WO2022143502A1 - 一种滚刀性能退化趋势评估方法 - Google Patents

一种滚刀性能退化趋势评估方法 Download PDF

Info

Publication number
WO2022143502A1
WO2022143502A1 PCT/CN2021/141537 CN2021141537W WO2022143502A1 WO 2022143502 A1 WO2022143502 A1 WO 2022143502A1 CN 2021141537 W CN2021141537 W CN 2021141537W WO 2022143502 A1 WO2022143502 A1 WO 2022143502A1
Authority
WO
WIPO (PCT)
Prior art keywords
hob
vibration signal
performance degradation
frequency band
resonance frequency
Prior art date
Application number
PCT/CN2021/141537
Other languages
English (en)
French (fr)
Inventor
从飞云
周琦皓
陈立
林枫
Original Assignee
浙江大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 浙江大学 filed Critical 浙江大学
Publication of WO2022143502A1 publication Critical patent/WO2022143502A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M7/00—Vibration-testing of structures; Shock-testing of structures
    • G01M7/08—Shock-testing
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts

Definitions

  • the invention relates to the technical field of signal processing, in particular to a method for evaluating the performance degradation trend of a hob based on the energy of a vibration signal resonance frequency band.
  • Gear hobbing machines are widely used in various machinery manufacturing industries such as automobiles, tractors, machine tools, construction machinery, mining machinery, metallurgical machinery, petroleum, instruments, and aircraft and spacecraft. It is the most widely used machine tool in gear processing machine tools. As one of the key components of a gear hobbing machine, the tool is undoubtedly the top priority. During the working process of the gear hobbing machine, the tool cuts and squeezes the workpiece, which is one of the most easily worn parts. The tool wear state not only directly affects the surface quality and dimensional accuracy of the workpiece, but also determines the timing of tool change in batch processing.
  • the tool performance degradation tracking relies on feature extraction technology, mainly by extracting the signal obtained by the sensor through a certain signal processing method to obtain features that can describe the degradation process; for example, Chinese patent document CN111967640A discloses a method that considers the amount of tool wear.
  • Chinese patent document CN110119551A discloses a machine learning-based tool wear degradation correlation feature analysis method for shield machine tools, which is obtained by training the LightGBM feature sorting model through data resource sets. , the data resource set contains all the features except the over-cumulative feature and the feature affected by the control of the shield machine driver.
  • the research in the prior art has a weak ability to characterize the degradation process of the features extracted by the hob.
  • the device signal is collected by the sensor, there is often noise interference, which makes the features extracted from the signal robust.
  • the research on the performance degradation method of hob based on strong robustness and characterization ability is less involved.
  • the purpose of the present invention is to provide a method for evaluating the performance degradation trend of the hob.
  • the method starts from the energy of the resonance frequency band and uses the root mean square value as the standard to measure the performance degradation of the hob, and has better sensitivity to the performance degradation process of the hob. This improves the ability of vibration signal features to characterize the degradation process.
  • the present invention provides the following technical solutions:
  • the present invention provides a method for evaluating the performance degradation trend of a hob, comprising the following steps:
  • Feature extraction is carried out on the vibration signal in the process of hob performance degradation, preferably the resonance frequency band;
  • step (1) determining the preferred resonance frequency band and preferred center frequency of the hob is as follows:
  • n is the number of sequences in the initial data sequence of the vibration signal
  • k is the spectral line sequence number of the preferred resonance frequency band
  • j is the imaginary unit
  • ⁇ is a constant with a value of 3.1415926
  • e is a constant with a value of 2.718281828459;
  • the frequency with the largest corresponding amplitude is selected as the natural frequency f c' of the hob, and the half-power point is found on the ordinate according to the vibration signal spectrum diagram, that is, the peak value of the vibration signal spectrum sk A horizontal line is drawn along the abscissa through this value, and the intersection point of the horizontal line and the spectrum curve in the vibration signal spectrogram is set as the lower limit frequency ⁇ 1 of the calculation band and the upper limit frequency ⁇ 2 of the calculation band, and the calculation formula of the calculation bandwidth B w is :
  • i is the spectral line sequence number in the resonance frequency band
  • Calculate the number of spectral lines for the bandwidth sk is the vibration signal spectrum;
  • fs is the sampling frequency
  • m is the sequence number corresponding to the largest root mean square value in the root mean square value sequence
  • B w is the computational bandwidth
  • N is the number of vibration signal sampling points.
  • step (3) the feature extraction is the root mean square value of the vibration signal in the preferred resonance frequency band, and this value is also used as the hob performance degradation index;
  • Step (4) The specific process of constructing the hob performance degradation index p is as follows:
  • the impact energy generated by the contact surface between the hob and the workpiece will increase, and the corresponding RMS value in the preferred resonance frequency band will also increase. Therefore, the average value of the vibration signal in the preferred resonance frequency band will also increase.
  • the square root value is used as an indicator of the performance degradation of the hob.
  • the calculation formula of the performance degradation index is as follows:
  • k is the spectral line serial number of the preferred resonance frequency band; preferably the lower limit frequency of the resonance frequency band Preferred upper limit frequency of resonance frequency band Calculate the number of spectral lines for the bandwidth sk is the vibration signal spectrum;
  • k is the spectral line serial number of the preferred resonance frequency band
  • j is an imaginary unit
  • ⁇ is a constant with a value of 3.1415926
  • e is a constant with a value of 2.718281828459;
  • n is the number of sequences in the initial data sequence of the vibration signal
  • N is the number of sampling points
  • B 1 is the lower limit frequency of the preferred resonance frequency band
  • B 2 is the upper limit frequency of the preferred resonance frequency band
  • A is the number of spectral lines to calculate the bandwidth
  • sk is the vibration signal spectrum
  • k is the spectral line sequence number of the preferred resonance frequency band
  • d is the workpiece serial number.
  • the present invention provides a hob performance degradation trend evaluation method, which is based on the frequency spectrum resonance frequency band of the spindle vibration signal during the hob cutting process, greatly reduces the vibration signal error caused by human error and environmental factors, and can effectively avoid signal acquisition.
  • the impact of medium and high peak pulses on performance degradation indicators can effectively improve the accuracy of performance degradation indicators;
  • the present invention uses the root mean square value of a certain spectrum range to adaptively determine the preferred resonance frequency band, eliminates the influence of noise and human interference, and improves the sensitivity of the characteristic information; extracts the vibration signal spectrum
  • the root mean square of the preferred resonance frequency band The value is used as the hob degradation feature, which has a high sensitivity to the degradation process, which improves the feature's ability to characterize the hob degradation process.
  • the method proposed in the present invention has a certain reference significance for the effective extraction of hob degradation features, and also provides a certain promotion significance for further realizing the tool replacement strategy formulation and process adjustment based on hob state detection.
  • FIG. 1 is a flow chart of the method of the present invention.
  • FIG. 2 is a schematic diagram of an energy harvesting and performance degradation evaluation device of the present invention.
  • FIG. 3 is a schematic diagram of the installation of the energy harvesting and performance degradation evaluation device of the present invention on a gear hobbing machine
  • Figure 4 is a flowchart of an algorithm for determining the preferred resonance frequency band and the preferred center frequency.
  • FIG. 5 is a time domain waveform diagram of the pulse vibration signal to be processed.
  • FIG. 6 is a frequency domain waveform diagram of the pulse vibration signal to be processed.
  • FIG. 7 is a waveform diagram of the root mean square value corresponding to the preferred resonance frequency band under different center frequencies.
  • Figure 8 is a sequence diagram of the performance degradation index in the whole life cycle of the hob.
  • a method for evaluating the trend of hob performance degradation includes the following steps:
  • this embodiment also provides a vibration signal-based optimal resonance frequency band energy collection and performance degradation evaluation device, so as to realize the z-direction vibration signal acquisition and real-time performance degradation evaluation of the hob spindle;
  • the device includes: vibration Acceleration sensor 1, data acquisition card 2, personal computer 3, acquisition system 4; the vibration acceleration sensor is installed on the hob spindle of the gear hobbing machine to obtain the vibration signals of pulse excitation and hob cutting on the hob spindle respectively; data acquisition card
  • the vibration signal collected by the vibration acceleration sensor is transmitted to the computer; the acquisition system displays and stores the vibration signal in real time.
  • the installation of a vibration signal-based optimal resonance frequency band energy collection and performance degradation evaluation device on a gear hobbing machine is as follows:
  • the vibration acceleration sensor 1 is fixed on the hob spindle in the z direction by magnetic adsorption. , adjust the position so that it is fixed above the hob processing position, and check the installation stability of the vibration sensor 1.
  • Use a plastic hose to wrap the connection wire of the vibration sensor 1 to guide the arrangement of the connection wire to prevent the connection wire from entering the processing area and causing danger during processing.
  • Use insulating tape to wrap the end of the vibration sensor 1 to prevent damage to the vibration sensor 1 caused by cutting splashes during processing.
  • the other end of the connection line of the vibration sensor 1 extends to the workbench and is connected to the data acquisition card 2 .
  • n is the number of sequences in the initial data sequence of the vibration signal
  • k is the spectral line sequence number of the preferred resonance frequency band
  • j is the imaginary unit
  • ⁇ is a constant with a value of 3.1415926
  • e is a constant with a value of 2.718281828459;
  • the frequency with the largest corresponding amplitude is selected as the natural frequency f c' of the hob, and the half-power point is found on the ordinate according to the vibration signal spectrum diagram, that is, the peak value of the vibration signal spectrum sk A horizontal line is drawn along the abscissa through this value, and the intersection point of the horizontal line and the spectrum curve in the vibration signal spectrogram is set as the lower limit frequency ⁇ 1 of the calculation band and the upper limit frequency ⁇ 2 of the calculation band, and the calculation formula of the calculation bandwidth B w is :
  • i is the spectral line sequence number in the resonance frequency band
  • Calculate the number of spectral lines for the bandwidth sk is the vibration signal spectrum;
  • fs is the sampling frequency
  • m is the sequence number corresponding to the largest root mean square value in the root mean square value sequence
  • B w is the computational bandwidth
  • N is the number of vibration signal sampling points.
  • the maximum root mean square value indicates that under the same bandwidth, the energy in the frequency band corresponding to this center frequency is the largest, that is, the preferred center frequency and the preferred resonance frequency band.
  • the z-direction vibration acceleration signal of the main shaft of the hob during the whole life cycle of the hob is collected, and a part processing cycle is set as a set of data.
  • the sampling point set with a certain number of elements in the stable cutting stage of each set of data is selected as the initial data, which can better characterize the hob cutting law of the hob.
  • the impact energy generated by the contact surface between the hob and the workpiece will increase, and the corresponding RMS value in the preferred resonance frequency band will also increase. Therefore, the average value of the vibration signal in the preferred resonance frequency band will also increase.
  • the square root value is used as an indicator of the performance degradation of the hob.
  • the calculation formula of the performance degradation index is:
  • k is the spectral line serial number of the preferred resonance frequency band; preferably the lower limit frequency of the resonance frequency band Preferred upper limit frequency of resonance frequency band Calculate the number of spectral lines for the bandwidth sk is the vibration signal spectrum;
  • k is the spectral line serial number of the preferred resonance frequency band
  • j is an imaginary unit
  • ⁇ is a constant with a value of 3.1415926
  • e is a constant with a value of 2.718281828459;
  • n is the number of sequences in the initial data sequence of the vibration signal
  • N is the number of sampling points
  • B 1 is the lower limit frequency of the preferred resonance frequency band
  • B 2 is the upper limit frequency of the preferred resonance frequency band
  • A is the number of spectral lines to calculate the bandwidth
  • sk is the vibration signal spectrum
  • k is the spectral line sequence number of the preferred resonance frequency band
  • d is the workpiece serial number.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

一种滚刀性能退化趋势评估方法，包括如下步骤：进行滚刀主轴冲击试验(S1)；利用均方根值法确定滚刀主轴冲击试验中的优选共振频带B和优选中心频率f c(S2)；对滚刀性能退化过程中的振动信号共振频带进行特征提取(S3)；构建滚刀性能退化指标p(S4)；计算滚刀全寿命周期的性能退化指标序列P(S5)；将性能退化指标序列P与常用时域指标序列对比评估提出的性能退化指标；以滚刀切削过程中主轴振动信号频谱共振频带为基础，可以有效避免采集信号中高峰值脉冲对性能退化指标的影响。

Description

一种滚刀性能退化趋势评估方法 技术领域
本发明涉及信号处理技术领域，具体涉及一种基于振动信号共振频段能量的滚刀性能退化趋势评估方法。
背景技术
滚齿机广泛应用汽车、拖拉机、机床、工程机械、矿山机械、冶金机械、石油、仪表、飞机航天器等各种机械制造业，是齿轮加工机床中应用最广泛的一种机床。作为滚齿机的关键部件之一，刀具无疑是重中之重。滚齿机工作过程中刀具对工件进行切削、挤压，是最易磨损的部件之一。刀具磨损状态不仅直接影响加工工件的表面质量和尺寸精度，也决定着批量化加工工序中换刀的时机。通过对滚刀磨损状态的监控和评估，一方面能够有效有效避免刀具磨损带来的加工误差，改善工件的加工精度，另一方面能够实现换刀时机的优选与控制，从而提高加工效率，降低加工成本，能够实现较大的经济效益。因此基于状态监测方法提出有效评价滚刀磨损程度性能退化指标对于监测刀具磨损和设备运行状态具有重要的意义。
滚刀从正常状态到退化直至失效要经历一系列性能退化状态。迄今为止，已经提出了多种方法并应用于刀具性能退化领域。现有技术有的是通过对退化数据进行建模分析，但是目前存在着退化数据难以获得，获取退化数据时可能影响设备自身退化情况，测量精度不足等不利影响；现有技术有的是通过实时反映设备健康状态的状态监测方法，刀具性能退化跟踪依赖于特征提取技术，主要是将传感器获得的 信号通过一定的信号处理方法提取得到可以描述退化过程的特征；例如中国专利文献CN111967640A公开了一种考虑刀具磨损量和表面粗糙度的刀具剩余寿命预测方法，通过建立基于非线性Wiener过程刀具磨损退化模型、表面粗糙度退化模型，并采用Copula函数建立考虑二者相关性的多退化指标刀具剩余寿命预测模型，实现同时考虑刀具磨损量和表面粗糙度的刀具剩余寿命预测；例如中国专利文献CN110119551A公开了基于机器学习的盾构机刀具磨损退化关联特征分析方法，通过数据资源集对LightGBM特征排序模型进行训练获取的，数据资源集包含有除过累计量特征和受盾构机司机控制影响的特征的其他所有特征。
总体来说，现有技术中的研究对滚刀提取出的特征对退化过程的表征能力较弱，同时通过传感器采集设备信号时，往往会存在噪声干扰而导致从信号中提取得到的特征鲁棒性差的问题，而对基于具有较强鲁棒性和表征能力的滚刀性能退化方法研究涉及较少。
发明内容
本发明的目的是提供一种滚刀性能退化趋势评估方法，该方法从共振频段的能量出发，以均方根值作为衡量滚刀性能退化的标准，对滚刀性能退化过程具有较好的敏感性，提高了振动信号特征对退化过程的表征能力。
为了实现上述目的本发明提供以下技术方案：
本发明提供一种滚刀性能退化趋势评估方法，包括如下步骤：
(1)进行滚刀主轴锤击法共振模态试验；
(2)利用均方根值法确定滚刀主轴锤击法振动模态试验中的优选共振频带B和优选中心频率f c；
(3)对滚刀性能退化过程中的振动信号优选共振频带进行特征提取；
(4)构建滚刀性能退化指标p；
(5)计算滚刀全寿命周期的性能退化指标序列P；
步骤(1)确定滚刀优选共振频带和优选中心频率的具体过程如下：
设置冲击试验中采集到的振动信号采样点数为N，采样频率为f s，对对冲击试验过程中采集到的振动信号进行离散傅里叶变换计算振动信号频谱s k并绘制振动信号频谱图，振动信号频谱s k的计算公式如下：
Figure PCTCN2021141537-appb-000001
其中，n是振动信号初始数据序列中的序列个数；
k是优选共振频带的谱线序列号；
j是虚数单位；
π是常数，数值为3.1415926；
e是常数，数值为2.718281828459；
所述均方根值法的具体计算过程如下：
根据振动信号离散傅里叶变换选取对应幅值最大的频率作为滚刀固有频率f c’，在根据所述振动信号频谱图的纵坐标上寻找半功率点，即振动信号频谱s k峰值的
Figure PCTCN2021141537-appb-000002
并过此值沿横坐标作一水平线，所述水 平线与振动信号频谱图中频谱曲线的交点设为计算频带的下限频率σ 1和计算频带的上限频率σ 2，计算带宽B w的计算公式为：
B w＝σ 2-σ 1；
设置计算频带为[sp，sp+B w]，以sp为迭代变量，sp的初始值为0，设置迭代步长step＝f s/N，迭代变量sp的范围为[0，f s/2-B w]，计算振动信号在计算频带[sp，sp+B w]上的多个均方根值，并得到均方根值序列，所述均方根值序列对一个振动信号取不同计算频带，每迭代(即为取不同计算频带)一次产生一个均方根值；
均方根值计算公式为：
Figure PCTCN2021141537-appb-000003
其中：i是共振频带中的谱线序列号；
X rms(i)为均方根值序列中第i个值，i＝1，2，3，…，I；
Figure PCTCN2021141537-appb-000004
Figure PCTCN2021141537-appb-000005
计算带宽的谱线数
Figure PCTCN2021141537-appb-000006
s k为振动信号频谱；
均方根值序列中最大均方根值R的计算公式如下：
Figure PCTCN2021141537-appb-000007
设置均方根值序列的最大均方根值为X rms(m)，则优选中心频率的计算公式为：
Figure PCTCN2021141537-appb-000008
优选共振频带B带宽的计算公式为：
Figure PCTCN2021141537-appb-000009
其中，f s是采样频率；
m是均方根值序列中最大均方根值所对应的序列号；
B w是计算带宽；
N是振动信号采样点数。
步骤(3)中，特征提取是振动信号在优选共振频带的均方根值，也将此值作为滚刀性能退化指标；
步骤(4)构建滚刀性能退化指标p具体过程如下：
采集滚齿机加工过程滚刀全生命周期主轴z向振动加速度信号，设置一个零件加工周期为一组数据；选取每组数据切削稳定阶段一定元素数量的采样点集合作为初始数据，能较好得表征滚刀的滚切规律。
随着滚刀磨损程度的增大，滚刀与工件接触表面产生的冲击能量将提高，相应的在优选共振频带上的均方根值也将增大，因此将振动信号在优选共振频带的均方根值作为滚刀性能退化指标。性能退化指标的计算公式如下：
Figure PCTCN2021141537-appb-000010
其中，k是优选共振频带的谱线序列号；优选共振频带的下限频率
Figure PCTCN2021141537-appb-000011
优选共振频带的上限频率
Figure PCTCN2021141537-appb-000012
计算带宽的谱线数
Figure PCTCN2021141537-appb-000013
s k为振动信号频谱；
步骤(5)中计算滚刀全寿命周期的性能退化指标序列P的具体过程如下：
测定待评估滚刀全寿命周期加工工件c个，工件序号为d，d＝1，...，c；设置X n＝(x n(1)，x n(2)，...，x n(c))为振动信号初始数据序列，其中x n(d)为第d个工件的振动信号初始数据；对振动信号初始数据序列进行离散傅里叶变换，得到频谱序列S k＝(s k(1)，s k(2)，…，s k(c))，其 中s k(d)为第d个工件的振动信号初始数据的频谱，振动信号初始数据的频谱s k(d)计算公式为：
Figure PCTCN2021141537-appb-000014
其中，k是优选共振频带的谱线序列号；
j为虚数单位；
π是常数，数值为3.1415926；
e是常数，数值为2.718281828459；
n是振动信号初始数据序列中的序列个数；
N是采样点数；
计算频谱序列S k在优选共振频带的均方根值可以得到性能退化指标序列P＝(p(1)，p(2)，…，p(c))，p(d)为加工第d个工件时滚刀的性能退化指标，性能退化指标序列的表达式如下：
Figure PCTCN2021141537-appb-000015
其中，B 1是优选共振频带的下限频率；
B 2是优选共振频带的上限频率；
A是计算带宽的谱线数；
s k是振动信号频谱；
k是优选共振频带的谱线序列号；
d是工件序号。
本发明具有以下有益技术效果：
(1)本发明提供一种滚刀性能退化趋势评估方法，以滚刀切削过程中主轴振动信号频谱共振频带为基础，大大减少了人为失误和环 境因素造成的振动信号误差，可以有效避免采集信号中高峰值脉冲对性能退化指标的影响，有效提高性能退化指标的准确性；
(2)本发明使用一定频谱范围的均方根值来自适应的确定优选共振频带，剔除了噪声和人为干扰的影响，提高了特征信息的敏感度；提取振动信号频谱优选共振频带的均方根值作为滚刀退化特征，对退化过程有较高的敏感度，改善了特征对滚刀退化过程的表征能力。
(3)本发明提出的方法对于滚刀退化特征的有效提取起到一定的参考意义，也为进一步实现基于滚刀状态检测的刀具更换策略制定及工艺调整等提供了一定的促进意义。
附图说明
图1为本发明的方法流程图。
图2是本发明的能量采集及性能退化评估装置示意图。
图3是本发明的能量采集及性能退化评估装置在滚齿机上的安装示意图
图4为确定优选共振频带和优选中心频率的算法流程图。
图5为待处理脉冲振动信号的时域波形图。
图6为待处理脉冲振动信号的频域波形图。
图7为不同中心频率下对应优选共振频带的均方根值的波形图。
图8为滚刀全寿命周期性能退化指标序列图。
具体实施方式
以下结合附图对本发明的具体实施方式做详细描述，应当指出的是，实施例只是对本发明的具体阐述，不应视为对本发明的限定，实 施例的目的是为了让本领域技术人员更好地理解和再现本发明的技术方案，本发明的保护范围仍应当以权利要求书所限定的范围为准。
如图1所示，一种滚刀性能退化趋势评估方法，包括如下步骤：
(1)进行滚刀主轴锤击法振动模态试验；
(2)利用均方根值法确定滚刀主轴锤击法振动模态试验中的优选共振频带B和优选中心频率f c；
(3)对滚刀性能退化过程中的振动信号共振频带进行特征提取；
(4)构建滚刀性能退化指标p；
(5)计算滚刀全寿命周期的性能退化指标序列P；
如图2所示，本实施例还提供了一种基于振动信号优选共振频带能量的采集及性能退化评估装置，以实现滚刀主轴z向振动信号采集和实时性能退化评估；该装置包括：振动加速度传感器1，数据采集卡2，个人电脑3，采集系统4；振动加速度传感器安装在滚齿机滚刀主轴上，用于分别获取脉冲激励和滚刀切削在滚刀主轴上的振动信号；数据采集卡将振动加速度传感器采集的振动信号传输到电脑上；采集系统对振动信号进行实时显示和存储。
如图3所示，本实施例提供的一种基于振动信号优选共振频带能量的采集及性能退化评估装置在滚齿机上安装示意如下：振动加速度传感器1通过磁力吸附方式固定于滚刀主轴z方向上，调整位置使其固定于滚刀加工位置上方，检查确认振动传感器1安装稳定性。使用塑料软管包裹振动传感器1连接线，引导连接线排布，防止加工过程中连接线进入加工区域引起危险。使用绝缘胶带包裹振动传感器1端 部，防止加工过程中切削飞溅对振动传感器1的损伤。振动传感器1连接线另一端延伸至工作台与数据采集卡2连接，数据采集卡2通过连接线与个人电脑3连接，将采集信号输入采集系统4。
如图4所示，获取优选共振频带和优选中心频率的具体过程如下：
用模态锤敲击滚刀，滚刀主轴上布置的加速度传感器获取其产生的脉冲振动信号，振动信号的时域波形图和频域波形图如图5-6，确定优选中心频率和优选共振频带的过程如下：
首先设置滚刀主轴冲击试验中采集到的振动信号采样点数为N，采样频率为f s，对冲击试验过程中采集到的振动信号进行离散傅里叶变换计算振动信号频谱s k并绘制振动信号频谱图，振动信号频谱s k的计算公式如下：
Figure PCTCN2021141537-appb-000016
其中，n是振动信号初始数据序列中的序列个数；
k是优选共振频带的谱线序列号；
j是虚数单位；
π是常数，数值为3.1415926；
e是常数，数值为2.718281828459；
其次计算优选共振频带和优选中心频率的算法描述如下：
所述均方根值法的具体计算过程如下：
根据振动信号离散傅里叶变换选取对应幅值最大的频率作为滚刀固有频率f c’，在根据所述振动信号频谱图的纵坐标上寻找半功率点，即振动信号频谱s k峰值的
Figure PCTCN2021141537-appb-000017
并过此值沿横坐标作一水平线，所述水 平线与振动信号频谱图中频谱曲线的交点设为计算频带的下限频率σ 1和计算频带的上限频率σ 2，计算带宽B w的计算公式为：
B w＝σ 2-σ 1；
设置计算频带为[sp，sp+B w]，以sp为迭代变量，sp的初始值为0，设置迭代步长step＝f s/N，迭代变量sp的范围为[0，f s/2-B w]，计算振动信号在计算频带[sp，sp+B w]上的多个均方根值，并得到均方根值序列，均方根值计算公式为：
Figure PCTCN2021141537-appb-000018
其中：i是共振频带中的谱线序列号；
X rms(i)为均方根值序列中第i个值，i＝1，2，3，…，I；
Figure PCTCN2021141537-appb-000019
Figure PCTCN2021141537-appb-000020
计算带宽的谱线数
Figure PCTCN2021141537-appb-000021
s k为振动信号频谱；
均方根值序列中最大均方根值R的计算公式如下：
Figure PCTCN2021141537-appb-000022
设置均方根值序列的最大均方根值为X rms(m)，则优选中心频率的计算公式为：
Figure PCTCN2021141537-appb-000023
优选共振频带B带宽的计算公式为：
Figure PCTCN2021141537-appb-000024
其中，f s是采样频率；
m是均方根值序列中最大均方根值所对应的序列号；
B w是计算带宽；
N是振动信号采样点数。
如图7所示，不同中心频率下对应频带的均方根值，最大均方根值说明在相同的带宽下这个中心频率对应的频带上能量最大，即为优选中心频率和优选共振频带。
采集滚齿机加工过程滚刀全生命周期主轴z向振动加速度信号，设置一个零件加工周期为一组数据。选取每组数据切削稳定阶段一定元素数量的采样点集合作为初始数据，能较好得表征滚刀的滚切规律。
随着滚刀磨损程度的增大，滚刀与工件接触表面产生的冲击能量将提高，相应的在优选共振频带上的均方根值也将增大，因此将振动信号在优选共振频带的均方根值作为滚刀性能退化指标。性能退化指标的计算公式为：
Figure PCTCN2021141537-appb-000025
其中，k是优选共振频带的谱线序列号；优选共振频带的下限频率
Figure PCTCN2021141537-appb-000026
优选共振频带的上限频率
Figure PCTCN2021141537-appb-000027
计算带宽的谱线数
Figure PCTCN2021141537-appb-000028
s k为振动信号频谱；
测定待评估滚刀全寿命周期加工工件c个，工件序号为d，d＝1，...，c；设置X n＝(x n(1)，x n(2)，...，x n(c))为振动信号初始数据序列，其中x n(d)为第d个工件的振动信号初始数据；对振动信号初始数据序列进行离散傅里叶变换，得到频谱序列S k＝(s k(1)，s k(2)，…，s k(c))，其中s k(d)为第d个工件的振动信号初始数据的频谱，测定待评估滚刀全寿命周期加工工件c个，工件序号为d，d＝1，...，c；设置X n＝(x n(1)，x n(2)，...，x n(c))为振动信号初始数据序列，其中x n(d)为第d个工件的振动信号初始数据；对振动信号初始数据序列X n进行离散傅里叶 变换，得到频谱序列S k＝(s k(1)，s k(2)，…，s k(c))，其中s k(d)为第d个工件的振动信号初始数据的频谱，振动信号初始数据的频谱s k(d)计算公式为：
Figure PCTCN2021141537-appb-000029
其中，k是优选共振频带的谱线序列号；
j为虚数单位；
π是常数，数值为3.1415926；
e是常数，数值为2.718281828459；
n是振动信号初始数据序列中的序列个数；
N是采样点数；
计算频谱序列S k在优选共振频带的均方根值，得到性能退化指标序列P＝(p(1)，p(2)，…p(c))；p(d)为加工第d个工件时滚刀的性能退化指标，c为加工工件的个数，性能退化指标p(d)的计算公式如下：
Figure PCTCN2021141537-appb-000030
其中，B 1是优选共振频带的下限频率；
B 2是优选共振频带的上限频率；
A是计算带宽的谱线数；
s k是振动信号频谱；
k是优选共振频带的谱线序列号；
d是工件序号。
如图8所示，计算的滚刀全寿命周期性能退化指标序列。
以上所述，仅为本发明较佳的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，根据本发明的技术方案及其发明构思加以等同替换或改变，都应涵盖在本发明的保护范围之内。

Claims (4)

  1. 一种滚刀性能退化趋势评估方法，其特征是，包括如下步骤：
    (1)进行滚刀主轴锤击法振动模态试验；
    (2)利用均方根值法确定滚刀主轴锤击法振动模态试验中的优选共振频带B和优选中心频率f c；
    (3)对滚刀性能退化过程中的振动信号优选共振频带进行特征提取；
    (4)构建滚刀性能退化指标p；
    (5)计算滚刀全寿命周期的性能退化指标序列P；
  2. 根据权利要求1所述的一种滚刀性能退化趋势评估方法，其特征是，步骤(2)中确定优选共振频带的带宽B和优选中心频率f c的具体过程如下：
    设置滚刀主轴冲击试验中采集到的振动信号采样点数为N，采样频率为f s，对冲击试验过程中采集到的振动信号(振动信号初始数据x n)进行离散傅里叶变换计算振动信号频谱s k并绘制振动信号频谱图，振动信号频谱s k的计算公式如下：
    Figure PCTCN2021141537-appb-100001
    其中，n是振动信号初始数据序列中的序列个数；
    k是优选共振频带的谱线序列号；
    j是虚数单位；
    π是常数，数值为3.1415926；
    e是常数，数值为2.718281828459；
    所述均方根值法的具体计算过程如下：
    根据振动信号离散傅里叶变换选取对应幅值最大的频率作为滚刀固有频率f c’，在根据所述振动信号频谱图的纵坐标上寻找半功率点，即振动信号频谱s k峰值的
    Figure PCTCN2021141537-appb-100002
    并过此值沿横坐标作一水平线，所述水平线与振动信号频谱图中频谱曲线的交点设为计算频带的下限频率σ 1和计算频带的上限频率σ 2，计算带宽B w的计算公式为：
    B w＝σ 2-σ 1；
    设置计算频带为[sp，sp+B w]，以sp为迭代变量，sp的初始值为0，设置迭代步长step＝f s/N，迭代变量sp的范围为[0，f s/2-B w]，计算振动信号在计算频带[sp，sp+B w]上的多个均方根值，并得到均方根值序列，均方根值计算公式为：
    Figure PCTCN2021141537-appb-100003
    其中：i是共振频带中的谱线序列号；
    X rms(i)为均方根值序列中第i个值，i＝1，2，3，…，I；
    Figure PCTCN2021141537-appb-100004
    Figure PCTCN2021141537-appb-100005
    计算带宽的谱线数
    Figure PCTCN2021141537-appb-100006
    s k为振动信号频谱；
    均方根值序列中最大均方根值R的计算公式如下：
    Figure PCTCN2021141537-appb-100007
    设置均方根值序列的最大均方根值为X rms(m)，则优选中心频率的计算公式为：
    Figure PCTCN2021141537-appb-100008
    优选共振频带B带宽的计算公式为：
    Figure PCTCN2021141537-appb-100009
    其中，f s是采样频率；
    m是均方根值序列中最大均方根值所对应的序列号；
    B w是计算带宽；
    N是振动信号采样点数。
  3. 根据权利要求1所述的滚刀性能退化趋势评估方法，其特征是，步骤(4)所述的滚刀性能退化指标p计算公式如下：
    Figure PCTCN2021141537-appb-100010
    其中，k是优选共振频带的谱线序列号；优选共振频带的下限频率
    Figure PCTCN2021141537-appb-100011
    优选共振频带的上限频率
    Figure PCTCN2021141537-appb-100012
    计算带宽的谱线数
    Figure PCTCN2021141537-appb-100013
    s k为振动信号频谱。
  4. 根据权利要求1所述的滚刀性能退化趋势评估方法，其特征是，步骤(5)中计算滚刀全寿命周期的性能退化指标序列P的具体过程如下：
    测定待评估滚刀全寿命周期加工工件c个，工件序号为d，d＝1，...，c；设置X n＝(x n(1)，x n(2)，...，x n(c))为振动信号初始数据序列，其中x n(d)为第d个工件的振动信号初始数据；对振动信号初始数据序列X n进行离散傅里叶变换，得到频谱序列S k＝(s k(1)，s k(2)，…，s k(c))，其中s k(d)为第d个工件的振动信号初始数据的频谱，振动信号初始数据的频谱s k(d)计算公式为：
    Figure PCTCN2021141537-appb-100014
    其中，k是优选共振频带的谱线序列号；
    j为虚数单位；
    π是常数，数值为3.1415926；
    e是常数，数值为2.718281828459；
    n是振动信号初始数据序列中的序列个数；
    N是采样点数；
    计算频谱序列S k在优选共振频带的均方根值，得到性能退化指标序列P＝(p(1)，p(2)，…p(c))；p(d)为加工第d个工件时滚刀的性能退化指标，c为加工工件的个数，性能退化指标p(d)的计算公式如下：
    Figure PCTCN2021141537-appb-100015
    其中，B 1是优选共振频带的下限频率；
    B 2是优选共振频带的上限频率；
    A是计算带宽的谱线数；
    s k是振动信号频谱；
    k是优选共振频带的谱线序列号；
    d是工件序号。
PCT/CN2021/141537 2020-12-29 2021-12-27 一种滚刀性能退化趋势评估方法 WO2022143502A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011589950.4 2020-12-29
CN202011589950.4A CN112781820B (zh) 2020-12-29 2020-12-29 一种滚刀性能退化趋势评估方法

Publications (1)

Publication Number Publication Date
WO2022143502A1 true WO2022143502A1 (zh) 2022-07-07

Family

ID=75753242

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/141537 WO2022143502A1 (zh) 2020-12-29 2021-12-27 一种滚刀性能退化趋势评估方法

Country Status (2)

Country Link
CN (1) CN112781820B (zh)
WO (1) WO2022143502A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112781820B (zh) * 2020-12-29 2022-01-11 浙江大学 一种滚刀性能退化趋势评估方法
DE102022107444A1 (de) * 2022-03-29 2023-10-05 MTU Aero Engines AG Zerspanungsmaschine und Verfahren zum Überwachen einer dynamischen Steifigkeit einer Zerspanungsmaschine
CN114812484B (zh) * 2022-03-30 2024-02-13 中国有研科技集团有限公司 一种楔焊劈刀有效寿命的高效检验方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1934433A (zh) * 2004-03-31 2007-03-21 中国电力股份有限公司 滚动轴承的剩余寿命诊断方法及剩余寿命诊断装置
CN109596354A (zh) * 2018-12-21 2019-04-09 电子科技大学 基于自适应共振频带识别的带通滤波方法
US20190301975A1 (en) * 2018-03-30 2019-10-03 Okuma Corporation Abnormality diagnostic method and abnormality diagnostic device for rolling bearing
CN112781820A (zh) * 2020-12-29 2021-05-11 浙江大学 一种滚刀性能退化趋势评估方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103802022B (zh) * 2014-03-04 2016-02-03 上海理工大学 数控机床动态特性衰变的评估方法
CN104568444B (zh) * 2015-01-28 2017-02-22 北京邮电大学 变转速火车滚动轴承故障特征频率提取方法
EP3462764B1 (en) * 2017-09-29 2021-12-01 Ordnance Survey Limited Radio frequency propagation simulation tool
CN110119551B (zh) * 2019-04-29 2022-12-06 西安电子科技大学 基于机器学习的盾构机刀具磨损退化关联特征分析方法
CN110370080A (zh) * 2019-07-19 2019-10-25 广东寰球智能科技有限公司 一种修边机切刀的监测方法及监测系统
CN111476430A (zh) * 2020-04-21 2020-07-31 南京凯奥思数据技术有限公司 一种基于机器学习回归算法的刀具剩余寿命预测方法
CN111644900B (zh) * 2020-05-21 2021-11-09 西安交通大学 一种基于主轴振动特征融合的刀具破损实时监测方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1934433A (zh) * 2004-03-31 2007-03-21 中国电力股份有限公司 滚动轴承的剩余寿命诊断方法及剩余寿命诊断装置
US20190301975A1 (en) * 2018-03-30 2019-10-03 Okuma Corporation Abnormality diagnostic method and abnormality diagnostic device for rolling bearing
CN109596354A (zh) * 2018-12-21 2019-04-09 电子科技大学 基于自适应共振频带识别的带通滤波方法
CN112781820A (zh) * 2020-12-29 2021-05-11 浙江大学 一种滚刀性能退化趋势评估方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
YANG ZHIFEI, ET AL.: "Bearing Fault Diagnosis Based on Variable Mode Decomposition and Entropy Value Method ", MODULAR MACHINE TOOL & AUTOMATIC MANUFACTURING TECHNIQUE, no. 11, 30 November 2017 (2017-11-30), pages 78 - 80, XP055947195, ISSN: 1001-2265, DOI: 10.13462/j.cnki.mmtamt.2017.11.020 *
YANG ZHIFEI: "Research on Fault Feature Extraction and Residual Life Prediction of Rolling Bearings", MASTER THESIS, TIANJIN POLYTECHNIC UNIVERSITY, CN, no. 2, 15 February 2018 (2018-02-15), CN , XP055947192, ISSN: 1674-0246 *

Also Published As

Publication number Publication date
CN112781820A (zh) 2021-05-11
CN112781820B (zh) 2022-01-11

Similar Documents

Publication Publication Date Title
WO2022143502A1 (zh) 一种滚刀性能退化趋势评估方法
Cui et al. Quantitative trend fault diagnosis of a rolling bearing based on Sparsogram and Lempel-Ziv
CN103575523B (zh) 基于FastICA-谱峭度-包络谱分析的旋转机械故障诊断方法
US7341410B2 (en) Dynamical instrument for machining
US6640205B2 (en) Method and device for investigating and identifying the nature of a material
CN105058165A (zh) 基于振动信号的刀具磨损量监测系统
CN113375939B (zh) 基于svd和vmd的机械件故障诊断方法
CN113343928B (zh) 变速路段高速铁路钢轨波磨检测方法及装置、计算机设备
CN107350900A (zh) 一种基于断屑时间提取的刀具状态监测方法
CN110346130A (zh) 一种基于经验模态分解和时频多特征的镗削颤振检测方法
CN107414600A (zh) 基于多传感器信号的内螺纹低频激振冷挤压机床的加工过程监测方法
CN112207631B (zh) 刀具检测模型的生成方法、检测方法、系统、设备及介质
Du et al. Intelligent turning tool monitoring with neural network adaptive learning
CN116432071A (zh) 一种滚动轴承剩余寿命预测方法
CN106338662A (zh) 基于数学形态学的变压器绕组变形诊断方法
CN103490830B (zh) 基于物联网电力测温设备的去噪声射频频谱峰值获取方法
CN114871850A (zh) 一种基于振动信号和bp神经网络的刀具磨损状态评估方法
CN102988041A (zh) 心磁信号噪声抑制中的信号选择性平均方法
CN114739671A (zh) 一种基于改进广义s变换的轴承故障诊断方法
CN112857806B (zh) 一种基于移动窗口时域特征提取的轴承故障检测方法
Pang et al. The evolved kurtogram: a novel repetitive transients extraction method for bearing fault diagnosis
CN111611533B (zh) 基于自适应vmd和改进功率谱的磨机负荷特征提取方法
CN111141830B (zh) 基于微纳耦合光纤传感器的线性定位系统及方法
CN110849928B (zh) 一种超声滚压加工温度测量分析方法
CN111898508B (zh) 一种基于听觉感知的电动冲击批零件缺损的检测方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21914239

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21914239

Country of ref document: EP

Kind code of ref document: A1