CN110653661A - Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm - Google Patents
Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm Download PDFInfo
- Publication number
- CN110653661A CN110653661A CN201910941657.0A CN201910941657A CN110653661A CN 110653661 A CN110653661 A CN 110653661A CN 201910941657 A CN201910941657 A CN 201910941657A CN 110653661 A CN110653661 A CN 110653661A
- Authority
- CN
- China
- Prior art keywords
- signal
- cutter
- fractal
- cutting force
- signals
- 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
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
- B23Q17/09—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
- B23Q17/0966—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 by measuring a force on parts of the machine other than a motor
-
- 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
-
- 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
- B23Q17/0971—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 by measuring mechanical vibrations of parts of the machine
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to a cutter state monitoring and identifying method based on signal fusion and a multi-fractal spectrum algorithm, which comprises the following steps of: step 1: collecting a cutting force signal and a vibration signal in a cutting process; step 2: and (3) carrying out noise reduction treatment on the cutting force signal and the vibration signal acquired in the step (1), and carrying out step (3): analyzing the multi-fractal characteristics of the signal sequence after noise reduction, searching the relation between the signal and the cutter abrasion through the multi-fractal characteristics, extracting related characteristic vectors from the multi-fractal spectrum obtained through calculation, representing the relation between the signal and the cutter abrasion by using the characteristic vectors, and 4: combining the feature vectors extracted in the step 3 into a feature matrix, using the feature matrix as an input parameter variable, constructing a support vector machine model for monitoring the wear state of the cutter, and diagnosing the cutter state of the unknown state by using the optimized support vector machine model.
Description
Technical Field
The invention relates to the technical field of machining and advanced manufacturing, in particular to a cutter state monitoring and identifying method based on signal fusion and a multi-fractal spectrum algorithm.
Background
The intelligent monitoring technology for the cutter state is the fusion of technologies in multiple aspects including sensor application, digital signal identification and analysis, computer programming, artificial intelligent machine learning and the like, has a remarkable promoting effect on the automation of machining and the unmanned manufacturing process, and occupies an increasingly important position in the field of intelligent manufacturing.
Currently, the intelligent monitoring of the cutter state mostly uses a single signal for monitoring, including other signals such as cutting force signals, vibration signals, acoustic emission signals and the like. The single signal monitoring has the advantages of simple process and easy operation, but has the disadvantages. Each signal can reflect the current tool wear state in the machining process on one aspect, but the current tool state cannot be comprehensively represented due to too large surface of the signal, so that the tool wear state is easily misjudged, and the tool wear state is also easily influenced by machining parameters, machine tool rigidity, workpiece material characteristics and surrounding environment noise. Various sensors are used for acquiring different signals, the current cutter state is mapped from different angles, the advantages and the characteristics of the different signals are fully utilized, and the cutter information is comprehensively reflected. Therefore, a signal fusion technology for monitoring the state of the tool is required.
Many people in the current tool state can only extract time domain characteristics and frequency domain characteristics of signals in monitoring, wherein the time domain characteristics and the frequency domain characteristics comprise mean values, variances, skewness, kurtosis, barycentric frequencies, frequency variances and the like. The inventor finds that most of extracted features are feature quantities obtained based on statistics, and although the extracted features have universality in the field of fault diagnosis, the relation between signals and tool wear cannot be accurately established only by utilizing the features of various related signals generated in the cutting process so as to judge the state of a tool. Therefore, aiming at the cutting process, analyzing the intrinsic mechanism of the cutting process, researching the uniqueness of the signal generated by the cutting process, and extracting the signal characteristics through the uniqueness to establish the relation between the signal and the tool wear is urgent.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, provides a cutter state monitoring and identifying method based on signal fusion and a multi-fractal spectrum algorithm, and has the advantages of good identification effect and high identification rate.
In order to achieve the purpose, the invention adopts the following technical scheme:
the cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm comprises the following steps:
step 1: and collecting a cutting force signal and a vibration signal of the cutter in the cutting process.
Step 2: and (3) denoising the cutting force signal and the vibration signal acquired in the step (1) by using a wavelet threshold denoising method.
And step 3: analyzing the multi-fractal characteristics of the signal by utilizing a multi-fractal detrending fluctuation analysis method for the signal sequence after noise reduction, searching the relation between the signal and the cutter abrasion through the multi-fractal characteristics, extracting related characteristic vectors from a multi-fractal spectrum obtained through calculation, and representing the relation between the signal and the cutter abrasion by using the characteristic vectors.
And 4, step 4: combining the feature vectors extracted in the step 3 into a feature matrix, using the feature matrix as an input parameter variable, determining important parameters in the model by using a parameter optimization method, constructing a support vector machine model for monitoring the wear state of the cutter, optimizing the support vector machine model, and diagnosing the state of the cutter in an unknown state by using the optimized support vector machine model.
Further, the specific steps of step 2 are:
step (1): and respectively performing wavelet decomposition on the collected cutting force signals and vibration signals by adopting a set wavelet basis function and a set decomposition layer number to obtain signals of different frequency segments, wherein the signals comprise low-frequency signals and high-frequency signals of different quantities.
Step (2): and performing threshold quantization processing on the cutting force and vibration signals of the high-frequency segment obtained by wavelet decomposition by adopting a set threshold and a set threshold function, removing signals higher than the threshold, and reserving signals lower than the threshold.
And (3): and performing wavelet reconstruction on the signal component obtained after the threshold value quantization and the signal component of the low frequency band to obtain a vibration signal of the cutting force signal subjected to wavelet noise reduction.
Further, in the step (1), a db3 wavelet basis function is selected to perform 4-layer wavelet decomposition on the signal.
Further, the specific steps of step 3 are:
step (a): for a time series of length N { xi I 1,2, … N, and calculating xiThe difference y (i) between the value of (a) and the mean value, xiThe collected cutting force value or vibration amplitude value.
Step (b): dividing the sequence y (i) into m subsequences according to the direction from the head end to the tail end, dividing the sequence y (i) into m subsequences according to the direction from the tail end to the head end, wherein the length of the subsequences is s when the subsequences are divided, and obtaining 2m subsequences, and m is int (N/s)
Step (c): and (3) fitting the local trend of each subsequence by adopting a least square method:
yv(i)＝a0+a1i+a2i2+...+akik,i＝1,2,...；k＝1,2,...
wherein k is a polynomial order, akIs a polynomial coefficient;
step (d): calculating a mean square error function based on the results of steps (a) and (c):
when v is 1,2, … m,
when v is m +1, m +2 … 2m,
a step (e): calculating a q-order ripple function from the result of step (d):
step (f): obtaining a q-order fluctuation function F according to step (e)qPower law relationship between(s) and subsequence length s: fq(s)～sh(q)To obtain a generalized hurst index h (q).
Step (g): and (f) obtaining the relation between the singularity index and the multi-fractal spectrum according to the generalized hessian index obtained in the step (f).
Further, the specific method of the step (a) is as follows:
wherein the content of the first and second substances,
further, the specific steps of step (f) are as follows: and obtaining a certain linear relation between the q-order fluctuation function and the subsequence length s in a log-log function coordinate system, wherein the slope of the linear relation is a generalized hestert index h (q).
Further, the specific steps of the step (g) are as follows: the generalized hurst index h (q) has the following relationship with the quality index τ (q):
τ(q)＝qh(q)-1
the relationship between the singularity index alpha and the multi-fractal spectrum f (alpha) can be obtained through Legendre transformation:
α＝τ′(q)＝h(q)+qh′(q)
f(α)＝qα-τ(q)
further, in the step (3), the extracted feature vector is: the multi-fractal spectrum characteristic vector corresponding to each component of the cutting force signal and the vibration signal is as follows: (alphamin，f(αmin)，αmax，f(αmax)，α0，△α，△f(α))。
Wherein alpha isminRepresents the minimum value of alpha, alphamaxRepresents the maximum value of alpha, alpha0Is a corresponding value of α when f (α) is maximum, and Δ α ═ αmax-αmin，Δf(α)＝f(αmax)-f(αmin)。
The invention has the beneficial effects that:
1. the invention adopts a multi-fractal detrending fluctuation analysis method to analyze signals, has local dynamic characteristics and multi-analysis characteristics, extracts multi-fractal spectrum parameters as characteristic vectors, can more comprehensively represent the characteristics of cutting signals, mainly shows the local dynamic characteristics and the multi-analysis characteristics, has comprehensive characteristic description of the signals, realizes obvious difference of the initial state, the normal wear state and the rapid wear state of a cutter, obtains higher recognition effect in a support vector machine model, and has the recognition rate of more than 90 percent.
2. The invention removes noise by adopting a wavelet noise reduction mode to the collected cutting force signal and vibration signal so as to obtain better signal characteristics. And selecting a db3 wavelet basis function to perform 4-layer wavelet decomposition on the signal, selecting a heuristic threshold to perform threshold quantization processing on the high-frequency coefficient, and then performing wavelet reconstruction, thereby realizing the removal of signal noise and obtaining better signal characteristics.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application.
FIG. 1 is a schematic flow chart of example 1 of the present invention;
FIG. 2 is a waveform of a cutting force signal after noise reduction in an initial wear state of the cutting tool of the present invention;
FIG. 3 is a waveform of a noise-reduced cutting force signal when the tool of the present invention is in a normal wear state;
FIG. 4 is a waveform diagram of a cutting force signal after noise reduction in a state of sharp wear of the cutter according to the present invention;
FIG. 5 is a waveform of a vibration signal after noise reduction in an initial wear state of the tool according to the present invention;
FIG. 6 is a waveform diagram of a vibration signal after noise reduction in a normal wear state of the tool according to the present invention;
FIG. 7 is a waveform diagram of a vibration signal after noise reduction in a state of sharp wear of the cutter according to the present invention;
FIG. 8 is a graph showing the relationship between the q-order wave function of the cutting force signal and the log-log function of the subsequence length in the initial wear state of the cutter according to the present invention;
FIG. 9 is a graph showing the relationship between the q-order wave function of the cutting force signal and the log-log function of the subsequence length in the normal wear state of the cutter according to the present invention;
FIG. 10 is a graph showing the relationship between the q-order wave function of the cutting force signal and the log-log function of the subsequence length in the case of the cutter being worn suddenly;
FIG. 11 is a graph showing the relationship between the q-order wave function of the vibration signal and the log-log function of the subsequence length in the initial wear state of the tool according to the present invention;
FIG. 12 is a graph showing the relationship between the q-order wave function of the vibration signal and the log-log function of the subsequence length in the normal wear state of the tool according to the present invention;
FIG. 13 is a graph showing the relationship between the q-order wave function of the vibration signal and the log-log function of the subsequence length in the case of the sharp wear of the cutter according to the present invention;
FIG. 14 is a graph of generalized Hurst index versus q for various tool states for a cutting force signal according to the present invention;
FIG. 15 is a graph showing the relationship between the generalized Hurst index and q of the vibration signal according to the present invention under different tool conditions;
FIG. 16 is a graph showing the relationship between the singular index and the fractal spectrum of the cutting force signal in the x direction of the cutting tool in different states;
FIG. 17 is a graph showing the relationship between the singular index and the fractal spectrum of the cutting force signal in the y direction under different conditions;
FIG. 18 is a graph showing the relationship between the singular index and the fractal spectrum of the cutting force signal in the z direction under different conditions;
FIG. 19 is a graph showing the relationship between the singular index and the fractal spectrum of the vibration signal in the x direction of the tool in different states;
FIG. 20 is a graph showing the relationship between the singular index and the fractal spectrum of the vibration signal in the y direction of the tool in different states;
FIG. 21 is a graph showing the relationship between the singular index and the fractal spectrum of the vibration signal in the z direction of the tool in different states;
FIG. 22 is a diagram showing the results of the recognition status of the tool of the support vector machine model after completion of step 4 in the present invention;
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
For convenience of description, the words "up", "down", "left" and "right" in the present invention, if any, merely indicate correspondence with up, down, left and right directions of the drawings themselves, and do not limit the structure, but merely facilitate the description of the invention and simplify the description, rather than indicate or imply that the referenced device or element must have a particular orientation, be constructed and operated in a particular orientation, and thus should not be construed as limiting the invention.
As introduced in the background art, the feature vectors acquired by the existing prop state monitoring and identifying method cannot represent the comprehensive features of the signals, and the identifying effect is poor.
In embodiment 1 of an exemplary embodiment of the present application, as shown in fig. 1, a method for monitoring and identifying a tool state based on signal fusion and a multi-fractal spectrum algorithm includes the following steps:
step 1: in the cutting process, a cutting force signal is acquired through a dynamometer arranged below a tool rest, and a vibration signal is acquired through an acceleration sensor arranged on a tool.
Step 2: and (3) denoising the cutting force signal and the vibration signal acquired in the step (1) by using a wavelet threshold denoising method to achieve the purpose of removing cutting noise, so that an effective signal in the cutting process becomes more obvious, and the influence of other factors on subsequent signal analysis is reduced.
The specific steps of the step 2 are as follows:
step (1): and performing 4-layer wavelet decomposition on the cutting force signal and the vibration signal by using a db3 wavelet basis function.
Step (2): and performing threshold quantization processing on the cutting force and vibration signals of the high-frequency segment obtained by wavelet decomposition by adopting a proper threshold and a proper threshold function, namely removing signals higher than the threshold and keeping signals lower than the threshold.
And (3): and performing wavelet reconstruction on the signal component obtained after the threshold value quantization and the signal component of the low frequency band to obtain a vibration signal of the cutting force signal subjected to wavelet noise reduction.
The steps (1) to (3) adopt the existing signal noise reduction processing mode, and the detailed process thereof is not described in detail herein.
Fig. 2 is a cutting force signal acquired after noise reduction in an initial wear state of the tool, fig. 3 is a cutting force signal acquired after noise reduction in a normal wear state of the tool, and fig. 4 is a cutting force signal acquired after noise reduction in a rapid wear state of the tool.
Fig. 5 is a vibration signal acquired after noise reduction in an initial wear state of the tool, fig. 6 is a vibration signal acquired after noise reduction in a normal wear state of the tool, and fig. 7 is a vibration signal acquired after noise reduction in a rapid wear state of the tool.
And step 3: and performing multi-fractal analysis on the cutting force signal and the vibration signal after noise reduction, and extracting multi-fractal spectrum parameters to form a characteristic vector after the multi-fractal analysis.
The step 3 comprises the following specific steps:
step (a): for a time series of length N { xiI is 1,2, … N, N is a natural number, and x is calculatediX, xiThe collected cutting force value or vibration amplitude (including the values in the x direction, the y direction and the z direction) is obtained.
Wherein the content of the first and second substances,
wherein N is 1,2 … N
Step (b): dividing the sequence y (i) into m subsequences which are not overlapped and have the length of s according to the direction from the head end to the tail end, wherein m is int (N/s), wherein s is a natural number, and after the division, the subsequence with the length of s is arranged at the tail end of the sequence, so that the sequence y (i) is divided into m (m is int (N/s)) subsequences which are not overlapped and have the length of s again according to the direction from the tail end to the head end so as not to lose information, and obtaining 2m subsequences in total, wherein m is a natural number.
Step (c): fitting the local trend of each subsequence by using a least square method:
yv(i)＝a0+a1i+a2i2+...+akik,i＝1,2,...；k＝1,2,...(3)
wherein k is a polynomial order, akIs a polynomial coefficient;
step (d): calculating a mean square error function based on the results of steps (a) and (c): when v is 1,2, … m,
when v is m +1, m +2 … 2m,
calculating a q-order ripple function from the result of step (d):
step (f): obtaining a q-order fluctuation function F according to step (e)qPower law relationship between(s) and subsequence length s: fq(s)～sh(q)To obtain a generalized hurst index h (q).
The specific method comprises the following steps: and obtaining a linear relation between the q-order fluctuation function and the subsequence length s in a log-log function coordinate system, wherein the slope of the linear relation is the generalized hestert index h (q).
Fig. 8 is a linear relation diagram of the cutting force signal subsequence length s and the q-order fluctuation function under the initial wear state of the cutter under the condition that q is 10, 0 and 10 under a log-log function coordinate system.
Fig. 9 is a linear relation diagram of the cutting force signal subsequence length s and the q-order fluctuation function under the normal wear state of the cutter under a log-log function coordinate system when q is 10, 0 and 10.
Fig. 10 is a linear relation diagram of the cutting force signal subsequence length s and the q-order fluctuation function under the condition of the cutter under the condition of acute wear when q is 10, 0 and 10 under a log-log function coordinate system.
Fig. 11 is a linear relation diagram of the vibration signal subsequence length s and the q-order fluctuation function of the tool in the initial wear state under a log-log function coordinate system when q is 10, 0 and 10.
Fig. 12 is a linear relation diagram of the vibration signal subsequence length s and the q-order fluctuation function of the tool in a normal wear state under a log-log function coordinate system when q is 10, 0 and 10.
Fig. 13 is a linear relation diagram of the vibration signal subsequence length s and the q-order fluctuation function of the tool in the state of abrupt wear under a log-log function coordinate system when q is 10, 0 and-10.
8-13, the length s of the subsequence and the log-log function of the q-order fluctuation function are approximately linear, and a slope value, i.e. the generalized hester index h (q), can be obtained at a set q value.
Different values of h and q can be obtained by taking different values of q, and further a relation graph of the generalized hurst exponent and q is obtained
Fig. 14 shows a generalized hessian index of the cutting force signal with respect to q, and fig. 15 shows a generalized hessian index of the vibration signal with respect to q.
And (g) obtaining the relation between the singularity index and the multi-fractal spectrum according to the generalized hessian index obtained in the step (f).
The specific steps of the step (g) are as follows: the generalized hurst index h (q) has the following relationship with the quality index τ (q):
τ(q)＝qh(q)-1 (8)
the relationship between the singularity index alpha and the multi-fractal spectrum f (alpha) can be obtained through Legendre transformation:
α＝τ'(q)＝h(q)+qh′(q) (9)
f(α)＝qα-τ(q)
further, the relationship between the singular index α of the cutting force signal in the x-direction, y-direction and z-direction and the multi-fractal spectrum f (α) and the relationship between the singular index α of the vibration signal in the x-direction, y-direction and z-direction and the multi-fractal spectrum f (α) are obtained, as shown in fig. 16 to 21.
Extracting multi-fractal spectrum characteristic vectors corresponding to components in three directions of the cutting force signal and components in three directions of the vibration signal: (alphamin，f(αmin)，αmax，f(αmax)，α0Δ α, Δ f (α)) has 42-dimensional feature vectors of 7x 6.
Wherein alpha isminRepresents the minimum value of alpha, alphamaxRepresents the maximum value of alpha, alpha0Is a corresponding value of α when f (α) is maximum, and Δ α ═ αmax-αmin，Δf(α)＝f(αmax)-f(αmin)。
And 4, step 4: and (g) inputting the 42-dimensional feature vectors extracted in the step (g) into a support vector machine model as a data set, carrying out tool state identification diagnosis, and adopting the existing method for the tool state identification diagnosis by using the support vector machine model.
The training set is input into a support vector machine model to train model parameters, the training set and the test set are normalized first, and specifically, data are scaled according to proportion, so that the data with different sizes in different intervals fall into the same numerical value interval according to the original size sequence and rule, and the interval is usually a [0,1] interval. In this embodiment, the data is normalized, that is, the data is mapped to the [0,1] interval uniformly. This method normalizes data based on the mean and standard deviation of the raw data. The processed data are in accordance with the standard normal distribution, namely the mean value is 0, the standard deviation is 1, and the conversion function is as follows: x is x-mu sigma, wherein mu is the mean value of all sample data, sigma is the standard deviation of all sample data, then c and g parameters of the model are trained by a grid search method, and finally the parameter c is 4, and g is 32; c is a penalty coefficient, in order to adjust the interval size and the preference weight of two important indexes of classification accuracy, namely tolerance to errors, the higher c is, the more intolerable the errors are, the overfitting condition is easy to occur, namely too many self-characteristics of the learning data set are easy to occur, the smaller c is, the under-fitting condition is easy to occur, namely the characteristics in the data sample are not completely learned, and the c value determines the generalization capability. gamma is a parameter of the RBF function after the RBF function is selected as the kernel. The distribution of the data after mapping from the low-dimensional space to the new high-dimensional feature space is determined. The number of support vectors affects the speed of training and prediction. And inputting the obtained optimized model into a test set for testing, wherein the test method adopts an analysis method of the existing support machine vector machine model, and the detailed process is not further detailed.
As shown in fig. 22, the ordinate 1 represents the initial wear state, 2 represents the normal wear state, 3 represents the rapid wear state, and the tool state recognition accuracy of 92.2% is finally obtained,
in this embodiment, the initial wear state refers to a state in which the actual contact area of the flank face and the workpiece is small when the tool is just put into use, and a unit area bears a large positive pressure, and the wear speed of the tool is fast due to the large roughness and the microscopic unevenness of the flank face of the newly sharpened tool, which is called as an initial wear stage of the tool
The normal wear state means that after the initial wear stage, the contact area of the rear cutter face and the workpiece is increased, the pressure born on the unit area is gradually reduced, and the rough surface of the cutter is gradually smooth, so that the wear enters a relatively stable normal wear period. The abrasion amount of the rear cutter face can be approximately proportionally increased along with the increase of the cutting time in the stage, the abrasion is slow and uniform, the duration is long, and the effective working interval of the cutter is
The rapid wear state is referred to as a rapid wear stage in which the wear rate of the tool is increased by increasing the cutting force and the cutting temperature, increasing the surface roughness of the part, and increasing the wear rate of the tool, because the wear of the tool gradually and uniformly occurs, and the width of the wear zone reaches a certain limit value, and the interaction between the tool and the workpiece is increased after the wear exceeds the limit value.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (8)
1. The cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm is characterized by comprising the following steps of:
step 1: collecting a cutting force signal and a vibration signal of a cutter in a cutting process;
step 2: carrying out noise reduction treatment on the cutting force signal and the vibration signal acquired in the step 1 by using a wavelet threshold noise reduction method;
and step 3: analyzing the multi-fractal characteristics of the signal by utilizing a multi-fractal detrending fluctuation analysis method for the signal sequence after noise reduction, searching the relation between the signal and the cutter abrasion through the multi-fractal characteristics, extracting related characteristic vectors from a multi-fractal spectrum obtained through calculation, and representing the relation between the signal and the cutter abrasion by using the characteristic vectors;
and 4, step 4: combining the feature vectors extracted in the step 3 into a feature matrix, using the feature matrix as an input parameter variable, determining important parameters in the model by using a parameter optimization method, constructing a support vector machine model for monitoring the wear state of the cutter, optimizing the support vector machine model, and diagnosing the state of the cutter in an unknown state by using the optimized support vector machine model.
2. The cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 1, wherein the specific steps of step 2 are:
step (1): respectively performing wavelet decomposition on the collected cutting force signals and vibration signals by adopting a set wavelet basis function and a set decomposition layer number to obtain signals of different frequency segments, wherein the signals comprise low-frequency signals and high-frequency signals of different quantities;
step (2): performing threshold quantization processing on the cutting force and vibration signals of the high-frequency segments obtained by wavelet decomposition by adopting a set threshold and a set threshold function, removing signals higher than the threshold, and reserving signals lower than the threshold;
and (3): and performing wavelet reconstruction on the signal component obtained after the threshold value quantization and the signal component of the low frequency band to obtain a vibration signal of the cutting force signal subjected to wavelet noise reduction.
3. The tool state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 2, wherein in the step (1), db3 wavelet basis function is selected to perform 4-layer wavelet decomposition on the signal.
4. The cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 1, characterized by comprising the following steps:
step (a): for a time series of length N { xiI 1,2, … N, and calculating xiThe difference y (i) between the value of (a) and the mean value, xiThe collected cutting force value or vibration amplitude value;
step (b): dividing a sequence y (i) into m subsequences according to the direction from the head end to the tail end, dividing y (i) into m subsequences according to the direction from the tail end to the head end, wherein the length of the subsequences is s when the subsequences are divided, and 2m subsequences are obtained, and m is int (N/s);
step (c): fitting the local trend of each subsequence by using a least square method:
yv(i)＝a0+a1i+a2i2+...+akik,i＝1,2,...；k＝1,2,...
wherein k is a polynomial order, akIs a polynomial coefficient;
step (d): calculating a mean square error function based on the results of steps (a) and (c):
when v is 1,2, … m,
when v is m +1, m +2 … 2m,
a step (e): calculating a q-order ripple function from the result of step (d):
step (f): obtaining a q-order fluctuation function F according to step (e)qPower law relationship between(s) and subsequence length s: fq(s)～sh(q)Obtaining generalized hurst index h (q);
step (g): and (f) obtaining the relation between the singularity index and the multi-fractal spectrum according to the generalized hessian index obtained in the step (f).
5. The cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 4, wherein the specific method of step (a) is as follows:
wherein the content of the first and second substances,
6. the cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 4, wherein the specific steps of step (f) are: and obtaining a linear relation between the q-order fluctuation function and the subsequence length s under a log-log function coordinate system, wherein the slope of the linear relation is the generalized hestert index h (q).
7. The cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm as claimed in claim 4, wherein the specific steps of step (g) are: the generalized hurst index h (q) has the following relationship with the quality index τ (q):
τ(q)＝qh(q)-1；
the relationship between the singularity index alpha and the multi-fractal spectrum f (alpha) can be obtained through Legendre transformation:
α＝τ′(q)＝h(q)+qh′(q)
f(α)＝qα-τ(q)。
8. the method for monitoring and identifying the cutter state based on the signal fusion and the multi-fractal spectrum algorithm as claimed in claim 4, wherein in the step 3, the extracted feature vectors are as follows: a multi-fractal spectral feature vector (alpha) corresponding to each component of the cutting force signal and the vibration signalmin，f(αmin)，αmax，f(αmax)，α0，△α，△f(α))；
Wherein alpha isminRepresents the minimum value of alpha, alphamaxRepresents the maximum value of alpha, alpha0Is a corresponding value of α when f (α) is maximum, and Δ α ═ αmax-αmin，Δf(α)＝f(αmax)-f(αmin)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910941657.0A CN110653661A (en) | 2019-09-30 | 2019-09-30 | Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910941657.0A CN110653661A (en) | 2019-09-30 | 2019-09-30 | Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110653661A true CN110653661A (en) | 2020-01-07 |
Family
ID=69040255
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910941657.0A Pending CN110653661A (en) | 2019-09-30 | 2019-09-30 | Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110653661A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111716150A (en) * | 2020-06-30 | 2020-09-29 | 大连理工大学 | Evolution learning method for intelligently monitoring cutter state |
| CN111761409A (en) * | 2020-07-09 | 2020-10-13 | 内蒙古工业大学 | Multi-sensor numerical control machine tool cutter wear monitoring method based on deep learning |
| 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 |
| CN112008495A (en) * | 2020-07-28 | 2020-12-01 | 成都飞机工业（集团）有限责任公司 | Cutter damage identification method based on vibration monitoring |
| CN113033486A (en) * | 2021-04-21 | 2021-06-25 | 上海交通大学 | Signal feature extraction and modulation type identification method based on generalized fractal theory |
| CN113752089A (en) * | 2021-10-19 | 2021-12-07 | 山东农业大学 | Cutter state monitoring method based on singularity Leersian index |
| CN114734301A (en) * | 2022-03-29 | 2022-07-12 | 天津大学 | Milling chatter recognition method based on p-leader |
| CN114952419A (en) * | 2022-05-11 | 2022-08-30 | 深圳吉兰丁智能科技有限公司 | Multi-feature fusion based broken cutter monitoring method and electronic equipment |
| CN117515131A (en) * | 2024-01-04 | 2024-02-06 | 之江实验室 | Method, device, storage medium and equipment for monitoring abrasion of planetary reducer |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003013797A1 (en) * | 2001-08-02 | 2003-02-20 | Paul-Heinz Wagner | Method for controlling an intermittently operating screw tool |
| 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 |
| CN108873813A (en) * | 2018-06-25 | 2018-11-23 | 山东大学 | Tool wear degree detection method based on main shaft of numerical control machine tool servo motor current signal |
| CN109612726A (en) * | 2018-11-20 | 2019-04-12 | 南京航空航天大学 | A kind of multiple superstage analysis method extracted for vibration signal characteristics |
-
2019
- 2019-09-30 CN CN201910941657.0A patent/CN110653661A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003013797A1 (en) * | 2001-08-02 | 2003-02-20 | Paul-Heinz Wagner | Method for controlling an intermittently operating screw tool |
| 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 |
| CN108873813A (en) * | 2018-06-25 | 2018-11-23 | 山东大学 | Tool wear degree detection method based on main shaft of numerical control machine tool servo motor current signal |
| CN109612726A (en) * | 2018-11-20 | 2019-04-12 | 南京航空航天大学 | A kind of multiple superstage analysis method extracted for vibration signal characteristics |
Non-Patent Citations (2)
| Title |
|---|
| 关山: "基于NSF-DFA 特征和LS-SVNS 算法的刀具磨损状态识别", 《农业工程学报》 * |
| 李圣怡: "《多传感器融合理论及在智能制造系统中的应用》", 30 November 1998 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111716150A (en) * | 2020-06-30 | 2020-09-29 | 大连理工大学 | Evolution learning method for intelligently monitoring cutter state |
| 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 |
| CN111761409A (en) * | 2020-07-09 | 2020-10-13 | 内蒙古工业大学 | Multi-sensor numerical control machine tool cutter wear monitoring method based on deep learning |
| CN112008495A (en) * | 2020-07-28 | 2020-12-01 | 成都飞机工业（集团）有限责任公司 | Cutter damage identification method based on vibration monitoring |
| CN113033486B (en) * | 2021-04-21 | 2022-11-11 | 上海交通大学 | Signal feature extraction and modulation type identification method based on generalized fractal theory |
| CN113033486A (en) * | 2021-04-21 | 2021-06-25 | 上海交通大学 | Signal feature extraction and modulation type identification method based on generalized fractal theory |
| CN113752089B (en) * | 2021-10-19 | 2023-05-30 | 山东农业大学 | Cutter state monitoring method based on singular Lev index |
| CN113752089A (en) * | 2021-10-19 | 2021-12-07 | 山东农业大学 | Cutter state monitoring method based on singularity Leersian index |
| CN114734301A (en) * | 2022-03-29 | 2022-07-12 | 天津大学 | Milling chatter recognition method based on p-leader |
| CN114734301B (en) * | 2022-03-29 | 2024-04-26 | 天津大学 | Milling chatter identification method based on p-leader |
| CN114952419A (en) * | 2022-05-11 | 2022-08-30 | 深圳吉兰丁智能科技有限公司 | Multi-feature fusion based broken cutter monitoring method and electronic equipment |
| CN114952419B (en) * | 2022-05-11 | 2023-12-22 | 深圳吉兰丁智能科技有限公司 | Multi-feature fusion-based broken cutter monitoring method and electronic equipment |
| CN117515131A (en) * | 2024-01-04 | 2024-02-06 | 之江实验室 | Method, device, storage medium and equipment for monitoring abrasion of planetary reducer |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110653661A (en) | Cutter state monitoring and identifying method based on signal fusion and multi-fractal spectrum algorithm | |
| Yang et al. | Early chatter identification based on an optimized variational mode decomposition | |
| Zheng et al. | Rolling bearing fault detection and diagnosis based on composite multiscale fuzzy entropy and ensemble support vector machines | |
| KR101936283B1 (en) | Diagnostic and prognostics method for machine fault | |
| CN102824172B (en) | EEG (electroencephalogram) feature extraction method | |
| TWI422460B (en) | Tool nose detection method for cutting machine tool | |
| CN110044623B (en) | Intelligent rolling bearing fault identification method based on empirical mode decomposition residual signal characteristics | |
| CN103962888A (en) | Tool abrasion monitoring method based on wavelet denoising and Hilbert-Huang transformation | |
| CN105834835A (en) | Method for monitoring tool wear on line based on multiscale principal component analysis | |
| CN110648480B (en) | Single variable alarm system and method based on change rate | |
| CN102488516A (en) | Nonlinear electroencephalogram signal analysis method and device | |
| CN110259648B (en) | Fan blade fault diagnosis method based on optimized K-means clustering | |
| WO2016004749A1 (en) | Method for recognizing tool abrasion degree of large numerical control milling machine | |
| CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| CN110775758B (en) | Elevator running health degree evaluation method based on car acceleration signal analysis | |
| He et al. | Feature extraction of acoustic signals based on complex Morlet wavelet | |
| CN113752089A (en) | Cutter state monitoring method based on singularity Leersian index | |
| Han et al. | An intelligent fault diagnosis method of variable condition gearbox based on improved DBN combined with WPEE and MPE | |
| CN108956783B (en) | HDP-HSMM-based grinding sound grinding wheel passivation state detection method | |
| CN115741235A (en) | Wear prediction and health management method based on five-axis machining center cutter | |
| CN116871978A (en) | Drilling tool state monitoring method based on multi-sensor fusion | |
| CN114871850A (en) | Cutter wear state evaluation method based on vibration signal and BP neural network | |
| Zhang et al. | An efficient porcine acoustic signal denoising technique based on EEMD-ICA-WTD | |
| CN112454008A (en) | Cutting platform service life prediction method and system based on network model and readable storage medium | |
| Gong et al. | An integrated health condition detection method for rolling bearings using time-shift multi-scale amplitude-aware permutation entropy and uniform phase empirical mode decomposition |
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: 20200107 |
|
| RJ01 | Rejection of invention patent application after publication |