基于小波多分辨率分析和RBF神经网络的刀具磨损状态识别

刀具磨损监测对于提高加工过程的精度和自动化程度具有重要意义。本文提出一种基于RBF函数神经网络的刀具磨损状态监测模式。该系统利用声发射传感器对切削过程进行监测,采用多分辨率小波分解技术从声发射信号中提取反映刀具磨损的特征向量,并输入RBF神经网络,实现了刀具磨损的自动识别。     

A probabilistic neural network applied in monitoring tool wear in the end milling operation via acoustic emission and cutting power signals

Tool condition monitoring, which is very important in machining, has improved over the past 20 years. Several process variables that are active in the cutting region, such as cutting forces, vibrations, acoustic emission (AE), noise, temperature, and surface finish, are influenced by the state of the cutting tool and the conditions of the material removal process. However, controlling these process variables to ensure adequate responses, particularly on an individual basis, is a highly complex task. The combination of AE and cutting power signals serves to indicate the improved response. In this study, a new parameter based on AE signal energy (frequency range between 100 and 300 kHz) was introduced to improve response. Tool wear in end milling was measured in each step, based on cutting power and AE signals. The wear conditions were then classified as good or bad, the signal parameters were extracted, and the probabilistic neural network was applied. The mean and skewness of cutting power and the root mean square of the power spectral density of AE showed sensitivity and were applied with about 91% accuracy. The combination of cutting power and AE with the signal energy parameter can definitely be applied in a tool wear-monitoring system.     

基于切削声信号与优化SVM的刀具磨损状态监测

提出了一种利用切削声实现刀具磨损状态多特征监测的方法。根据经验模态分解与Hilbert变换理论,提取切削声信号的内禀模态能量与不同频段的Hilbert谱能量作为监测信号的备选特征。采用支持向量机作为分类器,针对备选特征的有效筛选问题,利用多种群遗传算法对分类器的输入特征进行了优化,剔除备选特征中的干扰特征,利用多种群遗传算法对分类器的模型参数进行了优化。利用优化后的分类器对测试样本进行分类,并与优化前的分类结果进行了对比。结果表明,优化后分类器的分类性能得到了明显提升,该方法可以对刀具磨损状态进行有效识别。     

基于主轴电流信号多特征融合的刀具磨损状态监测

为实现在正常生产条件下进行刀具磨损的长期在线监测,提出了基于主轴电流信号和粒子群优化支持向量机模型(PSO-SVM)的刀具磨损状态间接监测方法.首先对数控机床主轴电机电流信号进行分析,将与刀具磨损相关的主轴电流信号多个特征参数和EMD能量熵进行特征融合作为输入特征向量;其次,通过粒子群寻优算法(PSO)对支持向量机模型...     

Application of backpropagation neural network for spindle vibration-based tool wear monitoring in micro-milling

This study develops a micro-tool condition monitoring system consisting of accelerometers on the spindle, a data acquisition and signal transformation module, and a backpropagation neural network. This study also discusses the effect of the sensor installations, selected features, and the bandwidth size of the features on the classification rate. To collect the vibration signals necessary for training the system model and verifying the system, an experiment was implemented on a micro-milling research platform along with a 700?μm diameter micro-end mill and a SK2 workpiece. A three-axis accelerometer was installed on a sensor plate attached to the spindle housing to collect vibration signals in three directions during cutting. The frequency domain features representing changes in tool wear were selected based on the class mean scatter criteria after transforming signals from the time domain to the frequency domain by fast Fourier transform. Using the appropriate vibration features, this study develops and tests a backpropagation neural network classifier. Results show that proper feature extraction for classification provides a better solution than applying all spectral features into the classifier. Selecting five features for classification provides a better classification rate than the case with four and three features along with the 30?Hz bandwidth size of the spectral feature. Moreover, combining the signals for tool condition from both direction signals provides a better classification rate than determining the tool condition using a one-direction single sensor.     

刀具磨损监测中信号特征值的提取与聚类分析

声发射是近年刀具监测研究中采用的新技术之一,本文用包络分析法求取刀具磨损中声发射信号的包络线,用其时序模型参数作为特征值,采用神经网络对刀具磨损状态进行分析,试验表明效果良好。     

Acoustic emission method for tool condition monitoring based on wavelet analysis

It is believed that the acoustic emission (AE) signals contain potentially valuable information for tool wear and breakage monitoring and detection. However, AE stress waves produced in the cutting zone are distorted by the transmission path and the measurement systems and it is difficult to obtain an effective result by these raw acoustic emission data. In this article, a technique based on AE signal wavelet analysis is proposed for tool condition monitoring. The local characterize of frequency band, which contains the main energy of AE signals, is depicted by the wavelet multi-resolution analysis, and the singularity of the signal is represented by wavelet resolution coefficient norm. The feasibility for tool condition monitoring is demonstrated by the various cutting conditions in turning experiments.     

基于神经网络的多特征融合刀具磨损量识别

采用切削力信号监测钻削过程钻头的磨损量 ,分别从时域、频域提取了切削力信号的均值、方差、峭度系数和特定频段能量作为刀具磨损的特征信号 ,讨论了特征信号随着刀具磨损量增加的变化规律 ,并将各个特征信号构成的特征矢量输入多层反传神经网络进行融合 ,实现钻削过程刀具磨损量的智能识别。试验结果表明该方法能有效实现多特征融合 ,但识别精度和推广能力有待进一步提高     

刀具磨损状态在线监测技术

通过对切削加工过程中声发射（ＡＥ）信号频域特改正的研究,指出了ＡＥ信号中与刀具磨损状态有关的成分在频域内的分布规律,在此基础上分析了ＡＥ信号的均方根值ＡＥＲＭＳ和振铃计数与刀具后面磨损量ＶＢ之间的对应关系,建立了一种基于ＡＥ信号能量的双阈值判断监测方法。试验表明该方法能很好地解决持定切削条件下刀具磨损的在线监测问题。     

模糊数据融合的刀具磨损状态智能识别

为实现高速加工时刀具渐变磨损状态的在线准确识别,提出了一种集合多种智能的间接检测刀具磨损状态方法的模糊数据融合方法。尽管这些方法具有算法实现较为简单、处理速度较快的优点,但单一的信号检测及单一的智能建模方法难以获得全面的加工状态信息和准确的识别结果。为此,利用F推理技术对上述方法的冗余和互补信息进行数据融合,应用Makino—Fanuc 74-A20型加工中心的测试数据验证了该方案的可行性,并将刀具后刀面磨损的预测值与基于机器视觉检测的实测值进行比较。实验结果分析表明,多参数模糊融合识别方法能快速获得切削刀具磨损状态更加准确的预测值。     

基于声发射技术监测刀具磨损的研究

刀具磨损的研究方法很多,本文针对近些年发展的声发射技术(AE)在监测刀具磨损上的应用,采用理论分析和现场试验的方法进行可行性分析和验证,结果表明:在考察的几个影响声发射信号强度的因素之中,刀具主切削刃后刀面磨损量对其影响最为显著,这为利用AE技术研究刀具磨损提供了可行性依据;通过对铣刀AE信号进行时域振铃分析,清晰再现了刀具在不同时刻的磨损情况。     

基于云理论与LS-SVM的刀具磨损识别方法

针对刀具磨损过程中产生声发射信号的不确定性以及神经网络学习算法收敛速度慢、易陷入局部极小值、对特征要求较高等问题,提出了基于云理论和最小二乘支持向量机的刀具磨损状态识别方法。首先,对声发射信号进行小波包分解与重构,滤除干扰频段对求取特征参数的影响;其次,对重构后的信号利用逆向云算法提取云特征参数:期望、熵、超熵,分析刀具磨损声发射信号的云特性及磨损状态与云特征参数之间的关系;最后,将云特征参数组成特征向量送入最小二乘支持向量机进行识别。研究结果表明:所提取的特征可以很好地反映刀具的磨损状态,云-支持向量机方法可以有效地实现刀具磨损状态的识别,与传统神经网络识别方法相比具有更高的识别率,识别率达到96.67%。     

Study of high-frequency sound signals for tool wear monitoring in micromilling

This study analyzed the sound signals obtained from the micromilling process for microtool wear monitoring. Various spans of spectral features were created by analyzing sound signals on tool wear monitoring in microcutting. The selection algorithm based on class mean scattering criteria and the hidden Markov model (HMM) model was developed to verify the effect of various feature selection algorithms on the system performance. The effect of the feature bandwidth size, the size of observation sequence, and choice of the hidden states for HMM parameters were also studied. The results indicate that the normalized sound signals obtained from the single microphone with a frequency range between 20 and 80 kHz demonstrated the potential to provide a solution to monitor micromills with the proper selection of feature bandwidth and other parameters.     

WEAR MONITORING IN TURNING OPERATIONS USING VIBRATION AND STRAIN MEASUREMENTS

For the efficient and reliable operation of automated machining processes, the implementation of suitable tool condition monitoring (TCM) strategy is required. Various monitoring systems, utilising sophisticated signal processing techniques, have been widely researched for a number of different processes. Most monitoring systems developed up to date employ force, acoustic emission and vibration, or a combination of these and other techniques with a sensor integration strategy. With this work, the implementation of a monitoring system utilising simultaneous vibration and strain measurements on the tool tip, is investigated for the wear of synthetic diamond tools which are specifically used for the manufacturing of aluminium pistons. Contrary to many of the earlier investigations, this work was conducted in a manufacturing environment, with the associated constraints such as the impracticality of direct measurement of the wear. Data from the manufacturing process was recorded with two piezoelectric strain sensors and an accelerometer, each coupled to a DSPT Siglab analyser. A large number of features indicative of tool wear were automatically extracted from different parts of the original signals. These included features from the time and frequency domains, time-series model coefficients (as features) and features extracted from wavelet packet analysis. A correlation coefficient approach was used to automatically select the best features indicative of the progressive wear of the diamond tools. The self-organising map (SOM) was employed to identify the tool state. The SOM is a type of neural network based on unsupervised learning. A near 100% correct classification of the tool wear data was obtained by training the SOM with two independent data sets, and testing it with a third independent data set.     

A Feature Extraction Method for the Wear of Milling Tools Based on the Hilbert Marginal Spectrum

Abstract

The Hilbert–Huang transform (HHT) can adaptively delineate complex non-linear, non-stationary signals when used as the Hilbert–Huang marginal spectrum through empirical mode decomposition (EMD) and the Hilbert transform, to highlight local features of signals. Characterized by high resolution, the Hilbert marginal spectrum has been widely applied in mechanical signal processing and fault diagnosis. In the research, an HHT based on the improved EMD was proposed to analyze the cutting force, vibration acceleration (AC), and acoustic emission (AE) signals during tool wear in the milling process. At first, the collected signals were subjected to range analysis, which revealed that tool wear was closely related to the signals collected during the cutting process. Then, EMD was applied to the signals, followed by variance analysis after calculating the energies of each intrinsic mode function (IMF) component. Afterwards, the IMF components significantly influenced by wear degree, while slightly influenced by the three cutting factors (cutting velocity, feed per tooth, and cutting depth), were selected as IMF sensitive to the degree of wear. The HHT was finally applied to the sensitive IMF components of signals containing major tool wear information, thus obtaining the Hilbert marginal spectra of the signals, which were able to reflect the changes in signal amplitude with frequency. On the basis of the Hilbert marginal spectrum, the method defined the feature energy function which was then used as the eigenvector for predicting tool wear in milling processes. The analysis of signals in four tool wear states indicated that the method can extract salient tool wear features.     

神经网络在非石棉垫片分类中的研究

提出运用神经网络的分类方法来对已知的15种非石棉垫片的配方数据进行分类。分别运用神经网络中的PNN,LVQ和SOM神经网络对其进行分类。结果表明,PNN神经网络和LVQ神经网络在所提供的数据中均能进行有效的分类,而SOM的分类结果则不太理想。     

Model development for tool wear effect on AE signal generation in micromilling

A model for the relation between the acoustic emission signal generation and tool wear was established for cutting processes in micromilling by considering the acoustic emission (AE) generation and propagation mechanisms. In addition, the effect of tool wear on the AE signal generation in frequency and amplitude was studied. In the model development, the finite element analysis was first used to calculate the shear strain rate distribution on the shear plane based on the orthogonal cutting assumption. Conversely, the contact stress distribution of workpiece on the flank wear face was established based on the Waldorf model. Following the finite element method, the dislocation density in materials was calculated based on Orowan’s law with the calculated stress rate. Finally, the AE signal detected by the sensor was calculated by considering the Gaussian probability density function for the distribution of AE source on the shear plane and the one-dimension wave equation for AE signal propagation. Based on the developed model, the effect of tool wear on the AE signal generation was investigated and compared to the experimental results. The results obtained from these investigations indicate that the proposed model can be used to predict the effect of tool wear on the AE signal generation.     

刀具磨损声发射信号处理中小波基选取的研究

通过对小波基性质和刀具磨损声发射（AE）信号特点的研究,从理论上分析了小波变换中刀具磨损AE信号处理中小波基选取的方法。在试验验证过程中,根据小波包信号分解遵循能量守恒原理,用四种小波基对刀具磨损AE信号进行三层小波包分解;以AE信号经小波包分解后各频带上的能量为特征参数,比较四种情况下特征参数的变化,验证了理论分析的正确性。     

刀具磨损监测及破损模式的识别

对于金属切削过程中的刀具磨损,提出了基于隐马尔可夫模型的模式识别理论来识别刀具的不同磨损状态,从而预报刀具破损.该方法对切削过程中切削力信号的动态分量和刀柄振动信号进行快速傅里叶变换特征提取,然后利用自组织特征映射对提取的特征矢量进行预分类编码,把矢量编码作为观测序列引入到隐马尔可夫模型中进行机器学习,建立了3个不同磨损状态的隐马尔可夫模型,并利用最大概率进行模式识别.试验表明,该方法对车刀磨损过程进行识别和预报是有效的.     

基于小波模糊神经网络刀具监控系统研究

针对切削过程中振动信号和AE信号的特点,利用小波分析技术提取信号深层特征,建立了新型的基于模糊推理的神经网络模型,该模型能融合振动和AE信号的特征和描述信号特征与刀具状态的非线性关系,以此识别刀具状态。试验表明小波模糊神经网络对提高在线刀具监控系统的可靠性极为有效。