CN108356606A - Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural - Google Patents
Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural Download PDFInfo
- Publication number
- CN108356606A CN108356606A CN201810222486.1A CN201810222486A CN108356606A CN 108356606 A CN108356606 A CN 108356606A CN 201810222486 A CN201810222486 A CN 201810222486A CN 108356606 A CN108356606 A CN 108356606A
- Authority
- CN
- China
- Prior art keywords
- cutting
- rbf neural
- tool wear
- wear
- tool
- 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.)
- Granted
Links
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
- B23Q17/09—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool
- B23Q17/0952—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool during machining
- B23Q17/0957—Detection of tool breakage
Abstract
The present invention relates to a kind of tool wear on-line monitoring method based on wavelet packet analysis and RBF neural, the shearing force coefficient and cutting edge force coefficient of tangential force and radial load when demarcating different cutting-tool wear states using instantaneous Cutting Force Coefficient method of identification;Correlation by analyzing Cutting Force Coefficient and tool wear inputs RBF neural network model as tool wear characteristic parameter after normalized.The input layer of RBF neural monitoring model training process is cutting force feature, the cutting vibration feature by normalized, shearing force coefficient, cutting edge force coefficient;Output layer is the cutter tool flank wear after normalized;Hidden layer is the neuron obtained by radial basis function iteration optimization;There is fast response time, the high advantage of accuracy of identification by Tool Wear Monitoring experimental verification RBF neural monitoring model.
Description
Technical field
The invention belongs to be based on wavelet packet analysis and radial base neural net (radial basis function neural
Network tool wear on-line monitoring method), it is more specifically to a kind of to use wavelet packet analysis and instantaneous cutting force system
Tool wear characteristic value in number method of identification extraction process, is trained characteristic value using RBF neural, to logical
After the method for the RBF neural monitoring model accurate measurements tool wear trained.
Background technology
Important component of the cutting tool state intelligent online monitoring technology as advanced manufacturing technology, has become in recent years
The theme of the research field.Direct executor of the cutter as cutting process, it is inevitable in the cutting process of workpiece
Ground is there is wearing and being damaged, and the variation of cutting tool state directly results in cutting force increase, cutting temperature increases, workpiece table
Surface roughness rises, workpiece size exceeds the generation of tolerance, cutting color change and cutting-vibration, and therefore, it is necessary to cutter
Carry out wear condition monitoring.Cutting-tool wear state Study On Intelligent Monitoring Techniques refer in product processing, it is each for what is detected
Class sensor signal carries out data processing, predicts the state of cutting tool wear and damage in real time using feature recognition model, to
It is accurate to remind the cutter changing time.Studies have shown that CNC machine can reduce the downtime after being equipped with tool monitoring system
75%, production efficiency improves 10-60%, and machine tool utilization rate improves 50%.
Cutting-tool wear state intellectual monitoring is generally divided into two processes:The training process and profit of tool wear associated eigenvalue
The process being monitored with training pattern Cutter wear.Wherein, the extraction of wear characteristic involved in training process, feature extraction
Method include three classes:Temporal analysis, frequency domain analysis, Time-Frequency Analysis method.Temporal analysis refers to directly in time-domain
The interior method that signal is handled, using time-domain signal as feature signal extraction, often by noise signal or by
Change in machined parameters and make the changed influence of signal, thus the accuracy of signal extraction can be caused to decline, and often
A large amount of experimental data is needed to carry out induction and conclusion signal rule.Frequency-domain analysis refers to will by Fourier space or Fourier transformation
Signal is transformed into frequency domain by time-domain and is analyzed and handled.But carry out conversion of the time domain to frequency domain using Fourier transformation
When have the shortcomings that it is certain, i.e. this method for when extracting signal spectrum, needing whole time-domain informations using signal, due to
This method is a kind of integral transformation, lacks the function of time domain positioning, it is therefore desirable to be improved this transform method.Time-frequency domain
Analytic approach can connect the time-domain and frequency domain of signal, so as to which the change of signal frequency at any time is explicitly described
Change, makes signal that can reach high resolution ratio in time domain and frequency domain.In recent years, neural network is applied in terms of feature recognition
Extensively, can rapidly and accurately characteristic value be trained and is identified.
A kind of Tool Wear Monitoring method based on electric current Yu sound emission composite signal of patent of invention (CN104723171A)
The acoustic emission signal by the current signal and lathe tool state of wear that detect the spindle motor in machining is proposed, and is passed through
Wavelet packet analysis, principal component analysis judge the degree of wear of cutter.But this method need to measure electric current and two kinds of signals of sound emission, increase
Equipment cost and the complexity of clamping are added;Meanwhile judging that tool wear degree can not accurately predict knife using principal component
The tool flank wear of tool.
Invention content
Technical problems to be solved
In order to avoid the shortcomings of the prior art, the present invention proposes that one kind being based on wavelet packet analysis and RBF neural
Tool wear on-line monitoring method, solve to distinguish resolution ratio over the frequency domain when time domain method and frequency domain method processing signal low
The problem of.
Technical solution
A kind of tool wear on-line monitoring method based on wavelet packet analysis and RBF neural, it is characterised in that step
It is as follows:
Step 1:Under a certain operating mode, material is processed using constant cutting parameter, cutter is in workpiece upper side edge
Milling Process measures the Cutting Force Signal during this, while measuring the tool flank wear of cutter after processing every time, and will return
One changes treated output valve of the knife face wear extent as RBF neural;
Step 2:
1, Cutting Force Signal is carried out three layers under time-frequency domain using the wavelet packet functions tool box in MATLAB and decomposes weight
Structure obtains the wavelet packet coefficient of each frequency range;Using the reconstruct wavelet packet coefficient of first frequency range as cutting force feature, second
The reconstruct wavelet packet coefficient of frequency range is as cutting vibration feature;
2, original Cutting Force Signal is filtered using MATLAB design filters;Utilize instantaneous Cutting Force Coefficient
Shearing force coefficient and cutting edge force coefficient when method of identification demarcates different cutting-tool wear states is as tool wear characteristic value;
Step 3:Cutting force feature that step 2 is handled, cutting vibration feature, shearing force coefficient, cutting edge force coefficient carry out
Then normalized is used as RBF neural input terminal, the cutter tool flank wear after step 1 normalized is made
It is trained for network deferent segment, obtains the RBF neural monitoring model of training completion;
Step 4:Under same operating, the Cutting Force Signal during processing on real-time is measured;It is carried using the method for step 2
It takes cutting force feature, cutting vibration feature, shearing force coefficient, cutting edge force coefficient and is normalized, by normalized
Data afterwards input RBF neural monitoring model, and the output of RBF neural monitoring model is cutter tool flank wear
Predicted value.
The Cutting Force Signal during this is measured using Kistler9123C rotation dynamometers.
Advantageous effect
A kind of tool wear on-line monitoring method based on wavelet packet analysis and RBF neural proposed by the present invention,
In terms of feature signal extraction, the present invention proposes to carry out data to the original Cutting Force Signal in cutting process using wavelet packet analysis
Processing, this method make signal have higher time and frequency resolution in low frequency and high frequency treatment:By segmenting letter under frequency domain
Number, reconstruction signal under time domain, extract in original Cutting Force Signal with the relevant cutting force feature of tool wear and cutting vibration
Feature is demarcated shearing force coefficient and cutting edge force coefficient under different state of wear further according to instantaneous Cutting Force Coefficient method of identification, is obtained
To four classes and the relevant feature value vector of tool wear.In terms of feature recognition, by analyzing present mode recognition methods, in conjunction with
The quality of response speed and accuracy of identification, patent of the present invention select side of the RBF neural as data fusion and feature recognition
The relevant feature vector of four classes and the tool wear of extraction is imported RBF neural and is trained, will be used for tool wear by method
The signal characteristic parameter of monitoring imports the neural network monitoring model trained, to which rapidly and accurately prediction cutter flank is ground
Damage amount, achievees the purpose that on-line monitoring.
Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural proposed by the invention have with
Lower advantage:
1, tool wear is measured since traditional direct method of measurement needs to shut down, makes production disruption, is unfavorable for processing efficiency
Raising.Therefore this method allows tool wear to reach using measurement indirectly and the relevant signal characteristic value of tool wear
The effect of line monitoring.
2, cutter is during actual processing, with the increase of wear extent, the signals such as cutting force, vibration cutting, sound emission
Significant changes can occur, however, equipment cost and the complexity of clamping can be increased if being acquired all coherent signals.
Therefore, patent of the present invention proposes Cutting Force Signal in only acquisition cutting process, by the original tangential and diameter in cutting process
Wavelet packet analysis is carried out to Cutting Force Signal, is separated and the relevant cutting force feature of tool wear and cutting vibration feature, use
In the training and monitoring of subsequent tool wear feature.
3, the present invention demarcates Cutting Force Coefficient under different cutting-tool wear states using instantaneous Cutting Force Coefficient method of identification, passes through
The correlation for analyzing the instantaneous Cutting Force Coefficient and tool wear of different wear stages, by the shearing force coefficient and cutting edge when cutting
Force coefficient is monitored on-line as tool wear feature for tool wear.
4, the present invention uses radial base neural net (RBFNN) as the mould for tool wear characteristic value training and monitoring
Type.Since when central range of the input signal close to radial basis function, hidden node will generate larger output, so the net
Network has partial approximation ability, and response speed and accuracy of identification are also superior to BP neural network.
Description of the drawings
Fig. 1 is tool wear on-line monitoring method flow chart.
Fig. 2 is extracting machining tool wear features process.
Fig. 3 is three layers of WAVELET PACKET DECOMPOSITION structure chart.
Fig. 4 is cutting force Coefficient Fitting figure.
Fig. 5 is radial cutting force Coefficient Fitting figure.
Fig. 6 is RBF neural network structure figure.
Specific implementation mode
In conjunction with embodiment, attached drawing, the invention will be further described:
The present invention is based on the step of wavelet packet analysis and tool wear on-line monitoring method of RBF neural to be:Step
One:Under a certain operating mode, titanium alloy material is processed using constant cutting parameter, cutter adds in workpiece upper side edge milling
Work 58 times measures the Cutting Force Signal during this using Kistler9123C rotation dynamometers, while measuring processing every time
The tool flank wear of cutter afterwards, extraction wherein 50 groups of tool wear measured values are used to train as the output of neural network.Step
Rapid two:1. filtered Cutting Force Signal is carried out decomposed and reconstituted, reconstruction signal under time-frequency domain using three layers of wavelet packet analysis
In first frequency range and second frequency band signals energy value and tool wear correlation it is higher, therefore, can be by first frequency
Segment signal is used for the training of follow-up neural network as cutting vibration feature as cutting force feature, second frequency band signals.2.
Original Cutting Force Signal is filtered using MATLAB design filters, is demarcated using instantaneous Cutting Force Coefficient method of identification
Shearing force coefficient and cutting edge force coefficient when different state of wear is as tool wear characteristic value;Step 3:By normalized
50 groups afterwards are with the relevant cutting force feature of tool wear, cutting vibration feature, shearing force coefficient, cutting edge force coefficient as RBF
The input terminal of neural network, by after 50 groups of cutter tool flank wear diameter normalizeds of measurement as RBF neural
Output end is trained network, obtains RBF neural monitoring model;Step 4:It will be in 58 groups of Cutting Force Signals of acquisition
Remaining 8 groups for Tool Wear Monitoring test, using in step 2 data extraction method extraction Cutting Force Signal in knife
Tool wears relevant cutting force feature, cutting vibration feature, shears force coefficient and cutting edge force coefficient, conduct after normalized
The output end of the input terminal of RBF neural monitoring model, this monitoring model is cutter tool flank wear, the results showed that,
This Tool Wear Monitoring model response speed and accuracy of identification are satisfied by the requirement of on-line monitoring.
RBF neural input terminal neuron node number determined by characteristic value type, hidden layer neuron number of nodes
It can be determined by empirical equation, output neuron number of nodes is determined by desired output.Input node is and tool wear phase
The characteristic quantity of pass shares four category features vector, cutting force feature respectively tangentially and radially and cutting vibration feature, and not
With the shearing force coefficient and cutting edge force coefficient under state of wear, output node is the tool flank wear of cutter.For the ease of RBF
Neural network model training and monitoring, need the input and output characteristic value first to neural network to make normalized.
Specific implementation mode:
With reference to attached drawing 1-6, by taking milling General Aviation material titanic alloy machining process as an example, to the specific embodiment party of the present invention
Formula is described.
The present invention proposes a kind of tool wear on-line monitoring method based on wavelet packet analysis and RBF neural, attached drawing
1-2 indicates the tool wear on-line monitoring method flow and characteristic extraction procedure, includes mainly 6 steps:
Step 1:Under a certain operating mode, part is processed using constant cutting parameter, cutter is in titanium alloy workpiece
Upper Milling Process radial distance 1mm, the side of axial cutting-in 2mm 58 times, cutter become blunt abrasion from running-in wear, measure every
The tool flank wear of cutter after secondary processing, extraction wherein 50 groups of tool wear measured values are carried out as the output valve of neural network
Training.Meanwhile the cutting force during this is measured using Kistler 9123C rotation dynamometers;Its effect is:Using constant
Cutting parameter can reduce variable, reduce difficulty in computation, the three-dimensional cutting force and torque arithmetic measured according to rotation dynamometer
Go out cutter tangentially and radial cutting force, rotation dynamometer have dynamometry signal stabilization, strong antijamming capability easy for installation etc. excellent
Point.
Step 2:
1. the original Cutting Force Signal of measurement carried out under time-frequency domain using three layers of wavelet packet analysis it is decomposed and reconstituted, it is discrete
Orthogonal wavelet packet transform is defined as observation signal s (t) in Orthogonal Wavelet Packet base { wN, j, k(t)}N ∈ Z/Z-, j ∈ Z, k ∈ ZOn projection system
Number, i.e.,:
Wherein, n is Frequency Index, and k is positioning index, and j is scale index, wnReferred to as about the Orthogonal Wavelet Packet of ψ (t)
Base, by the scaling function ψ (t) of an orthonormalization, wherein w0=ψ (t) is generated by double scale difference recursion equation groups
Group of functions:
Wherein, hk、gkFor a pair of of conjugate quadrature mirror filter coefficient derived from ψ (t).
Wavelet packet can be with the increase of resolution ratio 2j, and it is excellent that there is further segmentation to attenuate for the spectral window that broadens
Quality combines orthogonal filter h for given signal by one group of low high passk、gk, signal can be divided to arbitrary frequency range
On, make it that all there is higher time and frequency resolution on low frequency and high frequency, patent of the present invention is first by processed
Cutting Force Signal in journey carries out three layers of WAVELET PACKET DECOMPOSITION, extraction and the relevant cutting force of abrasion from wavelet packet coefficient reconstruct image
Feature and cutting vibration feature, Fig. 3 indicate that three layers of specific partition process of WAVELET PACKET DECOMPOSITION, j indicate Decomposition order;Its effect is:
The Cutting Force Signal of sample frequency 2kHZ can be subdivided into 8 sections by three layers of WAVELET PACKET DECOMPOSITION on frequency domain, in the time domain will be each
The wavelet packet coefficient of frequency range is reconstructed, and by solving reconstruction signal energy value, extracts and believes with the relevant frequency range of tool wear
Number, and using the energy value of this signal as tool wear characteristic value.
2. being filtered to original Cutting Force Signal using MATLAB design filters, instantaneous cutting force system is recycled
Cutting Force Coefficient when number method of identifications demarcate different state of wear is used as tool wear characteristic value, due to shearing force coefficient and cutting edge
Force coefficient is held essentially constant when tool wear variation is smaller, according to instantaneous cutting force formula (4) it is found that instantaneous cutting
Power changes with the variation of Tool in Cutting thickness, and shearing force coefficient can be acquired with cutting edge force coefficient using linear fit.
Wherein, FqIndicate instantaneous cutting force, KcIndicate shearing force coefficient, KeIndicate cutting edge force coefficient, apIt indicates axial to cut
It is deep,Thickness of cutting is indicated, shown in expression formula such as following formula (5):
Wherein, c indicates feed engagement,Cutter teeth parallactic angle is indicated, since cutter adds material upmilling in testing herein
Work, cutter entrance angleAnd radial cutting-in is ae=1mm knows that cutter is cut out according to cutter teeth parallactic angle calculation formula (6)
Angle:
In practical fit procedure, when cutter teeth parallactic angle is from entrance angle to when cutting out angle and changing, the variation of cutting force by zero by
Gradually reach peak value to terminate, shearing force coefficient when going out different cutting-tool wear states using formula (7) linear fit and cutting edge power system
Number, Fig. 4-5 indicate the process using MATLAB linear fits shearing force coefficient and cutting edge force coefficient.
Known according to data analysis, the shearing force COEFFICIENT K of tangential force and radial loadcWith cutting edge force coefficient KeWith tool wear shape
State correlation is larger, therefore can be used as a kind of tool wear characteristic value;Its effect is:By demarcating each cutting-tool wear state
Under Cutting Force Coefficient, judge the relationship of Cutting Force Coefficient and tool wear and using Cutting Force Coefficient as tool wear feature
Value.
Step 3:By 50 groups after normalized with the relevant cutting force feature of tool wear, cutting vibration feature, cut
Shear force coefficient, cutting edge force coefficient are trained as feature vector input RBF neural, and Fig. 6 indicates RBF neural network structure
Figure, RBF neural input layer are the tool wear characteristic parameter extracted, and hidden layer uses Gaussian function, the work of node
With being analyzed certain regional areas of input signal, determine that the important parameter of hidden layer node is in radial basis function
The heart and width, are indicated with c and σ respectively, are shown below:
Wherein,X is input vector, cjFor j-th Gaussian function in n n dimensional vector ns
Cluster centre is randomly selected according to input vector,It is Normalized Scale parameter, is equivalent to variance.RjIt indicates to imply layer functions
Output.
When the center of input vector and certain hidden layer node closest to when, the output of this hidden layer node is maximum, with arrow
Amount is separate with radial direction base center, the output of hidden layer node also can monotonic decreasing, the output of hidden layer node passes through weights ω
It linearly is transferred to output layer, corresponding function expression is as follows:
wiTo export weights, Y exports for RBF neural desired value, herein with extracted using wavelet packet analysis and knife
Tool wears relevant cutting force feature and cutting vibration feature, and the shearing force using the calibration of instantaneous Cutting Force Coefficient method of identification
As input feature value, output layer is the wear extent of cutter flank for coefficient and cutting edge force coefficient;Its effect is:Radial base god
Through network (RBFNN) since when central range of the input signal close to radial basis function, hidden layer node will generate larger
Output, so the network has partial approximation ability, response speed and accuracy of identification are made also superior to BP neural network
It is characterized identification model.
Step 4:Remaining 8 groups in 58 groups of Cutting Force Signals of acquisition are tested for Tool Wear Monitoring, using step
Data extraction method in two extract in Cutting Force Signal with the relevant cutting force feature of tool wear, cutting vibration feature,
Shear force coefficient and cutting edge force coefficient, the input terminal after normalized as RBF neural monitoring model, this monitoring model
Output end be monitor cutter tool flank wear, the results showed that, this Tool Wear Monitoring model response speed and identification
Precision is satisfied by the requirement of on-line monitoring.
Claims (2)
1. a kind of tool wear on-line monitoring method based on wavelet packet analysis and RBF neural, it is characterised in that step is such as
Under:
Step 1:Under a certain operating mode, material is processed using constant cutting parameter, cutter is in workpiece upper side edge milling
Processing measures the Cutting Force Signal during this, while measuring the tool flank wear of cutter after processing every time, and will normalization
Output valve of the knife face wear extent as RBF neural that treated;
Step 2:
(1)Cutting Force Signal carried out under time-frequency domain using the wavelet packet functions tool box in MATLAB three layers it is decomposed and reconstituted,
Obtain the wavelet packet coefficient of each frequency range;Using the reconstruct wavelet packet coefficient of first frequency range as cutting force feature, second frequency
The reconstruct wavelet packet coefficient of section is as cutting vibration feature;
(2)Original Cutting Force Signal is filtered using MATLAB design filters;Known using instantaneous Cutting Force Coefficient
Shearing force coefficient and cutting edge force coefficient when other method demarcates different cutting-tool wear states is as tool wear characteristic value;
Step 3:Cutting force feature that step 2 is handled, cutting vibration feature, shearing force coefficient, cutting edge force coefficient carry out normalizing
Change is handled, and RBF neural input terminal is then used as, using the cutter tool flank wear after step 1 normalized as net
Network deferent segment is trained, and obtains the RBF neural monitoring model of training completion;
Step 4:Under same operating, the Cutting Force Signal during processing on real-time is measured;It is cut using the method extraction of step 2
It cuts power feature, cutting vibration feature, shearing force coefficient, cutting edge force coefficient and is normalized, after normalized
Data input RBF neural monitoring model, and the output of RBF neural monitoring model is the prediction of cutter tool flank wear
Value.
2. the tool wear on-line monitoring method based on wavelet packet analysis and RBF neural according to claim 1, special
Sign is:The Cutting Force Signal during this is measured using Kistler9123C rotation dynamometers.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810222486.1A CN108356606B (en) | 2018-03-19 | 2018-03-19 | Tool wear online monitoring method based on wavelet packet analysis and RBF neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810222486.1A CN108356606B (en) | 2018-03-19 | 2018-03-19 | Tool wear online monitoring method based on wavelet packet analysis and RBF neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108356606A true CN108356606A (en) | 2018-08-03 |
CN108356606B CN108356606B (en) | 2020-05-01 |
Family
ID=63000698
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810222486.1A Active CN108356606B (en) | 2018-03-19 | 2018-03-19 | Tool wear online monitoring method based on wavelet packet analysis and RBF neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108356606B (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
CN109158953A (en) * | 2018-09-04 | 2019-01-08 | 温州大学激光与光电智能制造研究院 | A kind of cutting-tool wear state on-line monitoring method and system |
CN109333160A (en) * | 2018-10-09 | 2019-02-15 | 西北工业大学 | The on-line monitoring method of high temperature alloy drilling process bit wear form and state of wear |
CN109434564A (en) * | 2018-12-21 | 2019-03-08 | 哈尔滨理工大学 | A kind of cutter wear state monitoring method based on deep neural network |
CN109571141A (en) * | 2018-11-01 | 2019-04-05 | 北京理工大学 | A kind of Monitoring Tool Wear States in Turning based on machine learning |
CN109623494A (en) * | 2019-01-18 | 2019-04-16 | 四川大学 | A kind of three-in-one sensor fixture and multi-modal cutting-tool wear state monitor system |
CN109732450A (en) * | 2019-02-27 | 2019-05-10 | 重庆理工大学 | A kind of blade polishing processing method neural network based |
CN109968104A (en) * | 2019-03-07 | 2019-07-05 | 中南大学 | A kind of highly sensitive accurate monitoring method of high speed wire pulling method |
CN110057707A (en) * | 2019-05-31 | 2019-07-26 | 上海交通大学 | Carbon fibre reinforced composite/titanium alloy lamination drilling cutters biometrics method |
CN110103079A (en) * | 2019-06-17 | 2019-08-09 | 中国科学院合肥物质科学研究院 | The on-line monitoring method of tool wear in a kind of micro- milling process |
CN110509109A (en) * | 2019-07-16 | 2019-11-29 | 西安交通大学 | Tool Wear Monitoring method based on multiple dimensioned depth convolution loop neural network |
CN110908334A (en) * | 2018-09-18 | 2020-03-24 | 财团法人工业技术研究院 | Cutter wear monitoring method |
CN111660141A (en) * | 2020-05-14 | 2020-09-15 | 北京工业大学 | Milling cutter wear state identification method based on spindle driving current and irrelevant to working conditions |
CN111832432A (en) * | 2020-06-23 | 2020-10-27 | 华中科技大学 | Cutter wear real-time prediction method based on wavelet packet decomposition and deep learning |
CN113126564A (en) * | 2021-04-23 | 2021-07-16 | 重庆大学 | Digital twin driven numerical control milling cutter abrasion on-line monitoring method |
CN113352149A (en) * | 2021-08-11 | 2021-09-07 | 西安博澜数字科技有限公司 | Cutter working condition parameter monitoring method and device |
CN113359622A (en) * | 2021-07-09 | 2021-09-07 | 南京航空航天大学 | Data-driven numerical control machine tool cutting force self-monitoring method |
CN114378653A (en) * | 2022-01-27 | 2022-04-22 | 上海机床厂有限公司 | Cylindrical grinding chatter online identification and monitoring method based on BP neural network |
WO2022126678A1 (en) * | 2020-12-17 | 2022-06-23 | 广东工业大学 | Method and device for evaluating performance state of numerical control cutting tool bit of flexible material |
CN114888635A (en) * | 2022-04-27 | 2022-08-12 | 哈尔滨理工大学 | Cutter state monitoring method |
CN116008114A (en) * | 2023-01-17 | 2023-04-25 | 哈尔滨理工大学 | Method for detecting tool wear in bone material drilling process based on sensor signals |
-
2018
- 2018-03-19 CN CN201810222486.1A patent/CN108356606B/en active Active
Cited By (33)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
CN109158953A (en) * | 2018-09-04 | 2019-01-08 | 温州大学激光与光电智能制造研究院 | A kind of cutting-tool wear state on-line monitoring method and system |
CN109158953B (en) * | 2018-09-04 | 2019-07-02 | 温州大学激光与光电智能制造研究院 | A kind of cutting-tool wear state on-line monitoring method and system |
CN110908334A (en) * | 2018-09-18 | 2020-03-24 | 财团法人工业技术研究院 | Cutter wear monitoring method |
CN110908334B (en) * | 2018-09-18 | 2021-02-19 | 财团法人工业技术研究院 | Cutter wear monitoring method |
CN109333160B (en) * | 2018-10-09 | 2020-09-22 | 西北工业大学 | Online monitoring method for abrasion form and abrasion state of drill bit in high-temperature alloy drilling process |
CN109333160A (en) * | 2018-10-09 | 2019-02-15 | 西北工业大学 | The on-line monitoring method of high temperature alloy drilling process bit wear form and state of wear |
CN109571141A (en) * | 2018-11-01 | 2019-04-05 | 北京理工大学 | A kind of Monitoring Tool Wear States in Turning based on machine learning |
CN109434564A (en) * | 2018-12-21 | 2019-03-08 | 哈尔滨理工大学 | A kind of cutter wear state monitoring method based on deep neural network |
CN109623494B (en) * | 2019-01-18 | 2023-08-15 | 四川大学 | Three-in-one sensor clamp and multi-mode cutter wear state monitoring system |
CN109623494A (en) * | 2019-01-18 | 2019-04-16 | 四川大学 | A kind of three-in-one sensor fixture and multi-modal cutting-tool wear state monitor system |
CN109732450B (en) * | 2019-02-27 | 2021-02-26 | 重庆理工大学 | Blade polishing and grinding method based on neural network |
CN109732450A (en) * | 2019-02-27 | 2019-05-10 | 重庆理工大学 | A kind of blade polishing processing method neural network based |
CN109968104A (en) * | 2019-03-07 | 2019-07-05 | 中南大学 | A kind of highly sensitive accurate monitoring method of high speed wire pulling method |
CN109968104B (en) * | 2019-03-07 | 2021-10-08 | 中南大学 | High-sensitivity precise monitoring method for high-speed broaching machining |
CN110057707B (en) * | 2019-05-31 | 2021-06-25 | 上海交通大学 | Method for measuring service life of carbon fiber reinforced composite/titanium alloy laminated drilling tool |
CN110057707A (en) * | 2019-05-31 | 2019-07-26 | 上海交通大学 | Carbon fibre reinforced composite/titanium alloy lamination drilling cutters biometrics method |
CN110103079A (en) * | 2019-06-17 | 2019-08-09 | 中国科学院合肥物质科学研究院 | The on-line monitoring method of tool wear in a kind of micro- milling process |
CN110509109A (en) * | 2019-07-16 | 2019-11-29 | 西安交通大学 | Tool Wear Monitoring method based on multiple dimensioned depth convolution loop neural network |
CN110509109B (en) * | 2019-07-16 | 2021-04-13 | 西安交通大学 | Cutter wear monitoring method based on multi-scale depth convolution cyclic neural network |
CN111660141B (en) * | 2020-05-14 | 2022-02-15 | 北京工业大学 | Milling cutter wear state identification method based on spindle driving current and irrelevant to working conditions |
CN111660141A (en) * | 2020-05-14 | 2020-09-15 | 北京工业大学 | Milling cutter wear state identification method based on spindle driving current and irrelevant to working conditions |
CN111832432A (en) * | 2020-06-23 | 2020-10-27 | 华中科技大学 | Cutter wear real-time prediction method based on wavelet packet decomposition and deep learning |
WO2022126678A1 (en) * | 2020-12-17 | 2022-06-23 | 广东工业大学 | Method and device for evaluating performance state of numerical control cutting tool bit of flexible material |
CN113126564A (en) * | 2021-04-23 | 2021-07-16 | 重庆大学 | Digital twin driven numerical control milling cutter abrasion on-line monitoring method |
CN113359622A (en) * | 2021-07-09 | 2021-09-07 | 南京航空航天大学 | Data-driven numerical control machine tool cutting force self-monitoring method |
CN113352149B (en) * | 2021-08-11 | 2021-11-02 | 西安博澜数字科技有限公司 | Cutter working condition parameter monitoring method and device |
CN113352149A (en) * | 2021-08-11 | 2021-09-07 | 西安博澜数字科技有限公司 | Cutter working condition parameter monitoring method and device |
CN114378653A (en) * | 2022-01-27 | 2022-04-22 | 上海机床厂有限公司 | Cylindrical grinding chatter online identification and monitoring method based on BP neural network |
CN114888635A (en) * | 2022-04-27 | 2022-08-12 | 哈尔滨理工大学 | Cutter state monitoring method |
CN114888635B (en) * | 2022-04-27 | 2023-07-25 | 哈尔滨理工大学 | Cutter state monitoring method |
CN116008114A (en) * | 2023-01-17 | 2023-04-25 | 哈尔滨理工大学 | Method for detecting tool wear in bone material drilling process based on sensor signals |
CN116008114B (en) * | 2023-01-17 | 2023-09-12 | 哈尔滨理工大学 | Method for detecting tool wear in bone material drilling process based on sensor signals |
Also Published As
Publication number | Publication date |
---|---|
CN108356606B (en) | 2020-05-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108356606A (en) | Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural | |
Madhusudana et al. | Condition monitoring of face milling tool using K-star algorithm and histogram features of vibration signal | |
Madhusudana et al. | Fault diagnosis of face milling tool using decision tree and sound signal | |
CN105834835B (en) | A kind of tool wear on-line monitoring method based on Multiscale Principal Component Analysis | |
Nasir et al. | Acoustic emission monitoring of sawing process: artificial intelligence approach for optimal sensory feature selection | |
CN102091972B (en) | Numerical control machine tool wear monitoring method | |
CN110000610A (en) | A kind of Tool Wear Monitoring method based on multi-sensor information fusion and depth confidence network | |
CN109434564A (en) | A kind of cutter wear state monitoring method based on deep neural network | |
CN110647943B (en) | Cutting tool wear monitoring method based on evolution data cluster analysis | |
CN103345200B (en) | A kind of cut Identification of Chatter method based on generalized interval | |
Sevilla-Camacho et al. | Tool breakage detection in CNC high-speed milling based in feed-motor current signals | |
CN106002490B (en) | Milling workpiece roughness monitoring method based on Path and redundant eliminating | |
CN108629864B (en) | Vibration-based electric spindle radial precision characterization method and system | |
CN108747590A (en) | A kind of tool wear measurement method based on rumble spectrum and neural network | |
CN106271881B (en) | A kind of Condition Monitoring of Tool Breakage method based on SAEs and K-means | |
CN104390697B (en) | One kind is based on C0The milling parameter detection method of complexity and coefficient correlation | |
WO2016004749A1 (en) | Method for recognizing tool abrasion degree of large numerical control milling machine | |
Zhou et al. | Tool condition monitoring in milling using a force singularity analysis approach | |
CN109732406A (en) | A kind of intelligent tool method for diagnosing faults | |
Chen et al. | Chatter detection in milling processes using frequency-domain Rényi entropy | |
Li et al. | Online chatter detection in milling process based on VMD and multiscale entropy | |
CN114905336B (en) | Variable working condition cutter wear monitoring method and system based on cutting force component decoupling | |
Li et al. | Chatter identification of thin-walled parts for intelligent manufacturing based on multi-signal processing | |
Mou et al. | A study of tool tipping monitoring for titanium milling based on cutting vibration | |
CN111958321B (en) | Numerical control machine tool cutter wear degree identification method based on deep neural network |
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 |