CN109605127A - A kind of cutting-tool wear state recognition methods and system - Google Patents
A kind of cutting-tool wear state recognition methods and system Download PDFInfo
- Publication number
- CN109605127A CN109605127A CN201910053100.3A CN201910053100A CN109605127A CN 109605127 A CN109605127 A CN 109605127A CN 201910053100 A CN201910053100 A CN 201910053100A CN 109605127 A CN109605127 A CN 109605127A
- Authority
- CN
- China
- Prior art keywords
- cutting
- state
- tool
- tool wear
- force signal
- 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.)
- Pending
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
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Machine Tool Sensing Apparatuses (AREA)
Abstract
The invention discloses a kind of cutting-tool wear state recognition methods and systems.The recognition methods includes: the Cutting Force Signal and cutting-tool wear state obtained in cutter drilling process；Noise reduction process is carried out to the Cutting Force Signal using Threshold Denoising Method, the Cutting Force Signal after determining noise reduction；The feature in Cutting Force Signal after extracting the noise reduction using time domain, frequency domain and Harmonic wavelet packet analysis method, determines feature samples collection；It is input with the feature samples collection, is output with the cutting-tool wear state, establishes the cutting tool state identification model based on least square method supporting vector machine；The current state of wear of the cutter is identified according to the cutting tool state identification model.It can be improved the recognition efficiency of cutting-tool wear state using recognition methods provided by the present invention and system.
Description
Technical field
The present invention relates to machining tool and detection field, more particularly to a kind of cutting-tool wear state recognition methods and
System.
Background technique
Abrasion Monitoring is of great significance in intelligence manufacture production；Machining is as forming parts
In most important processing method, manufacture production in play an important role.Direct executor of the cutter as cutting process,
Its state is important to guaranteeing processing quality, improving production efficiency, reduction production cost and realizing that continuous and automatic processing has
Meaning.The method of current cutting-tool wear state monitoring is mainly indirect method, by related to tool wear in acquisition process
Various signals, signal processing appropriate is carried out to it, the signal characteristic of corresponding different state of wear is extracted, utilizes intelligent calculation
Method establishes cutting-tool wear state identification model, realizes the monitoring of Cutter wear state.Abrasion Monitoring at present
The problem of being primarily present two aspects: first is that for non-stationary signal, the characteristic reliability that signal processing extracts is not high；Second is that needing
It wants a large amount of sample data to ensure accuracy of identification, causes cutting-tool wear state recognition efficiency low.
Summary of the invention
The object of the present invention is to provide a kind of cutting-tool wear state recognition methods and systems, to solve current tool wear shape
State monitoring method signal processing extraction characteristic reliability is not high, the low problem of cutting-tool wear state recognition efficiency.
To achieve the above object, the present invention provides following schemes:
A kind of cutting-tool wear state recognition methods, comprising:
Obtain the Cutting Force Signal and cutting-tool wear state in cutter drilling process；
Noise reduction process is carried out to the Cutting Force Signal using Threshold Denoising Method, the cutting force letter after determining noise reduction
Number；
The feature in Cutting Force Signal after extracting the noise reduction, determines feature samples collection；The feature includes time domain spy
Sign, frequency domain character and frequency band energy；
It is input with the feature samples collection, is output with the cutting-tool wear state, establishes and supported based on least square
The cutting tool state identification model of vector machine；
The current state of wear of the cutter is identified according to the cutting tool state identification model.
Optionally, the feature for extracting the Cutting Force Signal, determines feature samples collection, specifically includes:
Temporal signatures and the frequency domain spy in the Cutting Force Signal are extracted using temporal analysis and frequency domain analysis
Sign, determines feature samples collection.
Optionally, the feature for extracting the Cutting Force Signal, determines feature samples collection, specifically includes:
Frequency band energy in the Cutting Force Signal is extracted using Harmonic wavelet packet analytic approach, determines feature samples collection.
Optionally, described with the feature samples collection is input, is output with the cutting-tool wear state, establishes based on most
Small two multiply the cutting tool state identification model of support vector machines, specifically include:
The feature samples collection is divided into training set and test set；
It is input with the training set, is output with the cutting-tool wear state, establishes base using particle swarm optimization algorithm
In the cutting tool state identification model of least square method supporting vector machine.
Optionally, it is described with the training set be input, with the cutting-tool wear state be output, utilize particle group optimizing
Algorithm is established after the cutting tool state identification model based on least square method supporting vector machine, further includes:
The cutting tool state identification model is verified using the test set.
Optionally, it is described with the training set be input, with the cutting-tool wear state be output, utilize particle group optimizing
Algorithm establishes the cutting tool state identification model based on least square method supporting vector machine, specifically includes:
Penalty factor and nuclear parameter are determined using particle swarm optimization algorithm；
Radial basis kernel function is determined according to the penalty factor and the nuclear parameter；
The cutting tool state identification model based on least square method supporting vector machine is established according to the Radial basis kernel function.
A kind of cutting-tool wear state identifying system, comprising:
Cutting Force Signal obtains module, for obtaining Cutting Force Signal and tool wear in cutter drilling process
State；
Noise reduction process module is determined for carrying out noise reduction process to the Cutting Force Signal using Threshold Denoising Method
Cutting Force Signal after noise reduction；
Feature samples collection determining module determines feature sample for extracting the feature in the Cutting Force Signal after the noise reduction
This collection；The feature includes temporal signatures, frequency domain character and frequency band energy；
Cutting tool state identification model establishes module, for being input with the feature samples collection, with the tool wear shape
State is output, establishes the cutting tool state identification model based on least square method supporting vector machine；
Current state of wear identification module, for identifying the current mill of the cutter according to the cutting tool state identification model
Damage state.
Optionally, the feature samples collection determining module specifically includes:
Fisrt feature sample set determination unit, for extracting the cutting force using temporal analysis and frequency domain analysis
Temporal signatures and frequency domain character in signal, determine feature samples collection.
Optionally, the feature samples collection determining module specifically includes:
Second feature sample set determination unit, for being extracted in the Cutting Force Signal using Harmonic wavelet packet analytic approach
Frequency band energy determines feature samples collection.
Optionally, the cutting tool state identification model is established module and is specifically included:
Division unit, for the feature samples collection to be divided into training set and test set；
Cutting tool state identification model establishes unit, for being input with the training set, is with the cutting-tool wear state
Output establishes the cutting tool state identification model based on least square method supporting vector machine using particle swarm optimization algorithm.
The specific embodiment provided according to the present invention, the invention discloses following technical effects: the present invention provides one kind
Cutting-tool wear state recognition methods and system are analyzed using Harmonic wavelet packet and carry out signal characteristic abstraction, effectively solution wavelet packet
The frequency band of decomposition overlaps problem, improves the reliability of feature samples；The cutter based on least square method supporting vector machine is established simultaneously
State recognition model establishes well contacting between tool condition monitoring signal and cutting-tool wear state, and is directed to kernel function
Penalty factor and nuclear parameter have this problem of larger impact to the accuracy of identification that cutting tool state is worn, utilize Particle Swarm Optimization
Method carries out parameter optimization to the penalty factor and nuclear parameter of kernel function and finds optimal parameter combination, effectively increases sample data
Recognition efficiency.
Detailed description of the invention
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is cutting-tool wear state recognition methods flow chart provided by the present invention；
Fig. 2 is cutting-tool wear state identifying system structure chart provided by the present invention；
Fig. 3 is the spectrogram under different machining states provided by the present invention；
Fig. 4 is Cutting Force Signal WAVELET PACKET DECOMPOSITION band energy figure under different machining states provided by the present invention；Fig. 4
It (a) is normal process stage schematic diagram provided by the present invention, Fig. 4 (b) is sharp wear stage provided by the present invention signal
Figure；
Fig. 5 is particle swarm optimization algorithm flow chart provided by the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The object of the present invention is to provide a kind of cutting-tool wear state recognition methods and systems, can be improved cutting-tool wear state
Recognition efficiency.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, with reference to the accompanying drawing and specific real
Applying mode, the present invention is described in further detail.
Fig. 1 is cutting-tool wear state recognition methods flow chart provided by the present invention, as shown in Figure 1, a kind of tool wear
State identification method, comprising:
Step 101: obtaining the Cutting Force Signal and cutting-tool wear state in cutter drilling process；
Step 102: noise reduction process, cutting after determining noise reduction being carried out to the Cutting Force Signal using Threshold Denoising Method
Cut force signal.
Step 103: the feature in Cutting Force Signal after extracting the noise reduction determines feature samples collection；The feature packet
Include temporal signatures, frequency domain character and frequency band energy.
The step 103 specifically includes: being extracted in the Cutting Force Signal using temporal analysis and frequency domain analysis
Temporal signatures and frequency domain character, determine feature samples collection.
Cutting Force Signal phase after calculating step 1 noise reduction according to the calculation method of time domain charactreristic parameter, frequency domain character parameter
The feature answered mainly includes mean value, standard deviation, root mean square, peak factor, kurtosis index, degree of bias index totally 5 temporal signatures,
Gravity frequency, frequency variance, square frequency totally 3 frequency domain characters.Using this 8 features as be used to characterize cutting-tool wear state
Feature, table 1 are time-domain analysis processing method table provided by the present invention, and table 2 is frequency-domain analysis processing side provided by the present invention
Method table, as shown in table 1 and table 2.
Table 1
Table 2
The step 103 specifically includes: the frequency band energy in the Cutting Force Signal is extracted using Harmonic wavelet packet analytic approach
Amount, determines feature samples collection.
Frequency Band Energy Analysis Using is carried out to the Cutting Force Signal after noise reduction using Fast Fourier Transform (FFT) and obtains each band energy
Distribution situation, effective band limits is determined according to the accounting of energy；Using db3 wavelet basis function to cutting after step 1 noise reduction
It cuts force signal and carries out four layers of Harmonic wavelet packets decomposition, by signal decomposition to each frequency range, according to band energy variation tendency and knife
The energy feature of 6 effective frequency ranges is as sample characteristics before the correlation extraction of tool state of wear.
Step 104: being input with the feature samples collection, be output with the cutting-tool wear state, establish based on minimum
Two multiply the cutting tool state identification model of support vector machines.
The step 104 specifically includes: the feature samples collection is divided into training set and test set；With the training set
For input, it is output with the cutting-tool wear state, is established using particle swarm optimization algorithm and be based on least square method supporting vector machine
Cutting tool state identification model；And the cutting tool state identification model is verified using the test set.
Use Radial basis kernel function K (x, xi)=exp (- | x-xi|2/2σ2) (wherein | x-xi|2It is considered as two features
Square Euclidean distance between vector) it establishes based on least square method supporting vector machine (Least Squares Support
Vector Machine, LS-SVM) cutting tool state identification model, and be iterated optimizing using particle group optimizing process to obtain
Obtain kernel function optimal penalty factor γ and nuclear parameter σ2.Tool abrasion accuracy of identification under different parameters distinguishes larger, core
Parameter σ value is related with the division fine degree of sample, and penalty factor γ is for weighing empiric risk and structure risk.
Specific step is as follows: initialization particle swarm optimization algorithm parameter, using penalty factor and kernel function as each particle
Two-dimensional coordinate, according to training sample training LS-SVM, and a cross validation stay to calculate the fitness of particle, i.e., instructed original
The sample practiced in sample set is successively used as verifying sample, remaining carries out model training as training sample, until each sample
All make to terminate algorithm when one-time authentication sample.
To each particle, fitness f is compared with itself optimal value, updates its own adaptive optimal control value；It will be each
The adaptive optimal control value of particle is compared with global optimum, global optimum's adaptive value of Population Regeneration.Check whether satisfaction knot
Beam condition, the most optimized parameter γ and σ are selected in output if meeting2, continue optimization algorithm parameter if being unsatisfactory for, be based on establishing
The cutting tool state identification model of least square method supporting vector machine.
Step 105: the current state of wear of the cutter is identified according to the cutting tool state identification model.
Using the feature samples collection of extraction as the input of LS-SVM, cutting tool state classification is as output.Parameter optimization is obtained
The γ and σ arrived2Optimum combination substitutes into LS-SVM classifier, and cutting tool state identification model of the training based on LS-SVM algorithm utilizes
Model carries out cutting-tool wear state identification.
Fig. 2 is cutting-tool wear state identifying system structure chart provided by the present invention, as shown in Fig. 2, a kind of tool wear
State recognition system, comprising:
Cutting Force Signal obtains module 201, for obtaining Cutting Force Signal and cutter in cutter drilling process
State of wear；The Cutting Force Signal is used to characterize the state of wear of the cutter.
Noise reduction process module 202, for carrying out noise reduction process to the Cutting Force Signal using Threshold Denoising Method, really
Cutting Force Signal after determining noise reduction.
Feature samples collection determining module 203 determines feature for extracting the feature in the Cutting Force Signal after the noise reduction
Sample set；The feature includes temporal signatures, frequency domain character and frequency band energy.
The feature samples collection determining module 203 specifically includes: fisrt feature sample set determination unit, for using time domain
Analytic approach and frequency domain analysis extract temporal signatures and frequency domain character in the Cutting Force Signal, determine feature samples
Collection.
The feature samples collection determining module 203 specifically includes: second feature sample set determination unit, for using harmonic wave
Wavelet packet analysis method extracts the frequency band energy in the Cutting Force Signal, determines feature samples collection.
Harmonic wavelet packet Decomposition order and frequency range are determined using Fast Fourier Transform (FFT), recycle Harmonic wavelet packet
Analytic approach extracts the feature (i.e. frequency band feature) in Cutting Force Signal.
Cutting tool state identification model establishes module 204, for being input with the feature samples collection, with the tool wear
State is output, establishes the cutting tool state identification model based on least square method supporting vector machine.
The cutting tool state identification model is established module 204 and is specifically included: division unit, is used for the feature samples collection
It is divided into training set and test set；Cutting tool state identification model establishes unit, for being input with the training set, with the knife
Having state of wear is output, establishes the cutting tool state based on least square method supporting vector machine using particle swarm optimization algorithm and identifies mould
Type.
Current state of wear identification module 205, for identifying working as the cutter according to the cutting tool state identification model
Preceding state of wear.
By taking drilling process as an example, main method is as follows:
(1) tool wear signal monitoring is tested.3 drilling test condition table of table, as shown in table 3:
Table 3
Test repeats four life cycle management wear tests using the above machining condition.Cutting set by this paper
Under machined parameters, to each measurement for cutter being carried out under certain processing hole count interval tool abrasion, until cutter
Until abrasion.Cutting-tool wear state is divided into initial wear (VB≤0.2), normal wear according to cutter tool flank wear VB
(0.2<VB≤0.8), sharp wear (VB>0.8).The Drill Force Signals for having chosen the test of third group herein are analyzed, this group
Test machined 85 holes in total, and the sampling number that drills every time is 20000, to avoid the influence for piercing, drilling out, extract herein every
Totally 15000 data points are analyzed for the 2501~17500 of secondary drilling.The cutting tool state type and its corresponding that experiment is obtained
Cutting Force Signal is organized into sample.
(2) signal processing and feature extraction.
Wavelet de-noising is carried out using Cutting Force Signal of the Threshold Denoising Method to experiment acquisition, improves signal-to-noise ratio；
Cutting Force Signal phase after calculating step 1 noise reduction according to the calculation method of time domain charactreristic parameter, frequency domain character parameter
The feature answered, mainly includes mean value, standard deviation, root mean square, kurtosis index, degree of bias index totally 5 temporal signatures, gravity frequency,
Frequency variance, square frequency totally 3 frequency domain characters.
Fig. 3 is the spectrogram under different machining states provided by the present invention, as shown in figure 3, being become using fast Fourier
It changes and the distribution situation that Frequency Band Energy Analysis Using obtains each band energy is carried out to the signal after noise reduction, had according to the determination of the accounting of energy
Band limits is imitated, the frequency content greater than 100Hz is considered noise；Fig. 4 is different processing shapes provided by the present invention
Cutting Force Signal WAVELET PACKET DECOMPOSITION band energy figure under state, as shown in figure 4, using db3 wavelet basis function to step 1 noise reduction after
Cutting Force Signal carry out four layers of Harmonic wavelet packet and decompose, by signal decomposition to each frequency range, according to band energy variation tendency
Energy feature with 6 before the correlation extraction of cutting-tool wear state effective frequency ranges is as sample characteristics.
The corresponding cutting tool state kind of the feature that time domain derived above, frequency domain and Harmonic wavelet packet are analyzed
Class carries out coding composition sample set, and cutting tool state sample set is further divided into training sample set and test sample collection, selects
48 groups in sample set are selected as training sample, 37 groups are used as test sample.
(3) parameter optimization.Optimizing is iterated using particle group optimizing process to obtain optimal penalty factor and core ginseng
Number.Table 4 is the relevant parameter Initialize installation table of PSO provided by the present invention, as shown in table 4:
Table 4
Fig. 5 is particle swarm optimization algorithm flow chart provided by the present invention, as shown in Figure 5, the specific steps are as follows: initialization
Particle swarm optimization algorithm parameter, using penalty factor and kernel function as the two-dimensional coordinate of each particle, according to training sample training
LS-SVM, and carry out the fitness for staying a cross validation to calculate particle, i.e., the sample that former training sample is concentrated successively is used as and is tested
Sample is demonstrate,proved, remaining carries out model training as training sample, terminates algorithm when each sample makees one-time authentication sample.
To each particle, fitness f is compared with itself optimal value, updates its own adaptive optimal control value；Most by each particle
Excellent adaptive value is compared with global optimum, global optimum's adaptive value of Population Regeneration.It checks whether and meets termination condition, if
Meet then output and selects the most optimized parameter γ and σ2, continue optimization algorithm parameter if being unsatisfactory for.
(4) cutting-tool wear state identification model is established.Using Cutting Force Signal as the input of LS-SVM, cutting tool state class
It Zuo Wei not export, cutting tool state identification model of the training based on LS-SVM algorithm carries out cutting tool state identification using model.
The present invention utilizes the drilling axis in the acquisition drilling processing of kistler three-dimensional dynamometer by taking drilling process as an example
To force signal；It is special that the preferred signal for extracting cutting tool state is analyzed using time-domain analysis, frequency-domain analysis and Harmonic wavelet packet
Frequency band energy seek peace as sample characteristics and forms sample set, marks off training sample set and test sample collection；It is excellent using population
The least square method supporting vector machine for changing algorithm constructs cutting tool state identification model, is known automatically with reaching Cutter wear state
Other purpose.This method establishes well contacting between tool condition monitoring signal and cutting-tool wear state, is cutting tool state
Recognition methods provides new optimum ideals, improves the precision of drilling cutters state recognition, with important theory significance and
Practical application value.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with other
The difference of embodiment, the same or similar parts in each embodiment may refer to each other.For system disclosed in embodiment
For, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is said referring to method part
It is bright.
Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said
It is bright to be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art, foundation
Thought of the invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (10)
1. a kind of cutting-tool wear state recognition methods characterized by comprising
Obtain the Cutting Force Signal and cutting-tool wear state in cutter drilling process；
Noise reduction process is carried out to the Cutting Force Signal using Threshold Denoising Method, the Cutting Force Signal after determining noise reduction；
The feature in Cutting Force Signal after extracting the noise reduction, determines feature samples collection；The feature includes temporal signatures, frequency
Characteristic of field and frequency band energy；
It is input with the feature samples collection, is output with the cutting-tool wear state, establishes and be based on least square supporting vector
The cutting tool state identification model of machine；
The current state of wear of the cutter is identified according to the cutting tool state identification model.
2. cutting-tool wear state recognition methods according to claim 1, which is characterized in that described to extract the cutting force letter
Number feature, determine feature samples collection, specifically include:
Temporal signatures and frequency domain character in the Cutting Force Signal are extracted using temporal analysis and frequency domain analysis, really
Determine feature samples collection.
3. cutting-tool wear state recognition methods according to claim 1, which is characterized in that described to extract the cutting force letter
Number feature, determine feature samples collection, specifically include:
Frequency band energy in the Cutting Force Signal is extracted using Harmonic wavelet packet analytic approach, determines feature samples collection.
4. cutting-tool wear state recognition methods according to claim 1, which is characterized in that described with the feature samples collection
For input, it is output with the cutting-tool wear state, establishes the cutting tool state identification model based on least square method supporting vector machine,
It specifically includes:
The feature samples collection is divided into training set and test set；
It is input with the training set, is output with the cutting-tool wear state, is established using particle swarm optimization algorithm based on most
Small two multiply the cutting tool state identification model of support vector machines.
5. cutting-tool wear state recognition methods according to claim 4, which is characterized in that it is described with the training set be it is defeated
Enter, be output with the cutting-tool wear state, establishes the knife based on least square method supporting vector machine using particle swarm optimization algorithm
After tool state recognition model, further includes:
The cutting tool state identification model is verified using the test set.
6. cutting-tool wear state recognition methods according to claim 4, which is characterized in that it is described with the training set be it is defeated
Enter, be output with the cutting-tool wear state, establishes the knife based on least square method supporting vector machine using particle swarm optimization algorithm
Have state recognition model, specifically include:
Penalty factor and nuclear parameter are determined using particle swarm optimization algorithm；
Radial basis kernel function is determined according to the penalty factor and the nuclear parameter；
The cutting tool state identification model based on least square method supporting vector machine is established according to the Radial basis kernel function.
7. a kind of cutting-tool wear state identifying system characterized by comprising
Cutting Force Signal obtains module, for obtaining Cutting Force Signal and tool wear shape in cutter drilling process
State；
Noise reduction process module determines noise reduction for carrying out noise reduction process to the Cutting Force Signal using Threshold Denoising Method
Cutting Force Signal afterwards；
Feature samples collection determining module determines feature samples collection for extracting the feature in the Cutting Force Signal after the noise reduction；
The feature includes temporal signatures, frequency domain character and frequency band energy；
Cutting tool state identification model establishes module, for being input with the feature samples collection, is with the cutting-tool wear state
The cutting tool state identification model based on least square method supporting vector machine is established in output；
Current state of wear identification module, for identifying the current abrasion shape of the cutter according to the cutting tool state identification model
State.
8. cutting-tool wear state identifying system according to claim 7, which is characterized in that the feature samples collection determines mould
Block specifically includes:
Fisrt feature sample set determination unit, for extracting the Cutting Force Signal using temporal analysis and frequency domain analysis
In temporal signatures and frequency domain character, determine feature samples collection.
9. cutting-tool wear state identifying system according to claim 7, which is characterized in that the feature samples collection determines mould
Block specifically includes:
Second feature sample set determination unit, for extracting the frequency band in the Cutting Force Signal using Harmonic wavelet packet analytic approach
Energy determines feature samples collection.
10. cutting-tool wear state identifying system according to claim 7, which is characterized in that the cutting tool state identifies mould
Type is established module and is specifically included:
Division unit, for the special sample set to be divided into training set and test set；
Cutting tool state identification model establishes unit, is output with the cutting-tool wear state for being input with the training set,
The cutting tool state identification model based on least square method supporting vector machine is established using particle swarm optimization algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910053100.3A CN109605127A (en) | 2019-01-21 | 2019-01-21 | A kind of cutting-tool wear state recognition methods and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910053100.3A CN109605127A (en) | 2019-01-21 | 2019-01-21 | A kind of cutting-tool wear state recognition methods and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109605127A true CN109605127A (en) | 2019-04-12 |
Family
ID=66017820
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910053100.3A Pending CN109605127A (en) | 2019-01-21 | 2019-01-21 | A kind of cutting-tool wear state recognition methods and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109605127A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN110153801A (en) * | 2019-07-04 | 2019-08-23 | 西南交通大学 | A kind of cutting-tool wear state discrimination method based on multi-feature fusion |
| CN110515364A (en) * | 2019-07-15 | 2019-11-29 | 北京工业大学 | A kind of cutting-tool wear state detection method based on variation mode decomposition and LS-SVM |
| CN110647943A (en) * | 2019-09-26 | 2020-01-03 | 西北工业大学 | Cutting tool wear monitoring method based on evolutionary data cluster analysis |
| CN110682159A (en) * | 2019-09-25 | 2020-01-14 | 武汉誉德节能数据服务有限公司 | Cutter wear state identification method and device |
| CN111791090A (en) * | 2020-07-02 | 2020-10-20 | 重庆邮电大学 | Cutter life abrasion judgment method based on edge calculation and particle swarm optimization |
| CN113334144A (en) * | 2021-05-20 | 2021-09-03 | 武汉理工大学 | System and method for detecting wear state of milling cutter |
| CN113554621A (en) * | 2021-07-23 | 2021-10-26 | 江苏科技大学 | Tool wear state identification system and method based on wavelet scale map and depth migration learning |
| CN114453630A (en) * | 2022-01-20 | 2022-05-10 | 湖北文理学院 | Method and device for controlling machine tool to mill non-stick tool, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09323239A (en) * | 1996-04-01 | 1997-12-16 | Mitsubishi Materials Corp | Wear-out detector of rotary cutting tool and detecting method |
| CN102152172A (en) * | 2011-02-24 | 2011-08-17 | 天津大学 | Cutter abrasion monitoring method based on co-integration modeling |
| CN105834835A (en) * | 2016-04-26 | 2016-08-10 | 天津大学 | Method for monitoring tool wear on line based on multiscale principal component analysis |
| CN106002487A (en) * | 2016-07-30 | 2016-10-12 | 湖北知本信息科技有限公司 | Diamond cutter abrasion monitoring method based on cutting force model |
| CN106392774A (en) * | 2016-10-14 | 2017-02-15 | 山东理工大学 | Fractal-theory-based online monitoring method for tool wear condition |
| CN107186547A (en) * | 2017-05-25 | 2017-09-22 | 重庆大学 | Numerical control turning batch machining tool wear on-line monitoring method based on cutting power |
-
2019
- 2019-01-21 CN CN201910053100.3A patent/CN109605127A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09323239A (en) * | 1996-04-01 | 1997-12-16 | Mitsubishi Materials Corp | Wear-out detector of rotary cutting tool and detecting method |
| CN102152172A (en) * | 2011-02-24 | 2011-08-17 | 天津大学 | Cutter abrasion monitoring method based on co-integration modeling |
| CN105834835A (en) * | 2016-04-26 | 2016-08-10 | 天津大学 | Method for monitoring tool wear on line based on multiscale principal component analysis |
| CN106002487A (en) * | 2016-07-30 | 2016-10-12 | 湖北知本信息科技有限公司 | Diamond cutter abrasion monitoring method based on cutting force model |
| CN106392774A (en) * | 2016-10-14 | 2017-02-15 | 山东理工大学 | Fractal-theory-based online monitoring method for tool wear condition |
| CN107186547A (en) * | 2017-05-25 | 2017-09-22 | 重庆大学 | Numerical control turning batch machining tool wear on-line monitoring method based on cutting power |
Non-Patent Citations (3)
| Title |
|---|
| 李威霖: "车铣刀具磨损状态监测及预测关键技术研究", 《中国博士学位论文全文数据库 工程科技I辑》 * |
| 李威霖等: "基于粒子群优化LS-SVM的车刀磨损量识别技术研究", 《计算机应用研究》 * |
| 董彩云等: "基于谐波小波包和BSA优化LS-SVM的铣刀磨损状态识别研究", 《中国机械工程》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN110153801A (en) * | 2019-07-04 | 2019-08-23 | 西南交通大学 | A kind of cutting-tool wear state discrimination method based on multi-feature fusion |
| CN110515364A (en) * | 2019-07-15 | 2019-11-29 | 北京工业大学 | A kind of cutting-tool wear state detection method based on variation mode decomposition and LS-SVM |
| CN110515364B (en) * | 2019-07-15 | 2021-06-11 | 北京工业大学 | Cutter wear state detection method based on variational modal decomposition and LS-SVM |
| CN110682159A (en) * | 2019-09-25 | 2020-01-14 | 武汉誉德节能数据服务有限公司 | Cutter wear state identification method and device |
| CN110647943A (en) * | 2019-09-26 | 2020-01-03 | 西北工业大学 | Cutting tool wear monitoring method based on evolutionary data cluster analysis |
| CN111791090A (en) * | 2020-07-02 | 2020-10-20 | 重庆邮电大学 | Cutter life abrasion judgment method based on edge calculation and particle swarm optimization |
| CN113334144A (en) * | 2021-05-20 | 2021-09-03 | 武汉理工大学 | System and method for detecting wear state of milling cutter |
| CN113554621A (en) * | 2021-07-23 | 2021-10-26 | 江苏科技大学 | Tool wear state identification system and method based on wavelet scale map and depth migration learning |
| CN113554621B (en) * | 2021-07-23 | 2023-09-22 | 江苏科技大学 | Cutter wear state identification system and method based on wavelet scale map and deep migration learning |
| CN114453630A (en) * | 2022-01-20 | 2022-05-10 | 湖北文理学院 | Method and device for controlling machine tool to mill non-stick tool, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109605127A (en) | A kind of cutting-tool wear state recognition methods and system | |
| CN107956708B (en) | A kind of potential cavitation fault detection method of pump based on quick spectrum kurtosis analysis | |
| CN107378641B (en) | A kind of Monitoring Tool Wear States in Turning based on characteristics of image and LLTSA algorithm | |
| CN111037365B (en) | Cutter state monitoring data set enhancing method based on generative countermeasure network | |
| CN104723171B (en) | Cutter wear monitoring method based on current and acoustic emission compound signals | |
| CN106330385A (en) | Interference type identification method | |
| CN113074967B (en) | Abnormal sound detection method and device, storage medium and electronic equipment | |
| CN109545242A (en) | A kind of audio data processing method, system, device and readable storage medium storing program for executing | |
| CN102760444B (en) | Support vector machine based classification method of base-band time-domain voice-frequency signal | |
| CN107350900B (en) | A kind of tool condition monitoring method extracted based on the chip breaking time | |
| CN107274915A (en) | A kind of DAB of feature based fusion distorts automatic testing method | |
| He et al. | Feature extraction of acoustic signals based on complex Morlet wavelet | |
| CN103698687A (en) | Method and system for processing signals of hardware Trojan detection in integrated circuit | |
| CN103220241A (en) | Method for extracting box-dimension features from signals at low signal-to-noise ratio condition | |
| WO2022143502A1 (en) | Hob performance degradation tendency evaluation method | |
| CN111585662B (en) | Classification identification and parameter estimation method and system for phase modulation signal | |
| CN105118511A (en) | Thunder identification method | |
| CN114346761B (en) | Cutter abrasion condition detection method based on improved condition generation countermeasure network | |
| CN102411711A (en) | Finger vein recognition method based on individualized weight | |
| CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| Wu et al. | Envelope harmonic noise ratio based adaptive kurtogram and its application in bearing compound fault identification | |
| CN112688324B (en) | Power system low-frequency oscillation mode identification method based on FastICA and TLS-ESPRIT | |
| CN104708497A (en) | Tool wear monitoring system based on current and sound emission composite signals | |
| CN111600878A (en) | Low-rate denial of service attack detection method based on MAF-ADM | |
| CN103473640A (en) | Processing progress information collecting method based on stress state of machine tool cutter |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190412 |
|
| RJ01 | Rejection of invention patent application after publication |