数字孪生驱动的数控铣削刀具磨损在线监测方法

针对数控铣床不断老化导致刀具磨损预测模型误差较大,加工过程中动态数据难以在线采集等问题,提出一种数字孪生驱动的刀具磨损在线监测方法。采用神经网络对加工过程中的多源数据进行特征提取,建立考虑机床老化的刀具磨损时变偏差量化模型,并在此基础上提出数控铣削刀具磨损的在线预测方法;开发了面向刀具磨损的数控铣削数字孪生系统,在线感知加工过程中的动态数据并实时仿真刀具磨损过程;最后,将该方法应用于实际加工中并与其他的预测方法进行了对比,结果表明该方法有效降低了机床老化带来的误差,实现了刀具磨损的精确预测。     

Tool wear prediction considering uncovered data based on partial least square regression

Tool wear prediction plays an important role in guaranteeing the workpiece quality and improving the production efficiency. However, because of the uncertainty and complexity of tool wear process, it is hard to ensure that the samples related to all tool wear values can be collected during the training stage. Therefore, the accuracy of tool wear prediction for these uncovered data will deteriorate severely. In this paper, partial least square regression is presented to realize the tool wear prediction based on force signal. The main characteristic of this method is that the regression analysis is in the principal component space so that the multicollinearity between explanatory variables can be avoided effectively. Side milling experiment was carried out to validate the effectiveness of the proposed model. The analysis and comparison under different number of uncovered data show that the partial least square regression based tool wear prediction is more accurate.     

CHMM for tool condition monitoring and remaining useful life prediction

To solve the problems of tool condition monitoring and prediction of remaining useful life, a method based on the Continuous Hidden Markov Model (CHMM) is presented. With milling as the research object, cutting force is taken as the monitoring signal, analyzed by wavelet packet theory to reduce noise and extract the energy feature of the signal as a basis for diagnosis. Then, CHMM is used to diagnose tool wear state. Finally, a Gaussian regression model is proposed to predict the milling tool’s remaining useful life after the test sample data are verified to be consistent with the Gaussian distribution based on a reliable identification of the milling tool wear state. The probability models of tool remaining useful life prediction could be established for tools with different initial states. For example, when an unknown state of milling force signal is delivered to the milling tool online diagnostic system, the state and the existing time of this state could be predicted by the established prediction model, and then, the average remaining useful life from the present state to the tool failure state could be obtained by analyzing the transfer time between each state in the CHMM. Compared to the traditional probabilistic model, which requires a large amount of test samples, the experimental cost is effectively reduced by applying the proposed method. The results from the experiment indicate that CHMM for tool condition monitoring has high sensitivity, requires less training samples and time, and produces results quickly. The method using the Gaussian process to accurately predict remaining life has ample potential for application to real situations.     

Cutting force-based real-time estimation of tool wear in face milling using a combination of signal processing techniques

In this paper, combinations of signal processing techniques for real-time estimation of tool wear in face milling using cutting force signals are presented. Three different strategies based on linear filtering, time-domain averaging and wavelet transformation techniques are adopted for extracting relevant features from the measured signals. Sensor fusion at feature level is used in search of an improved and robust tool wear model. Isotonic regression and exponential smoothing techniques are introduced to enforce monotonicity and smoothness of the extracted features. At the first stage, multiple linear regression models are developed for specific cutting conditions using the extracted features. The best features are identified on the basis of a statistical model selection criterion. At the second stage, the first-stage models are combined, in accordance with proven theory, into a single tool wear model, including the effect of cutting parameters. The three chosen strategies show improvements over those reported in the literature, in the case of training data as well as test data used for validation—for both laboratory and industrial experiments. A method for calculating the probabilistic worst-case prediction of tool wear is also developed for the final tool wear model.     

铣削钛合金加工表面形貌特征参数的预测

钛合金在铣削过程中受迫振动明显,刀—工接触关系不断变化,加工表面形貌特征参数难以预测,已成为制约加工表面质量进一步提高的瓶颈。针对铣削振动与加工表面形貌的非线性随机变化特性进行了切削钛合金试验,采用高斯过程回归法构建铣削振动作用下的加工表面形貌高斯过程模型。分析刀齿误差和铣削振动对加工表面形貌特征参数的影响规律,为以加工表面质量分布一致性为前提的铣削钛合金工艺设计提供参考依据。     

模具钢铣削中刀具磨损的试验研究

以面铣刀刀片磨损为研究对象,结合类神经网络系统建构高速数控铣削加工的预测模型。以加工参数为模型输入条件,刀腹磨耗为输出条件。采用多因素试验方法,选择切削速度、进给速度、切削深度三个试验参数,利用直交表式的试验计划法设计试验点。依照试验点铣削工件后再测量刀具加工后的刀腹磨耗量,进而求得倒传递网络所需的36组训练范例与11组验证数据。刀腹磨耗预测模式是利用类神经网络中的倒传递网络原理,以田口法求得倒传递网络参数的最优值。试验结果显示,刀腹磨耗随着切削速度、进给速度、切削深度增加而上升。铣削模具钢后,刀具磨耗预测值的平均误差为4.72%,最大误差为11.43%,最小误差为0.31%。整体而言,类神经网络对于铣削加工可进行有效预测。     

线性回归模型诊断和在线预测刀具磨损量的方法研究

目的是研究诊断端面铣刀磨损量和在线预测铣刀的剩余寿命的方法.采用线性回归模型估计测刀面的磨损量.线性回归模型的输入是从铣刀受力信号提取出的特征和切削条件,比如进给量、转速等.在诊断了刀具的磨损量后,采用双指数平滑方法跟随诊断结果预测铣刀的使用寿命.最后,通过卖验验证了基于线性回归模型得到的刀具的磨损量和基于双指数平滑方法在线预测铣刀的剩余寿命的可行性.     

基于改进PCANet模型的铣刀磨损预测方法研究

铣刀健康状况直接影响实际生产加工过程,因此开展铣刀状态监测研究具有较大工程意义。以卷积神经网络为代表的深度学习模型已经逐渐用于监测加工过程中的刀具状态。但是这些模型的可解释性较差,预测结果的差异性也较大。作为一种新颖的卷积神经网络变种,主成分分析模型(Principal component analysis network, PCANet)的可解释性好,但是特征自监督学习能力有待提升,且相关应用案例较少。针对以上问题,拟对PCANet模型进行优化,进而提出了一种激活主成分分析-最大池化-支持向量回归(Activated PCANet with max pooling and support vector regression, APCANet-MP-SVR)模型,用于自适应提取敏感特征并准确预测刀具磨损情况。首先引入tanh激活函数,提高模型泛化能力;再采用最大池化层替代哈希编码和直方图用于特征选择,进一步降低冗余特征规模;最后建立支持向量回归模型实时预测刀具磨损值。应用案例充分证明了所提模型能够更好地用于加工现场刀具磨损值预测。     

基于主轴电流的铣削力间接监测方法

实时准确地监测铣削状态对于提高加工质量与加工效率具有重要意义,切削力作为重要的加工状态监测对象,因其监测设备昂贵且安装不便而受到限制,为此提出一种考虑刀具磨损的基于主轴电流的铣削力监测方法.首先基于切削微元理论建立了考虑后刀面磨损的铣削力模型,并通过铣削实验进行铣削力模型系数标定;然后对主轴电流与铣削力的关系进行理论建...     

基于狄利克雷混合模型的刀具磨损量在线估计

提出了一种基于狄利克雷混合模型的刀具磨损状态监测和磨损量估计的新方法。该方法将刀具磨损过程描述为磨损量的累积过程,通过对磨损增量的连续估计获得刀具当前的磨损量估计。首先对原始力信号进行特征提取,接着在不确定磨损增量状态数量的前提下采用狄利克雷混合模型对特征自动分类,然后利用吉布斯采样方法确定模型参数,最终得到描述力信号特征与磨损增量映射关系的刀具磨损状态混合模型。根据该混合模型以及当前的力信号信息即可完成刀具磨损量的在线估计。真实应用案例证明了该方法能自适应学习磨损状态并有效估计刀具的连续磨损值。     

钛合金铣削过程刀具前刀面磨损解析建模

钛合金Ti6Al4V作为典型的航空航天难加工材料,在其铣削过程中硬质合金刀具的磨损会降低加工过程稳定性,进而影响加工效率和已加工表面表面质量。刀具前刀面磨损会导致刀具刃口强度降低,并影响切屑的流向和折断情况。针对前刀面磨损机理进行分析并构建了月牙洼磨损深度预测模型。首先运用解析方法构建了前刀面应力场模型,得到切屑在前刀面滑动过程中的刀具前刀面应力分布情况及磨损位置。基于刀-屑接触关系的基础上建立了前刀面温度场模型。然后,基于所得刀具前刀面应力与温度分布,构建综合考虑磨粒磨损、粘结磨损与扩散磨损的铣刀月牙洼磨损深度预测模型,获得月牙洼磨损预测曲线;结合铣刀月牙洼磨损带沿切削刃方向分布的特点,建立了随时间变化的铣刀前刀面磨损体积预测模型。最后通过试验验证了切削宽度对前刀面磨损的影响规律,预测结果与试验测量值具有较好的吻合性。结果表明随着切削宽度的增加,月牙洼磨损深度及前刀面磨损体积都随之增加。研究结果为钛合金铣削用刀具的设计和切削参数的合理选择提供了理论基础。     

球头铣刀刀具磨损建模与误差补偿

针对刀具磨损度量方式和模型建立的问题,以球头刀具为研究对象,提出球头铣刀刀具磨损的度量方式,建立球头刀具磨损模型.以复映磨损在硬度较软加工材料上的方式测量球头刀具磨损,确定刀具磨损模型系数,给出刀具磨损模型系数确定的具体实现方法.加工试验验证球头刀具磨损度量方式的合理性和所建立刀具磨损模型的正确性,同时针对数控铣削加工中球头铣刀刀具磨损引起的误差提出离线仿真误差补偿算法,给出离线仿真误差补偿算法的具体实现步骤,通过建立的刀具磨损引起的加工误差模型仿真获得加工走刀步的误差.对于误差超差的走刀步,预先修改数控加工(Numerical control,NC)程序,保证实际加工零件满足精度要求.误差补偿验证试验表明所提出的离线仿真误差补偿算法的正确性和有效性.     

In-situ process monitoring and adaptive control for precision micro-EDM cavity milling

The linear tool wear compensation method (LCM) is commonly applied in micro-EDM 3-D milling to compensate the tool length wear in order to achieve high machining accuracy. Traditional LCMs mainly rely on empirical models and off-line wear measurements, whereas the process dynamics are not taken into account. When machining complex 3D cavities, an increasing number of tool wear compensation cycles have usually to be performed in order to maintain the targeted machining accuracy. This negatively affects the duration of the overall machining cycle. To realize efficient precision micro-EDM cavity milling, without the necessity to predefine Z-axis tool feed in the NC trajectory before machining, an in-situ process control system is developed to adaptively control the tool wear compensation factor based on the discharge pulse behavior. Experiments have shown that the change of the compensation factor can be detected and also a continuous increase of the factor (over compensation) leads to the saturation of the mean effective pulse frequency. Pulse monitoring therefore provides valuable information for understanding the process dynamics and for selecting the machining parameters towards better machining efficiency. Furthermore, the information gathered in-situ can be utilized to predict the tool wear and perform in-situ tool wear prediction. To implement this on machine-level, a combined off-line and in-line adaptive control of the tool wear compensation factor is proposed and experimentally validated by milling different 3D cavities. The off-line adaptive control is only necessary when the predicted machining depth error exceeds a certain limit. In this way, more than 80% of the off-line adaptive control cycles can be eliminated, whereby a total save of cycle time up to 18% has been reached, while still maintaining the desired dimensional and form accuracy.     

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.     

薄壁件不一致刀齿铣削时铣削力系数构造与预测

针对薄壁件铣削过程中刀齿半径不一致现象引起的铣削力系数计算失真问题,提出构造刀齿半径不一致时的实际铣削力系数,并采用核偏最小二乘法对不同铣削用量时的实际铣削力系数进行预测。针对两齿螺旋铣刀铣削过程推导理论铣削力系数,根据刀齿半径不一致铣削过程引入名义铣削力,推导刀齿半径误差,构造实际铣削力系数;基于核分析方法突出的非线性分析及预测能力,提出采用核偏最小二乘法在高维空间建立实际铣削力系数关于铣削用量及其组合量的预测模型,分析该方法中核主元个数、高斯核函数核参数对预测模型精度的影响并确定其取值范围。最后分析考虑刀齿半径误差与不考虑时的铣削力系数,并比较核偏最小二乘预测方法与偏最小二乘预测方法,结果表明所提铣削力系数构造过程及预测方法具有较高的计算精度和预测能力。     

基于自适应动态无偏最小二乘支持向量机的刀具磨损预测建模

由于训练样本数量有限,滑动时间窗长度以及监测模型不能自适应调整和更新等因素,传统基于机器学习的刀具磨损预测模型存在精度和效率较低等问题,因此提出了一种基于自适应动态无偏最小二乘支持向量机（ADNLSSVM）的刀具磨损预测模型。采用公开数据库中的铣削加工数据集,通过时频域分析和小波包分解等手段从振动信号中提取特征量,并进一步利用相关性分析从中选择有效特征量作为模型输入。试验结果表明该方法所建模型具有较高的建模效率和预测精度。     

基于有限元仿真的拼接模具铣削用刀具优化

对ABAQUS软件进行了二次开发，以实现拼接模具铣削过程仿真前处理的快速建模。对铣削力、应力和刀具温度进行仿真，并与铣削实验结果对比，验证了仿真模型的准确性。对不同前角、后角、螺旋角及刃口半径的球头铣刀铣削拼接模具的过程进行模拟仿真，采用遗传算法优化铣刀结构，将优化后的结构参数与传统结构参数代入刀具磨损、工件表面质量的对比实验，从而验证了优化的有效性。研究表明，对仿真前处理进行快速建模的二次开发运行成功，模拟结果准确。研究结果为降低制造成本、提高铣刀寿命和工件表面质量提供了理论参考。     

Prediction of tool wear using regression and ANN models in end-milling operation

Tool wear prediction plays an important role in industry for higher productivity and product quality. Flank wear of cutting tools is often selected as the tool life criterion as it determines the diametric accuracy of machining, its stability and reliability. This paper focuses on two different models, namely, regression mathematical and artificial neural network (ANN) models for predicting tool wear. In the present work, flank wear is taken as the response (output) variable measured during milling, while cutting speed, feed and depth of cut are taken as input parameters. The Design of Experiments (DOE) technique is developed for three factors at five levels to conduct experiments. Experiments have been conducted for measuring tool wear based on the DOE technique in a universal milling machine on AISI 1020 steel using a carbide cutter. The experimental values are used in Six Sigma software for finding the coefficients to develop the regression model. The experimentally measured values are also used to train the feed forward back propagation artificial neural network (ANN) for prediction of tool wear. Predicted values of response by both models, i.e. regression and ANN are compared with the experimental values. The predictive neural network model was found to be capable of better predictions of tool flank wear within the trained range.     

结合域对抗自适应的刀具磨损预测方法

数控加工中存在刀具几何误差及安装误差、刀具及工件材料性能的随机波动等因素,导致刀具之间的磨损过程与监测信号上存在较大差异的问题,使得刀具磨损值难以精确预测。为此,本文提出了一种结合域对抗自适应的多尺度分布式卷积长短时记忆网络模型（Multiscale time-distributed convolutional long short-term memory,MTDCLSTM）。将加工过程中采集到的多传感器信号作为模型输入,通过域分类器与预测器之间的对抗学习,提取出可有效表征刀具磨损且与域无关的多尺度时空特征,经预测器的非线性映射,实现对刀具磨损值的精确预测。实验结果表明,结合域对抗自适应的MTDCLSTM模型预测性能明显优于分布式卷积神经网络、长短时记忆网络、卷积神经网络与支持向量机模型。与基于迁移成分分析的支持向量回归模型相比,本文模型的均方根误差与平均绝对误差分别降低了59.8%和62.5%,决定系数提高了66.1%,可有效缩小刀具个体之间的差异,提高磨损值预测精度。     

Surface generation modeling of micro milling process with stochastic tool wear

Micro milling is widely used to manufacture miniature parts and features at high quality with low set-up cost. To achieve a higher quality of existing micro products and improve the milling performance, a reliable analytical model of surface generation is the prerequisite as it offers the foundation for surface topography and surface roughness optimization. In the micro milling process, the stochastic tool wear is inevitable, but the deep influence of tool wear hasn't been considered in the micro milling process operation and modeling. Therefore, an improved analytical surface generation model with stochastic tool wear is presented for the micro milling process. A probabilistic approach based on the particle filter algorithm is used to predict the stochastic tool wear progression, linking online measurement data of cutting forces and tool vibrations with the state of tool wear. Meanwhile, the influence of tool run-out is also considered since the uncut chip thickness can be comparable to feed per tooth compared with that in conventional milling. Based on the process kinematics, tool run-out and stochastic tool wear, the cutting edge trajectory for micro milling can be determined by a theoretical and empirical coupled method. At last, the analytical surface generation model is employed to predict the surface topography and surface roughness, along with the concept of the minimum chip thickness and elastic recovery. The micro milling experiment results validate the effectiveness of the presented analytical surface generation model under different machining conditions. The model can be a significant supplement for predicting machined surface prior to the costly micro milling operations, and provide a basis for machining parameters optimization.