钻削声发射信号高阶谱研究及应用

为解决批量钻削工序质量检测问题,采集各工步加工过程声发射监测信号,提出一种基于声发射信号高阶谱分析的批量钻削质量检测方法。基于统计意义上正常钻削过程声发射信号符合高斯分布的假说,对采集的信号进行小波包消噪后,计算批量钻削工步信号的双谱切片,描述信号偏离高斯分布的程度,并分析其与钻削加工工步质量的映射关系,实现钻削加工工步质量检测。实验及分析结果表明:基于声发射信号双谱切片提取的信号特征可有效辨别批量钻削工步中的质量不合格品。     

基于振动信号过程特征的批量钻削工序质量检测

为实现批量钻削质量的快速检测分析,采用振动传感器监控钻削过程,提出一种基于振动信号处理的批量钻削质量监测分析方法.基于钻刃与工件接触位置变化状况,在钻削过程中提炼出与钻削加工质量密切相关的三个阶段:引钻、钻孔和出钻三阶段.采用正交包络法求解振动信号的瞬时频率及瞬时幅值,定位这三个阶段初始时刻的瞬态时间点.利用这些瞬态时间点对振动信号进行分割,获得与每一个钻削过程中引钻、钻孔和出钻三阶段对应的钻削过程振动信号数据.采用主成分分析方法,综合此三个阶段内振动信号的统计特征对批量钻孔质量分布进行分析.计算和分析结果证明,可在3%的误差内提取瞬态特征点,分析和评估批量钻削质量准确度高.     

钻削振动信号小波包分频谱减特征增强方法

针对在钻削加工噪声背景下振动信号微弱特征识别和提取困难的问题,提出了一种基于小波包分频谱减的钻削振动信号特征增强方法。首先,在经典谱减法原理的基础上,将钻削前机床空转信号视为监测信号的"加性噪声";其次,根据钻削过程振动信号的特点,采用小波包分解方法将"加性噪声"和监测信号分成多个子频带;最后,对每个子频带内"加性噪声"的相应频带进行谱减处理后,重构钻削振动信号。仿真和实验结果表明,该方法能有效降低环境噪声对钻削过程特征提取的影响,从而达到增强监测信号特征的目的,同时建立了钻削过程与监测信号之间很好的映射关系模型。     

孔系钻削振动信号特征波动可视化研究及其应用

为解决工程应用中切削参数一致的孔系加工质量一致性评估的难题,提出了一种基于振动信号特征波动可视化的聚类分析方法。首先采用振动传感器监控孔系钻削过程,提取各孔振动信号小波包能量谱和高阶统计量特征;然后利用雷达图得到各孔振动信号特征矩阵分布图,提取信号特征雷达图多边形重心特征;最后采用模糊C-均值（FCM）算法对雷达图平面重心点集进行聚类分析。理论分析结果与人工检测结果对比表明：该方法可直观呈现孔系钻削质量分布情况,简便、可靠地实现孔系钻削质量的一致性评估。     

深孔啄钻振动信号DTW算法改进及应用

针对深孔加工质量检测难、效率低和成本高等问题,提出一种基于振动监测信号分割的改进动态时间规整算法,实现深孔加工质量一致性快速无损检测。首先,以啄钻间歇进给式加工采集的振动信号为研究对象,利用抗噪性能优良的双窗双谱算法对其进行分割;其次,对分割信号在时域和频域内进行特征降维;然后,针对信号长度不等引起的特征不等情况,采用改进的动态时间规整算法(dynamic time warping,简称DTW)进行规整对齐,同时达到减小时间复杂度和防止病态规整的目的;最后,利用求得的累积最短距离评估啄钻阶段振动信号的相似性程度,从而判别啄钻加工质量的一致性。仿真和试验结果表明,该方法能快速有效完成对深孔加工质量一致性无损检测,振动信号分析结果与实际物理检测结果相吻合。     

基于BP神经网络的超声振动钻削钻头磨损状态监测试验研究

搭建了超声轴向振动钻削钻头磨损状态的钻削力和声发射信号采集系统,采集不同磨损状态下钻中区域的钻削力和声发射信号进行小波分解,得到与钻头磨损状态相关的特征量作为识别钻头磨损状态的特征参数,输入到建立的6-13-3的三层BP神经网络模型中进行融合,识别钻头磨损状态。试验结果表明,通过BP神经网络技术将钻削力和声发射信号融合识别钻头磨损的准确率约88.9%,能够有效监测钻头磨损状态。     

三区段变参数振动钻削微孔的研究

分析了振动钻削微孔时钻入、钻削和钻出三个区段上不同的振动钻削机理,通过二次回归试验求出了各区段的最佳振动参数。提出了随钻孔区段的改变而改变振动参数,使各区段都工作在最佳振动钻削状态的三区段变参数振动钻削新方法。研究结果表明,这一新方法克服了定参数振动钻削微孔时各项钻削质量顾此失彼的矛盾,是全面提高微孔加工质量的有效方法。     

基于小波包能量谱的HMM钻头磨损监测

从工程应用的角度论述了小波包分解原理及其能量谱监测理论，并将该理论应用于钻削力信号特征提取中，针对钻削过程特征矢量与钻头磨损之间具有较强的随机性和不确定性的特点，提出一种基于隐马尔可夫模型（HMM）的钻头磨损监测方法。实验结果表明，通过对钻削力信号进行多层小波包分解，提取各频段能量谱作为特征矢量可准确刻画工艺系统随钻头磨损的演化规律，利用HMM建立的各钻头磨损状态小波包能量谱的统计模型可有效跟踪钻头磨损的发展趋势，实现钻头磨损状态和寿命的监测。     

面向钻削过程监测的振动信号处理及状态分类

易变形结构在钻削过程中因受力而产生形变,不能根据刀具的钻削距离判断刀具所处状态.钻削过程中刀具切入易变形结构不同的位置会产生不同幅度的振动,通过对加速度传感器采集到的振动信号进行快速傅里叶变换(fast Fourier transform,简称FFT),将钻削过程分为5个状态.通过计算系统基频整数次谐波分量幅值的变异系...     

钻削加工中声发射的特性研究

本文介绍了钻削加工中应用声发射(AE)技术时,传感器安装位置对声发射信号的影响,随钻深增加AE信号的变化及钻头断裂前AE信号的特征,对消除外部干扰影响、提高检测可靠性、减少测量误差所采取的方法也作了简述,有助于AE用于钻削在线测量的实用化。     

Bi-spectrum based-EMD applied to the non-stationary vibration signals for bearing faults diagnosis

Empirical mode decomposition (EMD) has been widely applied to analyze vibration signals behavior for bearing failures detection. Vibration signals are almost always non-stationary since bearings are inherently dynamic (e.g., speed and load condition change over time). By using EMD, the complicated non-stationary vibration signal is decomposed into a number of stationary intrinsic mode functions (IMFs) based on the local characteristic time scale of the signal. Bi-spectrum, a third-order statistic, helps to identify phase coupling effects, the bi-spectrum is theoretically zero for Gaussian noise and it is flat for non-Gaussian white noise, consequently the bi-spectrum analysis is insensitive to random noise, which are useful for detecting faults in induction machines. Utilizing the advantages of EMD and bi-spectrum, this article proposes a joint method for detecting such faults, called bi-spectrum based EMD (BSEMD). First, original vibration signals collected from accelerometers are decomposed by EMD and a set of IMFs is produced. Then, the IMF signals are analyzed via bi-spectrum to detect outer race bearing defects. The procedure is illustrated with the experimental bearing vibration data. The experimental results show that BSEMD techniques can effectively diagnosis bearing failures.     

Monitoring of Ball Bearing Operation under Grease Lubrication Using a New Compound Diagnostic System Detecting Vibration and Acoustic Emission

Our new compound diagnostic system comprised of a compound sensor, a signal processor, and a personal computer installed signal processing software. The compound sensor made by an advanced sensor fusion technique was able to detect simultaneously the vibration acceleration and the acoustic emission by itself. The signal processor received a signal from the sensor and separated it into the vibration acceleration signal and the acoustic emission signal. The signal processor and the personal computer processed the acceleration signal and acoustic emission signal for diagnostic information. The rolling contact fatigue process of a ball bearing under grease lubrication was monitored using the compound system. The system outputs diagnostic information, for example, the means, the variance, the skewness, and the kurtosis of the vibration acceleration signal and the acoustic emission signal. In diagnosing the rolling contact fatigue failure, the root mean square (rms) value of the vibration acceleration was most effective, and the mean of the demodulated acoustic emission was second to the rms value of the acceleration in effectiveness. From the result of the evaluation, it became clear that the system was useful for diagnosing rolling contact bearings under grease lubrication.     

基于力信号的低频振动制孔刀具磨损状态分析

为分析碳纤维增强树脂基复合材料(CFRP)/钛合金(TC4)叠层材料低频振动制孔工艺下刀具磨损状态,开展基于切削力信号的制孔刀具磨损状态研究.通过采集CFRP/TC4叠层材料低频振动制孔过程中的切削力信号,进行时域和频域分析,探讨各信号特征量与刀具磨损状态之间的联系.研究结果表明:CFRP/TC4叠层材料低频振动制孔轴...     

混凝土结构中PZT传感器的多功能性分析及应用

针对混凝土结构被动监测中的压电陶瓷(也称锆钛酸铅,Pb-based lead zirconium titanate,简称PZT)传感器信号的多功能特性,提出了PZT传感器不同用途信号的提取方法。根据不同功能信号的频率范围差异,通过小波Mallat分解,得到用于反映结构整体动态信息的振动信号以及局部断裂破坏引起的声发射信号,通过与加速度传感器和声发射传感器信号比较,验证了提取方法的正确性,并将该方法应用到钢筋混凝土框剪结构模型地震破坏试验的实时监测中。试验结果表明,应用该方法提取到的振动信号能准确测得结构主频率等结构的动态信息,声发射信号部分能清楚捕捉局部损伤引起的能量释放情况。应用该方法可准确提取出结构的振动信号和声发射信号,利用同一PZT传感器能够实时评估和监测结构的整体动态特性和局部损伤状况。     

基于振动信号特征量SVM分类的配变故障识别方法

配变油箱表面的振动信号富含绕组和铁芯的各类状态信息,是绕组和铁芯工作状况的最直接体现。采用希尔伯特黄(Hilbert-Huang Transform,HHT)带通滤波提取配电变压器振动信号主成分,获得表征绕组状态的100Hz分量及表征铁芯状态的150~1000Hz分量;利用负载电流拟合法提取绕组振动信号的特征量,通过测到的已知振动信号估计指定负载下的绕组100Hz振动幅值,构成绕组振动的特征向量;利用具有良好泛化能力及鲁棒性的双谱奇异值表征铁芯振动的特征。提取实验室试验测得的绕组松动、绕组变形、铁芯松动、铁芯两点接地以及铁芯接地不良等故障振动信号的特征向量,用基于信息融合的支持向量机(Support Vector Machine,SVM)实现绕组和铁芯状态的识别,结果验证了本文方法的有效性和准确性。     

基于声发射的双谱分析在金刚笔状态特征提取中的应用

为了有效地识别砂轮修整过程中金刚笔的钝化状态，针对金刚笔修整砂轮过程中声发射信号的非平稳时变特点，研究了用于金刚笔状态特征提取的基于声发射的双谱分析方法。通过砂轮修整过程在线监测实验，分析了金刚笔在锐利、中等钝化和钝化状态下，修整过程中声发射信号的双谱特征，提出了归一化双谱模对角切片特征提取方法。研究表明，采用基于声发射的双谱分析方法可以有效监测砂轮修整过程中金刚笔的钝化状态。     

Acoustic emissions for particle sizing of powders through signal processing techniques

The present work introduces an innovative method for measuring particle size distribution of an airborne powder, based on the application of signal processing techniques to the acoustic emission signals produced by the impacts of the powder with specific metallic surfaces. The basic idea of the proposed methodology lies on the identification of the unknown relation between the acquired acoustic emission signals and the powder particle size distribution, by means of a multi-step procedure. In the first step, wavelet packet decomposition is used to extract useful features from the acoustic emission signals; the dimensionality of feature space is further reduced through multivariate data analysis techniques. As a final step, a neural network is properly trained to map the feature vector into the particle size distribution.The proposed solution has several advantages, such as low cost and low invasiveness which allow the system based on this technique to be easily integrated in pre-existing plants. It has been successfully applied to the PSD measurement of coal powder produced by grinding mills in a coal-fired power station, and the experimental results are reported in the paper. The measurement principle can also be applied to different particle sizing applications, whenever a solid powder is carried in air or in other gases.     

Improving prediction accuracy of thermal analysis for weld-based additive manufacturing by calibrating input parameters using IR imaging

In this paper, we describe a condition classification technique designed to detect fault occurrence in an automotive light assembly during endurance testing. Inputs to the classifier are features extracted from vibration measurement data. They contain time domain parameters and frequency band energy parameters calculated using wavelet packet transforms. A support vector machine with Gaussian radial basis function kernel is designed for multiclass classification. A multiplex parameter estimation is achieved by searching for a minimum bound of the support vector count to achieve structural risk minimization. Through experiments, we show that the combination of effective feature extraction and classification with good generalization capability allows the proposed condition-monitoring system to be accurate and reliable. Additionally, acoustic signals known to have low signal to noise ratio are used as tests. We show that with the proposed methodology, acoustic signals can be used with increased sensitivity and accuracy for condition-monitoring purposes.     

基于长短时记忆卷积神经网络的刀具磨损在线监测模型

为了提高机械加工过程中刀具磨损在线监测的准确性，提出了一种基于长短时记忆卷积神经网络（LSTM-CNN）的刀具磨损在线监测模型。在该监测模型中，通过振动、力、声发射传感器对刀具切削过程中的振动、力和声发射信号进行采集，采集的数据其本质为时间序列数据。考虑采集数据的序列和多维度特性，采用LSTM-CNN网络对采集的数据进行序列和多维度特征提取，利用线性回归实现特征到刀具磨损值的映射。通过实验验证了该模型的有效性和可行性，模型的精度较其他几种方法有了较大的提高。