CN115351601B - Tool wear monitoring method based on transfer learning - Google Patents
Tool wear monitoring method based on transfer learning Download PDFInfo
- Publication number
- CN115351601B CN115351601B CN202211199862.2A CN202211199862A CN115351601B CN 115351601 B CN115351601 B CN 115351601B CN 202211199862 A CN202211199862 A CN 202211199862A CN 115351601 B CN115351601 B CN 115351601B
- Authority
- CN
- China
- Prior art keywords
- working condition
- monitoring
- value
- signal
- tool wear
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 115
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000013526 transfer learning Methods 0.000 title claims abstract description 30
- 238000005299 abrasion Methods 0.000 claims abstract description 35
- 230000006870 function Effects 0.000 claims abstract description 23
- 238000009826 distribution Methods 0.000 claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 10
- 238000003754 machining Methods 0.000 claims abstract description 5
- 238000005520 cutting process Methods 0.000 claims description 34
- 238000012545 processing Methods 0.000 claims description 11
- 238000013508 migration Methods 0.000 claims description 10
- 230000005012 migration Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 230000008030 elimination Effects 0.000 claims description 7
- 238000003379 elimination reaction Methods 0.000 claims description 7
- 238000013507 mapping Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 5
- 239000008186 active pharmaceutical agent Substances 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims 2
- 238000003860 storage Methods 0.000 claims 1
- 238000013135 deep learning Methods 0.000 abstract description 3
- 239000000463 material Substances 0.000 description 12
- 238000004519 manufacturing process Methods 0.000 description 9
- 238000011161 development Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000006978 adaptation Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000007514 turning Methods 0.000 description 1
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Mechanical Engineering (AREA)
- Machine Tool Sensing Apparatuses (AREA)
- Numerical Control (AREA)
Abstract
The tool wear monitoring method based on transfer learning solves the problem that the current tool wear monitoring model based on deep learning can only model for a single working condition, and belongs to the technical field of tool wear state prediction of a numerical control machining center. The invention comprises the following steps: s1, collecting monitoring signals of a machining process under an original working condition and a new working condition; s2, acquiring cutter wear value labels under an original working condition and a new working condition, and integrating the cutter wear value labels with a monitoring signal to obtain an original working condition data set and a new working condition data set with labels; s3, constructing a transfer learning model based on edge distribution self-adaption; s4, training by using the data set with the minimized transfer learning model loss function as a target, and completing training to obtain a new working condition cutter abrasion monitoring model; s5, acquiring a monitoring signal of the current moment of the new working condition, and inputting the monitoring signal into a new working condition cutter abrasion monitoring model for prediction.
Description
Technical Field
The invention belongs to the technical field of cutter wear state prediction of a numerical control machining center, and particularly relates to a cutter wear monitoring method based on transfer learning.
Background
The cutter performance is a basic guarantee of production efficiency and processing quality in the actual production of enterprises, and supports the production technical level and economic benefit of the whole manufacturing field. Accurate prediction cutter wear value can guarantee part machining precision when maximize cutter utilization ratio, reduction in production cost, avoids equipment damage. In actual production of factories, cutters of the same model often face different workpiece materials for cutting processing, and due to different physical and chemical properties of the workpiece materials, degradation processes of the cutters in the cutting processing also show different trends, but a wearing label collecting process is complicated, so how to quickly adapt a monitoring model trained by using historical data and corresponding wearing value labels to a new workpiece material processing process with only a small number of data samples for application is a problem to be solved currently.
Under the conditions of a large amount of historical data, corresponding abrasion label values and backgrounds, monitoring data of a small amount of new workpiece material cutting process are collected, the rapid adaptation migration method of the research model can avoid the need of collecting a large amount of new workpiece material cutting process monitoring data for reestablishing the monitoring model, time is saved, the development efficiency of the model is improved, and the method is a development need of the current tool state online monitoring technology.
Along with development and application of industrial big data, the traditional cutter wear monitoring method is gradually replaced by a deep network prediction method with characteristic autonomous learning, and the advantage of deep learning on big data processing is more highlighted under the background of massive processing process monitoring data. However, current deep learning models are typically only modeled for a single operating condition, such as when workpiece materials change, existing models typically fail.
Disclosure of Invention
Aiming at the problem that the current cutter wear monitoring model based on deep learning can only model for a single working condition, the invention provides a cutter wear monitoring method based on transfer learning.
The invention discloses a tool wear monitoring method based on transfer learning, which comprises the following steps:
s1, collecting monitoring signals of machining processes under an original working condition and a new working condition, wherein the monitoring signals comprise cutting force signals, vibration signals and sound emission signals;
S2, acquiring cutter wear value labels under an original working condition and a new working condition, and integrating the cutter wear value labels with a monitoring signal to obtain an original working condition data set and a new working condition data set with labels;
S3、
Constructing an edge distribution self-adaption-based transfer learning model, wherein the transfer learning model comprises a feature adapter, two original working condition cutter wear monitoring models, a regression network 1 and a regression network 2;
the original working condition cutter abrasion monitoring model 1 is used for extracting features of a high-dimensional embedding space of monitoring signals in an original working condition data set, and the extracted features are input into the regression network 1;
the original working condition cutter abrasion monitoring model 2 is used for extracting features of a high-dimensional embedding space of monitoring signals in a new working condition data set, and the extracted features are input into the regression network 2;
the regression network 1 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the original working condition data set;
the regression network 2 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the new working condition data set;
the feature adapter is used for acquiring the distribution difference of the features in the original working condition data set and the features in the new working condition data set;
S4, training the migration learning model based on the original working condition data set with the label and the new working condition data set by taking the minimized migration learning model loss function as a target, and training to obtain an original working condition tool wear monitoring model 2 and a regression network 2 as new working condition tool wear monitoring models;
s5, acquiring a monitoring signal of the current moment of the new working condition, and inputting the monitoring signal into a new working condition cutter abrasion monitoring model for prediction.
Preferably, in S4, the transfer learning model loss function is:
In the method, in the process of the invention,
D S represents the features extracted by the original working condition tool wear monitoring model 1;
d T represents the features extracted by the original working condition tool wear monitoring model 2;
MMD 2(DS,DT) represents the distribution differences obtained by the feature adapter;
Representing regression loss of regression network 1;
representing regression loss of regression network 2;
y S represents the tool wear value corresponding to D S for the original operating mode dataset;
y T represents the tool wear value corresponding to D T for the new operating mode dataset;
L represents the total loss.
Preferably, the method further comprises:
S6, forming a time sequence array by the cutter abrasion predicted value at the current moment predicted by the S5 and cutter abrasion predicted values in a plurality of previous time periods, carrying out least square fitting on the formed time sequence array to obtain the predicted value at the current moment, comparing the least square fitted predicted value with the cutter abrasion predicted value of the S5, if the difference value is larger than a set threshold value, using the least square fitted predicted value as the predicted abrasion value at the current moment, otherwise, reserving the cutter abrasion predicted value of the S5, and carrying out the next moment prediction.
Preferably, in S6, a least square fitting is performed on the time sequence of the components, and the objective function and the constraint condition include;
s.t.2θ2xi+θ1≥0
Wherein y i represents the predicted wear value at a certain time in the time sequence of predicted wear values of the tool;
x i represents the corresponding cutting stroke number; θ 0、θ1、θ2 represents the quadratic polynomial coefficient.
Preferably, in the step S1, the cutting force signal is acquired by arranging a cutting force sensor at the bottom of the workpiece, the vibration signal is acquired by arranging vibration sensors at the workpiece and the machine tool table, and the acoustic emission signal is acquired by arranging acoustic emission sensors at the machine tool table.
Preferably, the monitoring signal in S2 is a signal after preprocessing, and the preprocessing includes:
S21, performing offset elimination on the monitoring signal after cutting feeding starts, setting a time period, calculating an average value of each sensor signal, subtracting the average value from the signal as an offset value, and simultaneously calculating the standard deviation of the signal as a standard deviation basic value to represent the fluctuation degree of the monitoring signal when cutting is not performed;
s22, S22: periodically calculating the standard deviation of the signal in the feeding process, marking the moment as a cutting starting point when the standard deviation is more than 3 times of a basic value, and clearing a monitoring signal before the starting point to finish the elimination of the empty cutting segment;
s23, intercepting monitoring signals of the stable cutting section, and carrying out Z fraction standardization on each path of monitoring signals, wherein the expression is as follows:
wherein x represents the value of a certain sample point on a certain channel signal in the sample;
Mu represents the average value of all sample points of the channel signal data of x;
sigma represents the standard deviation of all sample points of the channel signal data of x;
S24, carrying out downsampling processing on the data after Z score standardization in a sliding window mode, and calculating the first fourth order moment statistic of the data in the sliding window: mean, variance, skewness, and kurtosis, i.e.: the monitoring signal is converted into four-dimensional data.
Preferably, in the step S2, a portable microscope photographing mode is adopted to record the abrasion change trend of the cutter in the cutting process, the abrasion value of the cutter is marked by combining the size, resolution and amplification factor of the microscope pixel, and then the abrasion value and the cutter monitoring signal in the corresponding time period are integrated to obtain the original working condition data set and the new working condition data set with labels.
Preferably, in the step S4, the feature adapter calculates the distribution difference by using a maximum mean difference method:
Wherein, phi (·) represents a mapping for mapping the original variable into a regenerated kernel hilbert space;
d Si represents the ith feature extracted by the original working condition tool wear monitoring model 1, and n features are added;
D Tj represents the j-th feature extracted by the original working condition tool wear monitoring model 2, and m features are taken as the total;
k (·) represents a kernel function;
representing the calculation of the sum of squares of the functions in hilbert space.
The invention has the beneficial effects that: the method and the device realize the purpose of predicting the tool wear value by combining the obtained prediction model trained under the original working condition with the monitoring data under a small amount of new working conditions and rapidly adapting to the processing process of the new workpiece material, effectively reduce the data quantity requirement of the new model training, improve the development efficiency of the monitoring model and ensure the production continuity. The invention realizes predictive maintenance of the numerical control cutter and has important significance for actual production.
Drawings
Fig. 1 is a flowchart of a tool wear monitoring model based on transfer learning according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of source domain vibration signal cancellation of the present invention.
Fig. 3 is a schematic diagram of a deep convolution residual network structure according to the present invention, input represents 4-dimensional time domain statistical feature data of an Input model, conv represents a convolution layer, BN represents a batch normalization layer, reLU represents a Linear rectification function, max Pool represents a maximum pooling layer, adaptive AvgPool represents an adaptive average pooling layer, flat represents a flattening layer, linear represents a Linear layer, and Output represents a predicted value of tool wear Output by the model.
Fig. 4 is a schematic diagram of edge distribution adaptation used in the present invention.
Fig. 5 is a schematic diagram of a migration learning model according to the present invention.
FIG. 6 is a flow chart illustrating the smoothing of predicted wear values in accordance with the present invention.
FIG. 7 is a graph of training results of the transfer learning model of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other.
The invention is further described below with reference to the drawings and specific examples, which are not intended to be limiting.
The tool wear monitoring method based on transfer learning of the embodiment comprises the following steps:
Step 1, collecting monitoring signals of a machining process under an original working condition and a new working condition, wherein the monitoring signals comprise cutting force signals, vibration signals and sound emission signals;
step 2, acquiring cutter wear value labels under an original working condition and a new working condition, and integrating the cutter wear value labels with a monitoring signal to obtain an original working condition data set and a new working condition data set with labels;
Step 3, constructing an edge distribution self-adaption-based transfer learning model, wherein the transfer learning model comprises a characteristic adapter, two original working condition cutter wear monitoring models and two regression networks;
The original working condition cutter abrasion monitoring model 1 is used for extracting features of a high-dimensional embedding space of monitoring signals in an original working condition data set, and the extracted features are respectively input into the regression network 1;
The original working condition cutter abrasion monitoring model 2 is used for extracting features of a high-dimensional embedding space of monitoring signals in a new working condition data set, and the extracted features are respectively input into two regression networks 2;
the regression network 1 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the original working condition data set;
the regression network 2 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the new working condition data set;
the feature adapter is used for acquiring the distribution difference of the features in the original working condition data set and the features in the new working condition data set;
step 4, training the migration learning model based on the original working condition data set with the label and the new working condition data set by taking the minimized migration learning model loss function as a target, and training to obtain an original working condition tool wear monitoring model 2 and a regression network 2 as new working condition tool wear monitoring models;
And 5, acquiring a monitoring signal of the current moment of the new working condition, and inputting the monitoring signal into a new working condition cutter abrasion monitoring model for prediction.
According to the method, the concept of transfer learning is utilized, the inherent similarity between prediction tasks is utilized to the maximum extent, the inherent relation and difference of the tool degradation trend under two working conditions are searched, the characteristic edge distribution adaptation under the two working conditions is carried out by utilizing an algorithm, and then the prediction model applied to the new processing material is obtained. The method comprises the steps of collecting monitoring signals under an original workpiece material working condition and a new workpiece material working condition, obtaining cutter abrasion labels under the original working condition and a small amount of new working conditions, and integrating to obtain an original working condition data set and a new working condition data set with labels; and a migration learning model is built, and then a monitoring model under a new working condition is built through migration learning, so that the requirement on target domain tagged data is reduced, and meanwhile, the problem of generalization application of the original tool wear monitoring model when workpiece materials change is solved.
In the preferred embodiment, in step 1, the cutting force signal is acquired by arranging a cutting force sensor at the bottom of the workpiece, the vibration signal is acquired by arranging vibration sensors at the workpiece and the machine tool table, and the acoustic emission signal is acquired by arranging acoustic emission sensors at the machine tool table.
In the preferred embodiment, in step 2, the monitoring signal is a preprocessed signal, a portable microscope photographing mode is adopted to record the abrasion change trend of the cutter in the cutting process, the cutter abrasion value is marked by combining the size, resolution and amplification factor of the microscope pixel, and then the abrasion value and the cutter monitoring signal in a corresponding time period are integrated to obtain an original working condition data set and a new working condition data set with labels.
The pretreatment step comprises the following steps:
step 21, monitoring signal offset elimination is performed after cutting feed begins. Calculating the average value of the sensor signals in the 120ms time period after the start of the feeding movement, subtracting the average value from the signals as an offset value, and simultaneously calculating the standard deviation of the signals in the process as a standard deviation basic value to represent the fluctuation degree of the monitoring signals when the cutting is not performed;
Step 22, calculating a standard deviation every 400ms in the feeding process, when the standard deviation is 3 times greater than a basic value, marking the moment as a cutting starting point, clearing a monitoring signal before the starting point, and completing the blank segment elimination, wherein the process of the blank segment elimination is shown in fig. 2;
Step 23, intercepting a monitoring signal of a stable cutting segment, reserving sampling points with the length of the monitoring signal being 100000, and carrying out Z fraction standardization on each path of monitoring signal, wherein the expression is as follows:
wherein x represents the value of a certain sample point on a certain channel signal in the sample;
Mu represents the average value of all sample points of the channel signal data of x;
sigma represents the standard deviation of all sample points of the channel signal data of x;
Step 24, performing downsampling processing on the data after Z score standardization in a sliding window mode, and calculating the first fourth order moment statistic of the data in the sliding window: mean, variance, skewness, and kurtosis, i.e.: converting the monitoring signal into four-dimensional data, so that the sample length is reduced, the calculated amount of a model is reduced, and the calculation formula of the statistic comprises the following steps:
Wherein, Representing the data mean; x ms represents the data variance; alpha represents data skewness; beta represents data kurtosis;
x i represents the ith sample point of a certain path of sensing signal; n represents the sample length of the sliding window.
In a preferred embodiment, the original working condition tool wear monitoring model in step 3 adopts a deep convolution residual error network to extract the characteristics of the high-dimensional embedded space of the input signal, the network is mainly formed by stacking residual basic blocks, the parameter setting of the overall architecture of the residual convolution neural network is shown in table 1, and the network structure is shown in fig. 3;
table 1 residual neural network structure settings
The feature adapter adopts the maximum mean value difference as the measurement of the distribution difference of the source domain feature and the target domain feature, and the adopted edge distribution self-adaption principle is shown in figure 4;
the feature adapter calculates distribution differences by using a maximum mean difference method:
Wherein, phi (·) represents a mapping for mapping the original variable into a regenerated kernel hilbert space;
d Si represents the ith feature extracted by the original working condition tool wear monitoring model 1, and n features are added;
d Tj represents the j-th feature extracted by the original working condition tool wear monitoring model 2, and m features are taken as the total; k (·) represents a kernel function; representing the calculation of the sum of squares of the functions in hilbert space.
The kernel function selects a Gaussian radial basis function, and maps the finite dimension data to a high-dimension space, and a calculation formula of the Gaussian radial basis function comprises:
wherein x' represents a kernel function center point;
x represents any point in space;
σ represents the bandwidth used to control the radial range of action of the function.
The regression network is composed of a fully connected neural network, and functions to fit a nonlinear relationship between the high-dimensional embedded features and the tool wear values.
The principle of the transfer learning model is shown in fig. 5.
In a preferred embodiment, the transfer learning model loss function in step 4 is:
Wherein D S represents the characteristics extracted by the original working condition tool wear monitoring model 1;
d T represents the features extracted by the original working condition tool wear monitoring model 2;
MMD 2(DS,DT) represents the distribution differences obtained by the feature adapter;
Representing regression loss of regression network 1;
representing regression loss of regression network 2;
y S represents the tool wear value corresponding to D S for the original operating mode dataset;
y T represents the tool wear value corresponding to D T for the new operating mode dataset;
L represents the total loss.
In a preferred embodiment, the method of this embodiment further includes:
As shown in fig. 6, step 6, forming a time sequence by the predicted value of the cutter wear at the current moment predicted in step 5 and the predicted values of the cutter wear in the previous time periods, performing least square fitting on the formed time sequence to obtain the predicted value of the current moment, comparing the predicted value of the least square fitting with the predicted value of the cutter wear in step 5, if the difference is greater than a set threshold, using the predicted value of the least square fitting as the predicted wear value of the current moment, otherwise, reserving the predicted value of the cutter wear in step 5, and turning to step 5 to predict the next moment.
In the step 6, performing least square fitting on the time sequence formed by the method, wherein the objective function and the constraint condition comprise;
s.t.2θ2xi+θ1≥0
Wherein y i represents the predicted wear value at a certain time in the time sequence of predicted wear values of the tool;
x i represents the corresponding cutting stroke number; θ 0、θ1、θ2 represents the quadratic polynomial coefficient.
The tool wear state monitoring model obtained by transfer learning is used for monitoring the tool wear state of a new workpiece material in the cutting process, as shown in fig. 7, the prediction result approaches to the tool wear true value, so that the tool wear state monitoring model provided by the invention has the capability of working condition transfer, can accurately predict the tool wear state under the new working condition, and has important significance for actual production.
Although the invention herein has been described with reference to particular embodiments, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present invention. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the different dependent claims and the features described herein may be combined in ways other than as described in the original claims. It is also to be understood that features described in connection with separate embodiments may be used in other described embodiments.
Claims (7)
1. A tool wear monitoring method based on transfer learning, the method comprising:
s1, collecting monitoring signals of machining processes under an original working condition and a new working condition, wherein the monitoring signals comprise cutting force signals, vibration signals and sound emission signals;
S2, acquiring cutter wear value labels under an original working condition and a new working condition, and integrating the cutter wear value labels with a monitoring signal to obtain an original working condition data set and a new working condition data set with labels;
S3, constructing an edge distribution self-adaption-based transfer learning model, wherein the transfer learning model comprises a characteristic adapter, two original working condition cutter wear monitoring models, a regression network 1 and a regression network 2;
the original working condition cutter abrasion monitoring model 1 is used for extracting features of a high-dimensional embedding space of monitoring signals in an original working condition data set, and the extracted features are input into the regression network 1;
the original working condition cutter abrasion monitoring model 2 is used for extracting features of a high-dimensional embedding space of monitoring signals in a new working condition data set, and the extracted features are input into the regression network 2;
the regression network 1 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the original working condition data set;
the regression network 2 is used for fitting a nonlinear relation between the input characteristics and the cutter abrasion values in the new working condition data set;
the feature adapter is used for acquiring the distribution difference of the features in the original working condition data set and the features in the new working condition data set;
The feature adapter calculates distribution differences by adopting a maximum mean difference method:
Wherein, phi (·) represents a mapping for mapping the original variable into a regenerated kernel hilbert space;
d Si represents the ith feature extracted by the original working condition tool wear monitoring model 1, and n features are added;
d Tj represents the j-th feature extracted by the original working condition tool wear monitoring model 2, and m features are taken as the total; k (·) represents a kernel function; Representing the calculation of the sum of squares of the functions under the hilbert space;
S4, training the migration learning model based on the original working condition data set with the label and the new working condition data set by taking the minimized migration learning model loss function as a target, and training to obtain an original working condition tool wear monitoring model 2 and a regression network 2 as new working condition tool wear monitoring models;
the transfer learning model loss function is:
Wherein D S represents the characteristics extracted by the original working condition tool wear monitoring model 1;
d T represents the features extracted by the original working condition tool wear monitoring model 2;
MMD 2(DS,DT) represents the distribution differences obtained by the feature adapter;
Representing regression loss of regression network 1;
representing regression loss of regression network 2;
y S represents the tool wear value corresponding to D S for the original operating mode dataset;
y T represents the tool wear value corresponding to D T for the new operating mode dataset;
L represents the total loss;
s5, acquiring a monitoring signal of the current moment of the new working condition, and inputting the monitoring signal into a new working condition cutter abrasion monitoring model for prediction.
2. The method of tool wear monitoring based on transfer learning of claim 1, further comprising:
S6, forming a time sequence array by the cutter abrasion predicted value at the current moment predicted by the S5 and cutter abrasion predicted values in a plurality of previous time periods, carrying out least square fitting on the formed time sequence array to obtain the predicted value at the current moment, comparing the least square fitted predicted value with the cutter abrasion predicted value of the S5, if the difference value is larger than a set threshold value, using the least square fitted predicted value as the predicted abrasion value at the current moment, otherwise, reserving the cutter abrasion predicted value of the S5, and carrying out the next moment prediction.
3. The tool wear monitoring method based on transfer learning according to claim 2, wherein in S6, the least square fitting is performed on the time series sequence of the components, and the objective function and the constraint condition include;
Wherein y i represents the predicted wear value at a certain time in the time sequence of predicted wear values of the tool;
x i represents the corresponding cutting stroke number; θ 0、θ1、θ2 represents the quadratic polynomial coefficient.
4. The tool wear monitoring method based on the transfer learning according to claim 1, wherein in S1, the cutting force signal is acquired by arranging a cutting force sensor at the bottom of the workpiece, the vibration signal is acquired by arranging vibration sensors at the workpiece and the machine table, and the acoustic emission signal is acquired by arranging acoustic emission sensors at the machine table.
5. The tool wear monitoring method based on transfer learning according to claim 4, wherein the monitoring signal in S2 is a pre-processed signal, and the pre-processing includes:
S21, performing offset elimination on the monitoring signal after cutting feeding starts, setting a time period, calculating an average value of each sensor signal, subtracting the average value from the signal as an offset value, and simultaneously calculating the standard deviation of the signal as a standard deviation basic value to represent the fluctuation degree of the monitoring signal when cutting is not performed;
S22, periodically calculating the standard deviation of the signal in the feeding process, and marking the moment as a cutting starting point and clearing a monitoring signal before the starting point when the standard deviation is more than 3 times of a basic value, so as to finish the elimination of the empty cutting segment;
s23, intercepting monitoring signals of the stable cutting section, and carrying out Z fraction standardization on each path of monitoring signals, wherein the expression is as follows:
wherein x represents the value of a certain sample point on a certain channel signal in the sample;
Mu represents the average value of all sample points of the channel signal data of x;
sigma represents the standard deviation of all sample points of the channel signal data of x;
S24, carrying out downsampling processing on the data after Z score standardization in a sliding window mode, and calculating the first fourth order moment statistic of the data in the sliding window: mean, variance, skewness, and kurtosis, i.e.: the monitoring signal is converted into four-dimensional data.
6. The tool wear monitoring method based on transfer learning according to claim 5, wherein in S2, a portable microscope photographing mode is adopted to record a wear change trend of a tool in a cutting process, a tool wear value is marked by combining a microscope pixel size, a resolution and a magnification, and then the wear value and a tool monitoring signal of a corresponding time period are integrated to obtain an original working condition data set and a new working condition data set with labels.
7. A computer-readable storage device storing a computer program, characterized in that the computer program, when executed, implements the tool wear monitoring method based on transfer learning according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211199862.2A CN115351601B (en) | 2022-09-29 | 2022-09-29 | Tool wear monitoring method based on transfer learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211199862.2A CN115351601B (en) | 2022-09-29 | 2022-09-29 | Tool wear monitoring method based on transfer learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115351601A CN115351601A (en) | 2022-11-18 |
| CN115351601B true CN115351601B (en) | 2024-05-28 |
Family
ID=84008301
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211199862.2A Active CN115351601B (en) | 2022-09-29 | 2022-09-29 | Tool wear monitoring method based on transfer learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115351601B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116540627B (en) * | 2023-02-07 | 2024-04-12 | 广东工业大学 | Machine tool thermal error prediction compensation group control method and system based on deep transfer learning |
| CN115847187B (en) * | 2023-02-27 | 2023-05-05 | 成都大金航太科技股份有限公司 | Real-time monitoring system for deep and narrow groove turning |
| CN116992954A (en) * | 2023-09-26 | 2023-11-03 | 南京航空航天大学 | UMAP data dimension reduction-based similarity measurement transfer learning method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005125480A (en) * | 2003-09-30 | 2005-05-19 | Allied Material Corp | Cutting method having function for diagnosing tool life |
| CN102091972A (en) * | 2010-12-28 | 2011-06-15 | 华中科技大学 | Numerical control machine tool wear monitoring method |
| CN102765010A (en) * | 2012-08-24 | 2012-11-07 | 常州大学 | Cutter damage and abrasion state detecting method and cutter damage and abrasion state detecting system |
| CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
| CN109822399A (en) * | 2019-04-08 | 2019-05-31 | 浙江大学 | Cutting tool for CNC machine state of wear prediction technique based on parallel deep neural network |
| CN112192318A (en) * | 2020-09-28 | 2021-01-08 | 北京天泽智云科技有限公司 | Machining tool state monitoring method and system |
| CN112518425A (en) * | 2020-12-10 | 2021-03-19 | 南京航空航天大学 | Intelligent machining cutter wear prediction method based on multi-source sample migration reinforcement learning |
-
2022
- 2022-09-29 CN CN202211199862.2A patent/CN115351601B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005125480A (en) * | 2003-09-30 | 2005-05-19 | Allied Material Corp | Cutting method having function for diagnosing tool life |
| CN102091972A (en) * | 2010-12-28 | 2011-06-15 | 华中科技大学 | Numerical control machine tool wear monitoring method |
| CN102765010A (en) * | 2012-08-24 | 2012-11-07 | 常州大学 | Cutter damage and abrasion state detecting method and cutter damage and abrasion state detecting system |
| CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
| CN109822399A (en) * | 2019-04-08 | 2019-05-31 | 浙江大学 | Cutting tool for CNC machine state of wear prediction technique based on parallel deep neural network |
| CN112192318A (en) * | 2020-09-28 | 2021-01-08 | 北京天泽智云科技有限公司 | Machining tool state monitoring method and system |
| CN112518425A (en) * | 2020-12-10 | 2021-03-19 | 南京航空航天大学 | Intelligent machining cutter wear prediction method based on multi-source sample migration reinforcement learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115351601A (en) | 2022-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115351601B (en) | Tool wear monitoring method based on transfer learning | |
| CN109396576B (en) | Deep learning-based electric spark machining stability and energy consumption state optimization decision system and decision method | |
| CN111633467B (en) | Cutter wear state monitoring method based on one-dimensional depth convolution automatic encoder | |
| WO2019001220A1 (en) | Decision tree based turning process energy consumption prediction system and method of numerically controlled lathe | |
| CN113780153B (en) | Cutter wear monitoring and predicting method | |
| CN111832432B (en) | Cutter wear real-time prediction method based on wavelet packet decomposition and deep learning | |
| CN102763048A (en) | Methods and apparatuses for utilizing adaptive predictive algorithms and determining when to use the adaptive predictive algorithms for virtual metrology | |
| CN117289668B (en) | Distributed speed reducer network cooperative control method, device, equipment and storage medium | |
| Kumar et al. | Performance evaluation for tool wear prediction based on Bi-directional, Encoder–Decoder and Hybrid Long Short-Term Memory models | |
| CN116520772A (en) | Sheet part numerical control machining quality prediction method based on multitasking transfer learning | |
| CN116204774A (en) | Cutter abrasion stability prediction method based on hierarchical element learning | |
| Li et al. | A hybrid remaining useful life prediction method for cutting tool considering the wear state | |
| CN112308298B (en) | Multi-scenario performance index prediction method and system for semiconductor production line | |
| CN113601261A (en) | Monitoring method of online rapid optimization model for cutter | |
| Amar Bouzid et al. | CNC milling cutters condition monitoring based on empirical wavelet packet decomposition | |
| CN117150235A (en) | Cutter life prediction method by cooperation of monitoring data and degradation model | |
| CN113052302A (en) | Machine health monitoring method and device based on recurrent neural network and terminal equipment | |
| CN112858901A (en) | System and method for monitoring operation state and service life prediction of cutter in real time | |
| CN117943892A (en) | Cutter monitoring data prediction optimization method based on time sequence data analysis | |
| CN116992954A (en) | UMAP data dimension reduction-based similarity measurement transfer learning method | |
| CN116821828A (en) | Multi-dimensional time sequence prediction method based on industrial data | |
| CN112488411B (en) | Machining stability assessment method based on approximate cycle process | |
| CN117341261B (en) | Intelligent control method and system for servo direct-drive screw press | |
| CN116540627B (en) | Machine tool thermal error prediction compensation group control method and system based on deep transfer learning | |
| CN118092314B (en) | Method for adjusting operation parameters of machine tool, operation parameter adjusting device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |