项目名称: 瞬态冲击信号的Hilbert时频谱表征方法与应用研究
项目编号: No.50805014
项目类型: 青年科学基金项目
立项/批准年度: 2009
项目学科: 电工技术
项目作者: 李宏坤
作者单位: 大连理工大学
项目金额: 20万元
中文摘要: 瞬态冲击信号包含大量反映设备运行特征的信息,但由于信号的非平稳和非线性性质、实测信号大量的噪声干扰,以及实际分析的瞬态冲击信号为短序列等原因,瞬态冲击信号的分析方法亟待提高。在对Hilbert 时频谱表征瞬态冲击信号的影响因素研究基础上,结合机械设备瞬态冲击振动信号的特点,本项目对Hilbert 时频谱表征瞬态冲击信号的方法进行优化研究:1、采用多项式相位估计法计算经验模拟式分解得到内蕴模式分量的瞬时频率,提高瞬时频率计算的实效性;2、建立Hilbert时频谱表征信号的性能评价模型,研究Hilbert 时频谱最佳频率分辨率的确定方法,提高信号分析中的准确性。3、对实际瞬态冲击信号的降噪方法和高频信号重构等预处理方法进行研究,结合故障诊断的实际,研究参数化的Hilbert 时频谱来表征瞬态冲击信号。本项目的研究能够有效提高机械设备故障诊断的准确性,促进我国设备预知维修的发展。并为瞬态冲击信号的特征提取方法,诸如医学、通讯、建筑等领域的信号分析提供借鉴。
中文关键词: 瞬态冲击信号;短序列;Hilbert 时频谱;瞬时频率;评价模型;
英文摘要: Instantaneous impact signal contains a lot of characteristics reflecting machine working condition. But there are a lot of reasons limited instantaneous impact signal analysis, such as nonstationary and nonlinear signal characteristics, noise interference for the practical testing signal and most of instantaneous impact signal belonging to short series. Therefore, it is urgent to improve instantaneous impact signal analysis method. According to investigation on Hilbert time-frequency spectrum characterized instantaneous impact signal and machine impact signal characteristics, optimization of Hilbert time-frequency spectrum characterized instantaneous impact signal will be investigated in this research. Firstly, to improve the effectiveness of instantaneous frequency calculation, phase estimation of polynomial phase is used to calculate instantaneous frequency for intrinsic mode functions obtained by empirical mode decomposition. Then, performance estimation model will be constructed for Hilbert time frequency spectrum characterized signal analysis. To improve signal analysis accuracy, the best frequency resolution of Hilbert time frequency spectrum will be investigated. In the end, parameter Hilbert time frequency spectrum characterized instantaneous impact signal corresponding to fault diagnosis of practical situation will be investigated according to practical instantaneous impact signal de-noising method and high frequency signal reconstruction pre-processing method research. The research of this project will improve the accuracy of machine fault diagnosis and contribute to development of preventative maintenance. At the same time, this research will be used for reference to instantaneous impact signal feature extraction during signal analysis in the area, such as medicine, communication, architecture and so on.
英文关键词: Instantaneous impact signal; Short series; Hilbert time-frequency spectrum; Instantaneous frequency; Estimation model;