共查询到18条相似文献,搜索用时 750 毫秒
1.
阐述了一种新的时频分析方法——基于Hilbert-Huang变换的Hilbert时频谱分析,并将其应用到纱线信号分析中。以毛条信号为例,将短时Fourier变换、小波分析和Hilbert谱分析方法进行比较,指出Hilbert谱分析能对信号的时频分布做出比短时Fourier变换和小波分析更为精确的描述,从而能更准确地揭示纱线信号中能量在时域和频域同时变化的过程。通过对纱线信号进行时频分析,可以得到每个瞬间的频率变化,进而对其进行在线谱分析,可以实时监测工艺及设备状况的变化,及时发现生产过程中的异常。 相似文献
2.
本研究针对轴承早期故障信号微弱不易检测的缺陷,提出了一种基于谱峭度Morlet小波变换法和经验模态分解的轴承早期故障信号检测与增强的方法.该方法把谱峭度引入到Morlet小波中,根据谱峭度变化寻找Morlet小波的最佳匹配参数,最大限度地增强故障弱信号;然后对信号进行经验模态分解,以信号相关系数大小确定重构信号的基本模态分量,对重构信号进行包络分析完成故障诊断.实验结果验证了此方法在滚动轴承早期故障诊断中的良好应用效果. 相似文献
3.
4.
5.
针对风力发电机组轴承故障振动信号特性,该信号为非平稳信号。提出基于希尔伯特黄的对风力发电机组轴承故障的诊断方法。该方法是用经验模式分解方法将原始信号分解成包含不同特征时间尺度的本征模态函数,并对分解后得到的每一个本征模态函数分量进行希尔伯特变换,从而得出时频平面上的幅值分布的尔伯特谱,并通过其反映出来的物理信息对风力发电机组轴承做故障诊断分析。 相似文献
6.
为解决集合经验模态(EEMD)存在分量重构误差大和提取的故障特征不明显问题,课题组提出一种自适应噪声完备集合经验模态分解(CEEMDAN)辅助快速谱峭度的故障诊断方法。首先采用CEEMDAN将故障信号分解为多个IMF分量,计算分量的谱峭度值,选择峭度和相关度最大的分量进行重构;然后通过快速谱峭度图确定最大共振频带,进行带通滤波分析,获得故障信息;最后采用某滚动轴承实验数据分别对内圈故障和外圈故障进行实验分析。结果表明:与原始故障信号相比,该方法获得的包络谱更清晰,故障频率更明显,内圈故障频率为162 Hz,外圈故障频率为107 Hz。该方法提取故障特征突出,可以得到有效的故障频带。 相似文献
7.
8.
9.
提出一种基于希尔伯特—黄变换(HHT)—线性判别分析(LDA)的枸杞产地电子舌辨识方法。以宁夏、新疆、甘肃、青海4个产地的枸杞为试验材料,采用伏安电子舌采集不同产地枸杞的"指纹图谱"信息,利用集合经验模态分解(EEMD)对电子舌原始信号进行多尺度分解得到一组本征模态函数(IMF),分别求取其奇异谱熵和Hilbert边际谱作为特征向量。在该基础上,利用LDA建立枸杞产地非线性组合预测模型。试验结果表明,HHTLDA与分别采用特征点提取(FPE)、主成分分析(PCA)和离散小波变换(DWT)的算法相比,具有更好的分类效果。对未知产地枸杞的总体分类精度和kappa系数分别达到98%和0.973,均表明该模型具有较好的鉴别效果。 相似文献
10.
为了实现对不同贮藏年限的红酒进行客观的辨别分析,提出一种采用电子舌结合集合经验模态分解(Ensemble empirical modal decomposition, EEMD)、鲸鱼算法(whale optimization algorithm, WOA)和最小二乘支持向量机(least square support vector machine, LSSVM)组合模型的区分方法。首先采用电子舌对4种不同贮藏年限红酒的“特征图谱”进行信息采集;然后利用EEMD对原始信号进行分解,提取分解后的本征模态函数奇异谱熵和希尔伯特边际谱作为特征数据;最后,采用鲸鱼算法优化最小二乘支持向量机建立红酒贮藏年限分析模型。结果表明,EEMD-WOA-LSSVM组合模型对不同贮藏年限的红酒的分类准确率、精确率、召回率、F1-score和Kappa系数分别达到97.5%、97.75%、97.5%、0.98和0.97,其区分能力优于GA-LSSVM、PSO-LSSVM和SVM模型。该研究可为红酒贮藏年限区分提供一种新的研究思路和技术手段。 相似文献
11.
Binita Baishya Kalita Seiko Jose Sunita Baruah Sanghamitra Kalita Smita Rani Saikia 《Journal of Natural Fibers》2019,16(1):49-57
Roselle is a bast fiber, and its utilization as a textile fiber for the development of textile products is still scanty. A work has been attempted to develop yarn from Roselle. Fibers were extracted from Roselle bark by decortication and degummed in alkaline medium. The degumming process was optimized based on fiber yield and strength. The degummed fibers were then bleached by the hydrogen peroxide bleaching process. Degummed and bleached fibers were characterized by scanning electron microscope, Fourier transform infrared spectroscopy, and X-ray diffraction analysis. Degumming and bleaching results in enhancement of density, fineness, and brightness of Roselle fiber. A marginal decrease in tenacity of the Roselle fiber was observed after bleaching; however, the strength was not affected by degumming. Fibers were converted into fine yarn in the jute spinning system. The yarn properties inferred that the yarn possessed essential properties for the preparation of apparels and home furnishing. 相似文献
12.
13.
基于LabVIEW的纱条数字信号发生器的实现 总被引:1,自引:0,他引:1
宋晓梅 《纺织高校基础科学学报》2005,18(1):90-93
通过对现场纱条信号模型的分析,在LabVIEW软件环境下,将由纤维直径、纤维长度和纤维根数及其它概率分布构成的纱条信号用计算机仿真产生,实现了纱条数字信号发生器.该信号发生器在纱条波谱分析仪的调试中获得验证. 相似文献
14.
Fabric surface unevenness and defects are usually created by yarn irregularity and defects in weaving process. These unevenness and irregularities appear on fabric and affect various fabric properties. In order to investigate weft yarn mass irregularities on fabric surface unevenness and defects, various plain fabric samples were prepared in which they differ only in weft yarn mass irregularity. One of the effective factors on fabric surface unevenness is yarn protrusion in fabric structure. Yarn protrusion in fabric structure is influenced by yarn’s physical and mechanical properties and fabric structural characteristics. In this work, relationship between mass irregularities of weft yarn and fabric surface unevenness was investigated using angular power spectrum curve, a measure of yarn protrusion in fabric structures. The results showed a high significant correlation between these two parameters. 相似文献
15.
Digital image analysis techniques in the spatial and frequency domains for twist measurement of yarns are described. A spatial technique is developed to extract the twist angle through the analysis of the yarn core image. Then, a Fourier transformation technique is applied to yarn images to measure the orientation of the fibre on the yarn surface. Finally, a hybrid method that incorporates frequency domain filtering prior to spatial analysis is proposed. The trials show that spatial analysis is a fast method and can successfully predict the twist in the yarn. Fourier transformation technique is quite sensitive to the protruding fibres obstructing the yarn surface, which may result in measurements having high variations. For yarns having little amount of hairs protruding from the core, the results agreed reasonably well with actual twist levels. Frequency domain filtering in conjunction with the spatial analysis of the yarn surface is found to be superior in terms of accuracy. The twist values calculated using the more reliable diameter measurements with back-lit images together with twist angles from the front-lit images are found to be more accurate when compared with the actual values. 相似文献
16.
设(?)和(?)为复Hiblert空间,给定算子A∈(?)((?)),B∈(?)((?))。当算子对(A,B)是可控算子对时,通过空间分解、极分解及构造算子矩阵的技巧,利用数学归纳法给出著名谱配置定理的一个比较清晰的证明,然后给出这个定理的一个应用. 相似文献
17.
18.
介绍了突发性纱疵的快速诊断方法。通过对工厂生产的常规品种的分析,提出优化牵伸比,使不同品种的生产工艺保持基本一致,这样在波谱图上出现的异常峰值的位置只与设备的状态相关,为快速诊断带来了便利。根据工厂生产的实际经验,总结出异常峰值的位置与其对应的设备机构缺陷的位置关系,查找出突发性纱疵产生的原因及排除设备缺陷的措施,快速诊断,及时调整设备与工艺,把纱疵消灭在萌芽之中,提高了纱条的质量。 相似文献