CN109605128B - Milling chatter online detection method based on power spectrum entropy difference - Google Patents

Milling chatter online detection method based on power spectrum entropy difference Download PDF

Info

Publication number
CN109605128B
CN109605128B CN201910019498.9A CN201910019498A CN109605128B CN 109605128 B CN109605128 B CN 109605128B CN 201910019498 A CN201910019498 A CN 201910019498A CN 109605128 B CN109605128 B CN 109605128B
Authority
CN
China
Prior art keywords
power spectrum
signal
spectrum entropy
flutter
milling
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN201910019498.9A
Other languages
Chinese (zh)
Other versions
CN109605128A (en
Inventor
李小虎
黄晓玮
万少可
苑俊朋
赵根
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
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 Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN201910019498.9A priority Critical patent/CN109605128B/en
Publication of CN109605128A publication Critical patent/CN109605128A/en
Application granted granted Critical
Publication of CN109605128B publication Critical patent/CN109605128B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
    • B23Q17/00—Arrangements for observing, indicating or measuring on machine tools
    • B23Q17/12—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring vibration

Abstract

The invention discloses a milling chatter online detection method based on power spectrum entropy difference, which comprises the steps of obtaining vibration information in a milling process through an acceleration sensor; decomposing the signal into a group of basic mode components by utilizing variation modal decomposition, and taking the basic mode components of the high-frequency part to reconstruct the signal to obtain the signal of the frequency band where the flutter component is located; and carrying out self-adaptive filtering on the reconstructed signal, calculating power spectrum entropy before and after signal filtering, and carrying out difference calculation on the obtained power spectrum entropy to obtain a power spectrum entropy difference so as to reflect the influence of the filtering on the frequency spectrum distribution of the flutter frequency band signal. Compared with the traditional flutter detection method, the method can effectively separate the flutter frequency band signals, the influence of threshold reflection filtering on different state signals has definite physical significance, the randomness of threshold selection can be avoided, the accuracy and reliability of milling flutter detection are improved, and the misjudgment rate and the missed judgment rate are reduced.

Description

Milling chatter online detection method based on power spectrum entropy difference
Technical Field
The invention belongs to the technical field of detection of machining states, and particularly relates to a milling chatter online detection method based on power spectrum entropy difference.
Background
Milling is the most common machining method, has the advantages of high efficiency, high precision, low cost and the like, and is widely applied to the manufacturing fields of aviation, aerospace, molds, automobiles and the like. Chatter is a strong self-excited vibration in the metal cutting process, which can cause the surface quality and size of a workpiece to be reduced, reduce the machining yield, cause the premature fatigue damage of machine tool parts and aggravate the tool abrasion. The selection of conservative machining parameters can avoid the outbreak of chatter vibration, but reduce the machining efficiency. Although the stability lobe graph of milling can be analyzed and obtained by an analytical method, and appropriate machining parameters are selected according to the stability lobe graph, due to model errors and time-varying characteristics of the milling process, the requirement of online detection of milling chatter vibration still exists.
The scholars at home and abroad carry out a lot of researches on the milling flutter detection, such as measuring the strength of periodic components in signals by utilizing the autocorrelation coefficient of vibration acceleration signals so as to judge the flutter state; separating the flutter component from the acquired acceleration signal by using a self-adaptive filter based on a minimum mean square error criterion, and measuring the strength of the flutter component in the original signal by using a variance ratio to realize the detection of early flutter; decomposing the vibration signal by using EEMD, screening out sensitive IMFs in the vibration signal, and calculating a normalized energy ratio, a variation coefficient and a center frequency to realize flutter detection; performing wavelet packet decomposition on the acquired signals, reconstructing flutter frequency band signals, extracting the energy ratio and the singular spectrum entropy coefficient of the frequency band signals, and training an artificial neural network model for identifying flutter; original data characteristics are extracted through a training stack type self-encoder, and the extracted characteristics are used for training a support vector machine to realize flutter detection.
Searching the existing documents shows that the flutter detection method can be divided into two categories, namely a pattern recognition method and a threshold value method. The pattern recognition method requires a large amount of labeled data for training; in the threshold method, the threshold value has no definite physical meaning, and the setting thereof requires a lot of experience and is blind. In the process of milling from stable development to flutter, signals are converted from narrow-band distribution of a low frequency band into wide-band distribution, and finally narrow-band distribution of a high frequency band is formed, so that a common basic mode component with the largest normalized energy ratio is selected as a sensitive basic mode component, and a flutter component can not be accurately screened out in the early stage of flutter.
Disclosure of Invention
The invention aims to provide a milling chatter detection method based on power spectrum entropy difference, which provides a threshold value with practical significance and avoids blindness in threshold value setting.
In order to achieve the purpose, the invention adopts the technical scheme that the milling chatter online detection method based on the power spectrum entropy difference comprises the following steps:
step 1, collecting an acceleration signal X ═ X in the milling process1,x2,…,xn]N represents a signal length;
step 2, carrying out variation modal decomposition on the acceleration signal X acquired in the step 1 and reconstructing flutter frequency band signals to obtain reconstructed signal Xrec；
Step 3, the reconstruction information obtained in the step 2Number XrecCarrying out self-adaptive filtering to obtain a reconstructed signal Xrec_filtered；
Step 4, respectively calculating the signals X obtained in the step 2recAnd the signal X obtained in step 3rec_filteredStarting from the mth frame data, the power spectrum entropy difference of the current frame for judging flutter is the mean value of the original power spectrum entropy differences of the previous m frames of the current frame, and a power spectrum entropy difference index is obtained;
and 5, judging the flutter state according to the power spectrum entropy difference index obtained in the step 4.
And acquiring an acceleration signal by adopting a three-way acceleration signal sensor.
The three-way acceleration signal sensor is installed at the end of the main shaft.
The three-direction acceleration signal sensor is connected with the data acquisition card, the data acquisition card is connected with the computer, the three-direction acceleration signal sensor transmits the monitored acceleration signals to the data acquisition card, and the data acquisition card transmits data to the computer.
In step 2, obtaining a group of basic mode components distributed from low frequency to high frequency through metamorphic modal decomposition, and recording as U-U1,u2,…,ukWhere k is the number of decomposition modalities; and reconstructing the flutter frequency band signal by using the high-frequency fundamental mode component of the middle-rear half part
Figure GDA0002328266650000031
Wherein [. ]]Indicating rounding.
In step 3, the reconstructed flutter frequency band signal is subjected to adaptive filtering under the minimum mean square error criterion to filter the frequency conversion and harmonic components thereof, and a filtered reconstructed signal X is obtainedrec_filtered。
In step 4, the power spectrum is calculated, then the probability density function of the power spectrum is calculated, and then the power spectrum entropy of the signal is calculated.
In step 4, the power spectrum entropy calculation method is as follows:
let Y ═ Y (k), k ═ 1,2, …, n } be a time series of length n, i.e. a frame XrecAnd Xrec_filtered(ii) a Fourier transform sequence thereof
Figure GDA0002328266650000032
Wherein
Figure GDA0002328266650000033
Figure GDA0002328266650000034
Represents an imaginary unit;
calculating a power spectrum s (j) of the signal, j representing the j-th value in the fourier transform sequence:
Figure GDA0002328266650000035
and [. cndot. ] represents rounding, and the probability density function of the power spectrum is calculated:
Figure GDA0002328266650000036
and [. cndot. ] represents rounding, and finally, the power spectrum entropy of the signal is calculated:
Figure GDA0002328266650000037
the power spectrum entropy difference obtained in the step 4 is as follows:
Figure GDA0002328266650000038
wherein m is 5-11.
a. When the power spectrum entropy difference is less than 0, the flutter frequency band signal is noise and weak frequency conversion and harmonic components, the frequency conversion and harmonic components are filtered to enable the frequency spectrum distribution of the signal to be uniform, the power spectrum entropy of the signal is enlarged after filtering, and the signal is in a stable milling state;
b. when the power spectrum entropy difference is larger than 0, the signal components of the flutter section are noise, frequency conversion and harmonic components and flutter components, the flutter components are amplified while the frequency conversion and harmonic components are filtered, the frequency spectrum distribution of the signal becomes concentrated, the power spectrum entropy of the signal becomes small after filtering, and the signal is in a flutter milling state.
Compared with the prior art, the invention has at least the following beneficial effects: the variation modal decomposition effectively relieves the modal aliasing phenomenon, accurately reconstructs the signal of the frequency band where the flutter exists, removes the stable components of the low frequency band with larger amplitude, improves the sensitivity of flutter detection, and the power spectrum entropy difference index measures the influence of filtering on the flutter frequency band signal, and has clear physical significance when 0 is used as a threshold value.
Compared with the traditional flutter detection method, the method disclosed by the invention separates the signal mainly reflecting the flutter information from the signal irrelevant to the flutter, the detection index threshold has definite physical significance, the blindness of threshold setting is avoided, and the sensitivity and accuracy of flutter detection are effectively improved.
Drawings
FIG. 1 is a schematic system diagram of the process of the present invention.
Fig. 2 is a schematic diagram of adaptive filtering.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
The milling chatter online detection method based on the power spectrum entropy difference comprises the following steps:
(1) signal acquisition:
as shown in fig. 1, the milling cutter 3 is clamped and mounted on the spindle 1 by the cutter handle 2, the three-way acceleration signal sensor 5 is mounted at the end of the spindle 1, the parameters of the three-way acceleration signal sensor 5 are shown in table 1, the three-way acceleration signal sensor 5 is matched with the data acquisition card 6, the parameters of the data acquisition card 6 are shown in table 2, the acceleration signal sensor 5 and the data acquisition card 6 are used for acquiring acceleration signals in the milling process, and the acquired acceleration signals are represented by X ═1,x2,…,xn]And n represents the length of the signal, and the acquired signal is transmitted to the computer 4 through a data line to realize the online detection of the milling vibration.
TABLE 1 acceleration sensor parameters
Figure GDA0002328266650000051
Table 2 data acquisition card parameters
Figure GDA0002328266650000052
(2) Carrying out variation modal decomposition on the signal X and reconstructing a flutter frequency band signal:
carrying out variation modal decomposition on the acquired original acceleration signal to obtain a group of fundamental mode components (IMFs) distributed from low frequency to high frequency, and recording the fundamental mode components as U-U1,u2,…,ukWherein k is the number of decomposition modes, and is usually 10-16; and reconstructing the flutter frequency band signal by using the high-frequency fundamental mode component of the middle-rear half part
Figure GDA0002328266650000053
Wherein [. ]]Indicating rounding.
(3) For the reconstructed signal XrecAnd (3) carrying out adaptive filtering:
the reconstructed flutter frequency band signal is subjected to self-adaptive filtering under the minimum mean square error criterion to filter the frequency conversion and harmonic components thereof, and a filtered reconstructed signal X is obtainedrec_filtered。
Figure 2 is a schematic diagram of an adaptive filter used in the present invention,
Figure GDA0002328266650000054
the filter consists of m groups of sine and cosine signals, and each group of sine and cosine signals forms a signal component to be filtered. m is the harmonic order and is typically taken as the smallest integer of the sampling frequency divided by the conversion frequency (i.e., the
Figure GDA0002328266650000055
fsSp is the main shaft rotating speed and the unit is revolution/min) of the signal, wherein the input signal corresponding to the ith harmonic wave is xi＝[xcixsi]＝[cos(i×w×nT)sin(i×w×nT)]W is the frequency conversion angular velocity, T is the sampling period, and n represents that the current time is the nth point in a frame signal;
Figure GDA0002328266650000056
at the nth point in a frame signal
Figure GDA0002328266650000057
The weight of (2), its initial value
Figure GDA0002328266650000058
Figure GDA0002328266650000059
The inner product of the input and the weight value represents the signal amplitude after the components needing to be filtered are mixed. Xrec_filtered(n)＝XrecAnd (n) -y (n) is the amplitude of the filtered signal. Accordingly, the weight at the n +1 th point is updated according to the formula
Figure GDA00023282666500000510
Where α is the step size of the filter,
(4) calculating the difference value of the power spectrum entropy of the reconstructed signal before and after filtering:
the variation range of the power spectrum entropy is [0,1 ]]The uniformity of the time series spectrum distribution is also reflected, the more uniform the distribution, the larger the power spectrum entropy, the X before and after filtering are respectively subjected torecAnd Xrec_filteredAnd calculating a power spectrum entropy index, and obtaining a power spectrum entropy difference index by difference, wherein the index reflects the influence of filtering on the frequency spectrum distribution of the signals of the flutter frequency band.
The power spectrum entropy is calculated as follows:
let Y ═ Y (k), k ═ 1,2, …, n be a time series of length n, i.e. a frame X in the methodrecAnd Xrec_filtered(ii) a Fourier transform sequence thereof
Figure GDA0002328266650000061
Wherein
Figure GDA0002328266650000062
Figure GDA0002328266650000063
Figure GDA0002328266650000064
Representing imaginary units.
Calculating a power spectrum s (j) of the signal, j representing the j-th value in the fourier transform sequence:
Figure GDA0002328266650000065
[·]represents rounding
Calculating the probability density function of the power spectrum:
Figure GDA0002328266650000066
[·]represents rounding
And finally, calculating the power spectrum entropy of the signal:
Figure GDA0002328266650000067
(6) judging the flutter state:
a. during stable milling, the flutter frequency band signal is noise and weak frequency conversion and harmonic components, the frequency conversion and harmonic components are filtered to enable the frequency spectrum distribution of the signal to be uniform, and the power spectrum entropy of the signal is increased after filtering;
b. when flutter occurs, the signal components of the flutter section are noise, frequency conversion and harmonic components and flutter components, the flutter components are amplified while the frequency conversion and harmonic components are filtered, the frequency spectrum distribution of the signal becomes centralized, and the power spectrum entropy of the signal becomes small after filtering;
starting from the m-th frame data, the power spectrum entropy difference of the current frame for flutter judgment is the mean value of the original power spectrum entropy differences of the previous m frames (including the current frame), and the influence of data fluctuation can be effectively eliminated, namely
Figure GDA0002328266650000071
In actual operation, m is 5-11.

Claims (10)

1. A milling chatter online detection method based on power spectrum entropy difference is characterized by comprising the following steps:
step 1, collecting an acceleration signal X ═ X in the milling process1,x2,…,xn]N represents a signal length;
step 2, carrying out variation modal decomposition on the acceleration signal X acquired in the step 1 and reconstructing flutter frequency band signals to obtain reconstructed signal Xrec；
Step 3, the reconstruction signal X obtained in the step 2recCarrying out self-adaptive filtering to obtain a reconstructed signal Xrec_filtered；
Step 4, respectively calculating the signals X obtained in the step 2recAnd the signal X obtained in step 3rec_filteredStarting from the mth frame data, the power spectrum entropy difference of the current frame for judging flutter is the mean value of the original power spectrum entropy differences of the previous m frames of the current frame, and a power spectrum entropy difference index is obtained;
and 5, judging the flutter state according to the power spectrum entropy difference index obtained in the step 4.
2. The milling chatter online detection method based on the power spectrum entropy difference as claimed in claim 1, wherein in step 1, a three-way acceleration signal sensor (5) is used for acquiring acceleration signals.
3. The milling chattering online detection method based on the power spectrum entropy difference as claimed in claim 2, wherein a three-way acceleration signal sensor (5) is installed at the end of the main shaft (1).
4. The milling chatter online detection method based on the power spectrum entropy difference as claimed in claim 3, wherein the three-way acceleration signal sensor (5) is connected to the data acquisition card (6), the data acquisition card (6) is connected to the computer (4), the three-way acceleration signal sensor (5) transmits the monitored acceleration signal to the data acquisition card (6), and the data acquisition card (6) transmits data to the computer (4).
5. The milling chattering online detection method based on the power spectrum entropy difference as claimed in claim 1, wherein in step 2, a group of basic mode components distributed from low frequency to high frequency is obtained through variation modal decomposition, and is recorded as U-U1,u2,…,ukWhere k is the number of decomposition modalities; and reconstructing the flutter frequency band signal by using the high-frequency fundamental mode component of the middle-rear half part
Figure FDA0002328266640000011
Wherein [. ]]Indicating rounding.
6. The milling flutter online detection method based on power spectrum entropy difference as claimed in claim 1, wherein in step 3, the reconstructed flutter frequency band signal is subjected to adaptive filtering under a minimum mean square error criterion to filter frequency conversion and harmonic components thereof, and a filtered reconstructed signal X is obtainedrec_filtered。
7. The milling chattering online detection method based on the power spectrum entropy difference as claimed in claim 1, wherein in step 4, the power spectrum is calculated first, then a probability density function of the power spectrum is calculated, and then a power spectrum entropy of the signal is calculated.
8. The milling chattering online detection method based on the power spectrum entropy difference as claimed in claim 7, wherein in the step 4, the power spectrum entropy is calculated as follows:
let Y ═ Y (k), k ═ 1,2, …, n } be a time series of length n, i.e. a frame XrecAnd Xrec_filtered(ii) a Fourier transform sequence thereof
Figure FDA0002328266640000021
Wherein
Figure FDA0002328266640000022
Figure FDA0002328266640000023
Represents an imaginary unit;
calculating a power spectrum s (j) of the signal, j representing the j-th value in the fourier transform sequence:
Figure FDA0002328266640000024
and [. cndot. ] represents rounding, and the probability density function of the power spectrum is calculated:
Figure FDA0002328266640000025
and [. cndot. ] represents rounding, and finally, the power spectrum entropy of the signal is calculated:
Figure FDA0002328266640000026
9. the milling chattering online detection method based on the power spectrum entropy difference as claimed in claim 8, wherein the power spectrum entropy difference obtained in step 4 is:
Figure FDA0002328266640000027
wherein m is 5-11.
10. The milling chattering online detection method based on power spectrum entropy difference as claimed in claim 1,
a. when the power spectrum entropy difference is less than 0, the flutter frequency band signal is noise and weak frequency conversion and harmonic components, the frequency conversion and harmonic components are filtered to enable the frequency spectrum distribution of the signal to be uniform, the power spectrum entropy of the signal is enlarged after filtering, and the signal is in a stable milling state;
b. when the power spectrum entropy difference is larger than 0, the signal components of the flutter section are noise, frequency conversion and harmonic components and flutter components, the flutter components are amplified while the frequency conversion and harmonic components are filtered, the frequency spectrum distribution of the signal becomes concentrated, the power spectrum entropy of the signal becomes small after filtering, and the signal is in a flutter milling state.
CN201910019498.9A 2019-01-09 2019-01-09 Milling chatter online detection method based on power spectrum entropy difference Active CN109605128B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910019498.9A CN109605128B (en) 2019-01-09 2019-01-09 Milling chatter online detection method based on power spectrum entropy difference

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910019498.9A CN109605128B (en) 2019-01-09 2019-01-09 Milling chatter online detection method based on power spectrum entropy difference

Publications (2)

Publication Number Publication Date
CN109605128A CN109605128A (en) 2019-04-12
CN109605128B true CN109605128B (en) 2020-03-31

Family

ID=66016710

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910019498.9A Active CN109605128B (en) 2019-01-09 2019-01-09 Milling chatter online detection method based on power spectrum entropy difference

Country Status (1)

Country Link
CN (1) CN109605128B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112394642B (en) * 2020-11-27 2022-09-20 华中科技大学 Robot milling chatter identification method based on super-parameter optimization
CN112496862B (en) * 2020-11-30 2021-10-29 上海交通大学 Milling flutter intelligent identification method based on theoretical model containing milling angle
CN112974945B (en) * 2021-03-19 2023-05-30 天津大学 Milling chatter monitoring method based on variation modal decomposition and tracking threshold
CN114800042B (en) * 2022-04-28 2023-05-05 华中科技大学 Robot milling vibration type identification method based on power spectrum entropy difference

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3621049A1 (en) * 1985-06-28 1987-01-02 Gen Electric ARRANGEMENT, MONITORING DEVICE AND METHOD FOR DETECTING THE INITIAL TOUCH BETWEEN A CUTTING TOOL AND A WORKPIECE
EP0714519A1 (en) * 1993-08-19 1996-06-05 Mallinckrodt Medical, Inc. Entropy based signal transmission, reception and signal analysis method and apparatus
CN103203670A (en) * 2013-04-03 2013-07-17 同济大学 Grinding chatter prediction method based on maximum information entropy and directional divergence
CN104390697A (en) * 2014-11-06 2015-03-04 西安交通大学 C0 complexity and correlation coefficient-based milling chatter detection method
CN105108584A (en) * 2015-07-21 2015-12-02 上海交通大学 Turning chatter detection method
CN106141815A (en) * 2016-07-15 2016-11-23 西安交通大学 A kind of high-speed milling tremor on-line identification method based on AR model
CN107330375A (en) * 2017-06-05 2017-11-07 天津大学 Lathe flutter on-line monitoring method based on empirical mode decomposition and Sample Entropy feature

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150304107A1 (en) * 2014-04-17 2015-10-22 International Business Machines Corporation Extracting Entropy from Machine Vibration

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3621049A1 (en) * 1985-06-28 1987-01-02 Gen Electric ARRANGEMENT, MONITORING DEVICE AND METHOD FOR DETECTING THE INITIAL TOUCH BETWEEN A CUTTING TOOL AND A WORKPIECE
EP0714519A1 (en) * 1993-08-19 1996-06-05 Mallinckrodt Medical, Inc. Entropy based signal transmission, reception and signal analysis method and apparatus
CN103203670A (en) * 2013-04-03 2013-07-17 同济大学 Grinding chatter prediction method based on maximum information entropy and directional divergence
CN104390697A (en) * 2014-11-06 2015-03-04 西安交通大学 C0 complexity and correlation coefficient-based milling chatter detection method
CN105108584A (en) * 2015-07-21 2015-12-02 上海交通大学 Turning chatter detection method
CN106141815A (en) * 2016-07-15 2016-11-23 西安交通大学 A kind of high-speed milling tremor on-line identification method based on AR model
CN107330375A (en) * 2017-06-05 2017-11-07 天津大学 Lathe flutter on-line monitoring method based on empirical mode decomposition and Sample Entropy feature

Also Published As

Publication number Publication date
CN109605128A (en) 2019-04-12

Similar Documents

Publication Publication Date Title
CN109605128B (en) Milling chatter online detection method based on power spectrum entropy difference
CN102091972B (en) Numerical control machine tool wear monitoring method
CN110263474A (en) A kind of cutter life real-time predicting method of numerically-controlled machine tool
CN106141815B (en) A kind of high-speed milling flutter on-line identification method based on AR models
CN109855874B (en) Random resonance filter for enhancing detection of weak signals in vibration assisted by sound
CN113780055B (en) MOMEDA and compression sensing rolling bearing fault diagnosis method
CN114619292B (en) Milling cutter wear monitoring method based on fusion of wavelet denoising and attention mechanism with GRU network
CN106112697A (en) A kind of milling parameter automatic alarm threshold setting method based on 3 σ criterions
CN102033108A (en) Method for extracting state discrimination characteristic parameter of metal drawing piece
Li et al. Online chatter detection in milling process based on VMD and multiscale entropy
CN111665050A (en) Rolling bearing fault diagnosis method based on clustering K-SVD algorithm
CN107297649A (en) A kind of lathe flutter on-line monitoring method based on wavelet packet and Hilbert transform
CN114486263A (en) Noise reduction and demodulation method for vibration signal of rolling bearing of rotary machine
CN103473640A (en) Processing progress information collecting method based on stress state of machine tool cutter
CN112975574B (en) Surface quality on-line detection system for aluminum alloy thin-wall part in milling process
CN111251187A (en) Method and device for fusing information and extracting characteristics of blade grinding burn
CN110744359A (en) Numerical control lathe cutter wear monitoring system and method
CN114714157B (en) Grinding chatter monitoring method based on time-varying filtering empirical mode decomposition and instantaneous energy ratio
CN116423292A (en) Milling chatter on-line monitoring method integrating energy ratio and amplitude standard deviation
CN114638265A (en) Milling flutter discrimination method based on signal convolution neural network
CN114923689A (en) Rolling bearing fault diagnosis method based on local feature scale decomposition
CN113807431A (en) Intelligent spindle state evaluation method and system based on multi-source information fusion
CN114757071A (en) Method for predicting vibration pitch diameter number of blade disc based on vibration parameter period
CN115310497A (en) Abnormity identification method for bearing of numerical control machine tool
Liu et al. Milling chatter detection of thin-walled parts based on GA-SE-SCK-VMD and RCMDE

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant