
LIU Shudan;CHEN Zhixing(Xuchang Gengxin Information Science Research Institute,Xuchang 461000,China;Aviation Engineering Institute,Xuchang Vocational Technical College,Xuchang 461000,China;School of Automation,Beijing Institute of Technology,Beijing 100081,China)
出处
《探测与控制学报》
CSCD
北大核心
2019年第3期37-42,共6页
Journal of Detection & Control
基金
北京市自然科学基金项目资助(1183027)
河南省教改重点项目资助([2015]061号)
作者简介
刘树聃(1974—),女,河南许昌人,硕士,副教授,研究方向:计算机应用技术。E-mail:xctax@163.com。
