CN113560955A - Method, system and application for predicting remaining service life of numerical control machine tool cutter - Google Patents
Method, system and application for predicting remaining service life of numerical control machine tool cutter Download PDFInfo
- Publication number
- CN113560955A CN113560955A CN202110642406.XA CN202110642406A CN113560955A CN 113560955 A CN113560955 A CN 113560955A CN 202110642406 A CN202110642406 A CN 202110642406A CN 113560955 A CN113560955 A CN 113560955A
- Authority
- CN
- China
- Prior art keywords
- signal
- service life
- signals
- machine tool
- numerical control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
-
- 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
- B23Q17/007—Arrangements for observing, indicating or measuring on machine tools for managing machine functions not concerning the tool
- B23Q17/008—Life management for parts of the machine
-
- 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
-
- 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
- B23Q17/0995—Tool life management
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention belongs to the technical field of machinery and discloses a method, a system and application for predicting the residual service life of a numerical control machine tool cutter, wherein the method for predicting the residual service life of the numerical control machine tool cutter comprises the following steps: collecting controller signals and sensor signals in the working process of the numerical control machine tool, preprocessing the signals, extracting features and selecting the features, and mining information related to tool abrasion in various signals; and establishing a tool residual service life prediction model by using a long-time memory network and an attention mechanism, so as to realize the prediction of the residual service life of the numerical control machine tool. According to the method, controller signals and sensor signals in the working process of the numerical control machine tool are collected, the residual service life prediction model of the tool is established by utilizing multi-source information, the tool abrasion conditions reflected by different types of signals are fully considered, the limitation that the prediction model is established by using a single signal in the prior art is effectively overcome, and the generalization capability of the residual service life prediction model of the tool is improved.
Description
Technical Field
The invention belongs to the technical field of machinery, and particularly relates to a method and a system for predicting the residual service life of a numerical control machine tool cutter and application of the method and the system.
Background
At present, as an important part in numerical control machining, the problem caused by tool abrasion is one of the main problems in the numerical control machining process. During milling of numerically controlled machine tools, wear degradation of the tool is inevitable. Once the cutter fails, the surface quality of the workpiece can not meet the requirement, so that the machining efficiency is low, and even the machine tool can be damaged when the abrasion is serious. Therefore, the method for effectively predicting the residual service life of the cutter has very important significance for improving the production efficiency of the numerical control machine tool. At present, a data driving and machine learning combined method is a mainstream method and technology in the technical field of tool life prediction, but certain limitations also exist.
The union technologies, Inc., filed by the patent literature, "tool life prediction method" (patent application No. 201811069400.2, publication No. CN109465676A), proposes a tool life prediction method. The method comprises the following steps: firstly, extracting a characteristic value of a current signal through characteristic learning, then performing data cleaning on the current signal, and then analyzing the relation between the current signal and the service life of a cutter through a machine learning and deep learning method to establish a cutter service life prediction model. The method can realize accurate estimation of the residual life of the cutter and improve the yield of the produced products. However, the method still has the disadvantages that only the spindle current signal of the processing machine is considered, different signals are not considered to reflect different wear conditions of the cutter, the single signal is used for establishing the prediction model, the limitation is large, the generalization capability of the model is poor, and the popularization and the use in actual production are not facilitated.
Through the above analysis, the problems and defects of the prior art are as follows: the existing method for predicting the service life of the cutter by adopting data driving and combining with a machine learning means only considers a spindle current signal of a processing machine and does not consider different signals to reflect different abrasion conditions of the cutter, the method for establishing a prediction model by using a single signal has great limitation, and the model has poor generalization capability and is not beneficial to popularization and use in actual production.
The difficulty in solving the above problems and defects is: the application limitation of modeling by using a single signal is eliminated, so that the model can learn different wear information of the tool from different signals.
The significance of solving the problems and the defects is as follows: in the prior art, the considered signal is too single, information related to cutter abrasion in different signals is extracted and excavated through signal characteristic extraction, and the information related to the abrasion is fused, so that the generalization capability of the model is improved. The invention aims to provide a method for predicting the residual service life of a cutter by multi-source information fusion.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method, a system and an application for predicting the residual service life of a numerical control machine tool cutter, in particular to a method, a system and an application for predicting the residual service life of the numerical control machine tool cutter based on multi-source information fusion, and aims to solve the problem that the generalization capability of a prediction model established by using a single signal in the prior art is poor.
The invention is realized in this way, a method for predicting the remaining service life of a numerical control machine tool cutter, which comprises the following steps:
collecting controller signals and sensor signals in the working process of the numerical control machine tool, preprocessing the signals, extracting features and selecting the features, and mining information related to tool abrasion in various signals; and establishing a tool residual service life prediction model by using a long-time memory network and an attention mechanism, so as to realize the prediction of the residual service life of the numerical control machine tool.
Further, the method for predicting the residual service life of the numerical control machine tool cutter comprises the following steps:
step one, signal acquisition and processing;
step two, extracting signal characteristics;
step three, signal characteristic selection;
and step four, modeling and predicting.
Further, in the first step, the signal acquisition and processing includes:
(1) collecting signals in the working process of the numerical control machine tool, namely a controller signal and a sensor signal, wherein the controller signal mainly comprises a main shaft load and three-direction mechanical coordinates, namely an x-axis mechanical coordinate, a y-axis mechanical coordinate and a z-axis mechanical coordinate, and the sensor signal mainly comprises a current signal and three-direction vibration signals, namely an x-axis vibration signal, a y-axis vibration signal and a z-axis vibration signal;
(2) preprocessing the acquired signals, firstly removing the signals acquired when the signals do not contact with a processing object according to the mechanical coordinates in three directions and the main shaft load, then processing missing values and abnormal values, and removing abnormal trend items in the signals by using a least square method.
Wherein, the removing the abnormal trend term in the signal by using the least square method comprises the following steps:
firstly, fitting a signal by using a high-order polynomial, selecting a proper polynomial coefficient according to the principle of a least square method to minimize the sum of squares of errors between the signal and the fitting signal, then substituting the selected polynomial coefficient into the high-order polynomial to obtain a trend term of the signal, and finally subtracting the trend term from the signal to obtain the signal without the trend term.
Further, in step two, the signal feature extraction includes:
and C, extracting the characteristics of the signal obtained in the step I from the aspects of time domain, frequency domain and time-frequency domain.
Further, in step three, the signal feature selection includes:
(1) and (3) screening the signal characteristics obtained in the second step by using two evaluation criteria of monotonicity and tendency, wherein the two evaluation criteria comprise:
calculating the monotonicity value and the trend value of each signal feature obtained in the step two according to the following formula:
wherein S ismonIs the monotonicity value of the single signal feature, T is the sample length of the single signal feature, and dH represents each sample and the previous sample in the single featureA difference of (d);
wherein S istredIs the trend value of a single signal feature, T is the sample length of the single signal feature, xiFor the ith sample of a single signal feature,
is the mean of the features of the individual signals, tiIs xiThe corresponding accumulated working time is calculated according to the working time,
is the average value of the accumulated working time sequence;
secondly, taking the mean value of the monotonicity value and the trend value of each signal feature as a comprehensive evaluation value, and forming a comprehensive evaluation matrix by the comprehensive evaluation values of all the signal features;
normalizing the comprehensive evaluation matrix, selecting signal characteristics with the comprehensive evaluation value larger than 0.5 to form a matrix, and obtaining the screened signal characteristics.
(2) Performing dimensionality reduction processing on the signal characteristics obtained by screening by using a kernel principal component analysis algorithm, wherein the dimensionality reduction processing comprises the following steps:
firstly, standardizing the signal characteristics obtained by screening;
selecting a kernel function and calculating a kernel matrix;
thirdly, centralizing the kernel matrix to obtain a centralized kernel matrix;
fourthly, calculating the eigenvalue and the eigenvector of the centralized kernel matrix, and ordering the eigenvalue according to the power reduction;
setting the threshold value of the accumulated contribution rate to be 90%, determining the number p of the principal elements, selecting eigenvectors corresponding to the previous p eigenvalues to form a matrix, and obtaining the signal characteristics after dimensionality reduction.
Further, in step four, the modeling prediction includes:
and (4) carrying out normalization processing on the signal characteristics obtained in the step three, taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, constructing a training set, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
Further, the establishing of the long-term and short-term memory network prediction model combining with the attention mechanism includes:
the structure of the long-time and short-time memory network combined with the attention mechanism is as follows in sequence: the number of output neurons of the LSTM network layer is set to be 70, the number of neurons of the full connection layer is set to be 1024, and the number of neurons of the output layer is set to be 1.
Another object of the present invention is to provide a system for predicting remaining useful life of a tool of a numerical control machine tool using the method for predicting remaining useful life of a tool of a numerical control machine tool, the system comprising:
the signal acquisition and processing module is used for acquiring signals in the working process of the numerical control machine tool and preprocessing the acquired signals;
the signal characteristic extraction module is used for extracting the characteristics of the acquired signals from the three aspects of time domain, frequency domain and time-frequency domain;
the signal characteristic selection module is used for screening the signal characteristics by utilizing two evaluation standards of monotonicity and tendency, and then performing dimensionality reduction processing on the screened signal characteristics by utilizing a kernel principal component analysis algorithm;
and the modeling prediction module is used for carrying out normalization processing on the signal characteristics, constructing a training set by taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
It is a further object of the invention to provide a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of:
collecting controller signals and sensor signals in the working process of the numerical control machine tool, preprocessing the signals, extracting features and selecting the features, and mining information related to tool abrasion in various signals; and establishing a tool residual service life prediction model by using a long-time memory network and an attention mechanism, so as to realize the prediction of the residual service life of the numerical control machine tool.
The invention also aims to provide an information data processing terminal which is used for realizing the system for predicting the residual service life of the cutter of the numerical control machine tool.
By combining all the technical schemes, the invention has the advantages and positive effects that: the invention provides a method for predicting the residual service life of a numerical control machine tool cutter, relates to a method for predicting the residual service life of the numerical control machine tool cutter based on multi-source information fusion, and can be used for predicting the residual service life of the numerical control machine tool cutter and solving the problem of model generalization capability of the prediction of the residual service life of the cutter.
According to the method, controller signals and sensor signals in the working process of the numerical control machine tool are collected, the tool residual service life prediction model is established by utilizing multi-source information, the tool abrasion conditions reflected by different types of signals are fully considered, the limitation that the prediction model is established by using a single signal in the prior art is effectively overcome, and the generalization capability of the tool residual service life prediction model is improved.
The method comprehensively extracts the vibration and current signal characteristics from three aspects of time domain, frequency domain and time-frequency domain, and utilizes monotonicity and tendency indexes to screen the characteristics, thereby improving the utilization efficiency of the multi-source signal. Meanwhile, the invention adopts a long-short term memory network to establish a tool residual service life prediction model, and adopts a local attention mechanism-based mode to add an attention mechanism in an output layer, thereby further improving the model prediction precision.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments of the present invention will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart of a method for predicting the remaining service life of a tool of a numerical control machine tool according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a method for predicting the remaining service life of a tool of a numerical control machine tool according to an embodiment of the present invention.
FIG. 3 is a block diagram of a system for predicting the remaining service life of a tool of a numerically-controlled machine tool according to an embodiment of the present invention;
in the figure: 1. a signal acquisition and processing module; 2. a signal feature extraction module; 3. a signal feature selection module; 4. and modeling a prediction module.
FIG. 4 is a diagram illustrating predicted results provided by an embodiment of the present invention.
Fig. 4(a) is a schematic diagram of a predicted result of the tool No. 02 of the experiment provided in the embodiment of the present invention.
Fig. 4(b) is a schematic diagram of a predicted result of the tool No. 03 of the experiment provided in the embodiment of the present invention.
Fig. 4(c) is a schematic diagram of a predicted result of the experiment No. 01 tool according to the embodiment of the present invention.
Fig. 4(d) is a schematic diagram of a predicted result of the experimental tool No. 03 according to the embodiment of the present invention.
Fig. 4(e) is a schematic diagram of a predicted result of the experimental tool iii 01 according to the embodiment of the present invention.
Fig. 4(f) is a schematic diagram of a predicted result of the tool No. three experiment 02 provided in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Aiming at the problems in the prior art, the invention provides a method, a system and an application for predicting the residual service life of a numerical control machine tool cutter, and the invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the method for predicting the remaining service life of the tool of the numerical control machine tool according to the embodiment of the present invention includes the following steps:
s101, signal acquisition and processing: collecting signals in the working process of the numerical control machine tool, and preprocessing the collected signals;
s102, extracting signal characteristics, namely extracting the characteristics of the signal acquired in the S101 from the three aspects of time domain, frequency domain and time-frequency domain;
s103, selecting signal characteristics, screening the signal characteristics by using two evaluation standards of monotonicity and tendency, and then performing dimensionality reduction processing on the screened signal characteristics by using a kernel principal component analysis algorithm;
s104, modeling and predicting: and (3) carrying out normalization processing on the signal characteristics, constructing a training set by taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
The schematic diagram of the method for predicting the residual service life of the numerical control machine tool cutter provided by the embodiment of the invention is shown in FIG. 2.
As shown in fig. 3, the system for predicting the remaining service life of a cutting tool of a numerical control machine according to an embodiment of the present invention includes:
the signal acquisition and processing module 1 is used for acquiring signals in the working process of the numerical control machine tool and preprocessing the acquired signals;
the signal feature extraction module 2 is used for extracting features of the acquired signals from the aspects of time domain, frequency domain and time-frequency domain;
the signal feature selection module 3 is used for screening the signal features by using two evaluation criteria of monotonicity and tendency, and then performing dimensionality reduction processing on the screened signal features by using a kernel principal component analysis algorithm;
and the modeling prediction module 4 is used for carrying out normalization processing on the signal characteristics, taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, constructing a training set, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
The technical solution of the present invention will be further described with reference to the following examples.
Example 1
The steps implemented by the present invention are described in further detail with reference to fig. 2.
And 2, preprocessing the acquired signals, firstly removing the signals acquired when the signals do not contact with the processing object according to the mechanical coordinates in the three directions and the main shaft load, then processing missing values and abnormal values, and removing abnormal trend items in the signals by using a least square method.
The step of removing the trend term in the signal by using a least square method comprises the following steps: firstly, fitting a signal by using a high-order polynomial, selecting a proper polynomial coefficient according to the principle of a least square method to minimize the sum of squares of errors between the signal and the fitting signal, then substituting the selected polynomial coefficient into the high-order polynomial to obtain a trend term of the signal, and finally subtracting the trend term from the signal to obtain the signal without the trend term.
And 2, extracting signal characteristics.
And (3) carrying out feature extraction on the signals obtained in the step (1) from three aspects of time domain, frequency domain and time-frequency domain.
And 3, selecting signal characteristics.
And (3) screening the signal characteristics obtained in the step (2) by using two evaluation criteria of monotonicity and tendency.
And step 1, calculating the monotonicity value and the trend value of each signal characteristic obtained in the step 2 according to the following formula.
Wherein S ismonIs the monotonicity value of the single signal feature, T is the sample length of the single signal feature, and dH represents the difference of each sample in the single feature from the previous sample.
Wherein S istredIs the trend value of a single signal feature, T is the sample length of the single signal feature, xiFor the ith sample of a single signal feature,
is the mean of the features of the individual signals, tiIs xiThe corresponding accumulated working time is calculated according to the working time,
is the average of the accumulated working time series.
And 2, taking the mean value of the monotonicity value and the trend value of each signal feature as a comprehensive evaluation value, and forming a comprehensive evaluation matrix by the comprehensive evaluation values of all the signal features.
And 3, normalizing the comprehensive evaluation matrix, selecting the signal features with the comprehensive evaluation value larger than 0.5 to form a matrix, and obtaining the screened signal features.
And then performing dimensionality reduction on the signal characteristics obtained by screening by using a kernel principal component analysis algorithm.
And step 1, standardizing the signal characteristics obtained by screening.
And step 2, selecting a kernel function and calculating a kernel matrix.
And 3, centralizing the kernel matrix to obtain a centralized kernel matrix.
And 4, calculating the eigenvalue and the eigenvector of the centralized kernel matrix, and sorting the eigenvalue according to the descending power.
And 5, setting a threshold value of the accumulated contribution rate to be 90%, determining the number p of the principal elements, selecting eigenvectors corresponding to the previous p eigenvalues to form a matrix, and obtaining the signal characteristics after dimensionality reduction.
And 4, modeling and predicting.
And (3) carrying out normalization processing on the signal characteristics obtained in the step (3), taking the wear ratio (the residual working time is divided by the total working time of the cutter) as a residual service life label, constructing a training set, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
The structure of the long-time and short-time memory network combined with the attention mechanism is as follows in sequence: the number of output neurons of the LSTM network layer is set to be 70, the number of neurons of the full connection layer is set to be 1024, and the number of neurons of the output layer is set to be 1.
Example 2
The data set used in the embodiment of the invention is the data of the Fuji machine tool in the second industrial big data innovation competition, and the data is acquired from the real machining process of the numerical control machine tool, and the acquisition of a brand new tool is stopped from the beginning of normal machining until the service life of the tool is terminated.
1) And (5) signal acquisition and processing.
1.1) collecting signals in the working process of the numerical control machine tool, namely controller signals and sensor signals, wherein the controller signals mainly comprise main shaft loads and mechanical coordinates in three directions, namely x-axis mechanical coordinates, y-axis mechanical coordinates and z-axis mechanical coordinates, and the sensor signals mainly comprise current signals and vibration signals in three directions, namely x-axis vibration signals, y-axis vibration signals, z-axis vibration signals and current signals.
The method comprises the steps of collecting signals of 3 cutters, namely a cutter No. 01, a cutter No. 02 and a cutter No. 03, wherein the sampling frequency of a controller signal is 33Hz, and the sampling frequency of a sensor signal is 25600 Hz.
1.2) preprocessing the acquired signals, firstly removing the signals acquired when the signals do not contact with a processing object according to the mechanical coordinates in three directions and the main shaft load, then processing missing values and abnormal values, and removing abnormal trend items in the signals by using a least square method.
The step of removing the trend term in the signal by using a least square method comprises the following steps: firstly, fitting a signal by using a high-order polynomial, selecting a proper polynomial coefficient according to the principle of a least square method to minimize the sum of squares of errors between the signal and the fitting signal, then substituting the selected polynomial coefficient into the high-order polynomial to obtain a trend term of the signal, and finally subtracting the trend term from the signal to obtain the signal without the trend term.
2) And (5) extracting signal characteristics.
And (2) carrying out feature extraction on the signals obtained in the step 1) from three aspects of time domain, frequency domain and time-frequency domain.
3) And selecting signal characteristics.
3.1) screening the signal characteristics obtained in the step 2) by using two evaluation criteria of monotonicity and tendency.
3.1.1) calculating the monotonicity value and the trend value of each signal characteristic obtained in the step 2) according to the following formula.
Wherein S ismonIs the monotonicity value of the single signal feature, T is the sample length of the single signal feature, and dH represents the difference of each sample in the single feature from the previous sample.
Wherein S istredIs a sheetTrend value of individual signal features, T is sample length of individual signal feature, xiIs the ith sample of a single signal feature, x is the mean of the single signal feature, tiIs xiThe corresponding accumulated working time is calculated according to the working time,
is the average of the accumulated working time series.
3.1.2) taking the mean value of the monotonicity value and the trend value of each signal feature as a comprehensive evaluation value, and forming a comprehensive evaluation matrix by the comprehensive evaluation values of all the signal features.
3.1.3) normalizing the comprehensive evaluation matrix, selecting signal characteristics with the comprehensive evaluation value larger than 0.5 to form a matrix, and obtaining the screened signal characteristics.
And 3.2) performing dimensionality reduction on the signal characteristics obtained by screening by using a kernel principal component analysis algorithm.
3.2.1) the signal features obtained by screening are standardized.
3.2.2) selecting a kernel function and calculating a kernel matrix.
3.2.3) centralizing the nuclear matrix to obtain a centralized nuclear matrix.
3.2.4) calculating the eigenvalue and the eigenvector of the centralized kernel matrix, and sorting the eigenvalue by descending power.
3.2.5) setting the threshold value of the accumulated contribution rate to be 90%, determining the number p of principal elements, selecting eigenvectors corresponding to the previous p eigenvalues to form a matrix, and obtaining the signal characteristics after dimensionality reduction.
4) And modeling and predicting.
Normalizing the signal characteristics obtained in the step 3), constructing a training set by taking a wear ratio (the residual working time is divided by the total working time of the cutter) as a residual service life label, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
The structure of the long-time and short-time memory network combined with the attention mechanism is as follows in sequence: the number of output neurons of the LSTM network layer is set to be 70, the number of neurons of the full connection layer is set to be 1024, and the number of neurons of the output layer is set to be 1.
Different experiments are set to prove the accuracy and the generalization capability of the long-time memory network prediction model combined with the attention mechanism in the residual service life prediction problem of the cutter. The training set of experiment one is cutter No. 01, and the testing set is cutter No. 02 and cutter No. 03; the training set of experiment two is No. 02 cutter, and the testing set is No. 01 cutter and No. 03 cutter; the training set of experiment three is tool 03, the testing set is tool 01 and tool 02, and the prediction results of the experiments are shown in fig. 4.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. The method for predicting the residual service life of the cutter of the numerical control machine tool is characterized by comprising the following steps of: collecting controller signals and sensor signals in the working process of the numerical control machine tool, preprocessing the signals, extracting features and selecting the features, and mining information related to tool abrasion in various signals; and establishing a tool residual service life prediction model by using a long-time memory network and an attention mechanism, so as to realize the prediction of the residual service life of the numerical control machine tool.
2. The method for predicting the remaining service life of the cutter of the numerical control machine tool according to claim 1, wherein the method for predicting the remaining service life of the cutter of the numerical control machine tool comprises the steps of:
step one, signal acquisition and processing;
step two, extracting signal characteristics;
step three, signal characteristic selection;
and step four, modeling and predicting.
3. The method for predicting the remaining service life of the numerical control machine tool cutter according to claim 2, wherein in the first step, the signal acquisition and processing comprises:
(1) collecting signals in the working process of the numerical control machine tool, namely a controller signal and a sensor signal, wherein the controller signal mainly comprises a main shaft load and three-direction mechanical coordinates, namely an x-axis mechanical coordinate, a y-axis mechanical coordinate and a z-axis mechanical coordinate, and the sensor signal mainly comprises a current signal and three-direction vibration signals, namely an x-axis vibration signal, a y-axis vibration signal and a z-axis vibration signal;
(2) preprocessing the acquired signals, firstly removing the signals acquired when the signals do not contact with a processing object according to mechanical coordinates in three directions and a main shaft load, then processing missing values and abnormal values, and finally removing abnormal trend items in the signals by using a least square method;
wherein, the removing the abnormal trend term in the signal by using the least square method comprises the following steps:
firstly, fitting a signal by using a high-order polynomial, selecting a proper polynomial coefficient according to the principle of a least square method to minimize the sum of squares of errors between the signal and the fitting signal, then substituting the selected polynomial coefficient into the high-order polynomial to obtain a trend term of the signal, and finally subtracting the trend term from the signal to obtain the signal without the trend term.
4. The method for predicting the remaining service life of the numerical control machine tool cutter according to claim 2, wherein in the second step, the signal feature extraction comprises the following steps:
and C, extracting the characteristics of the signal obtained in the step I from the aspects of time domain, frequency domain and time-frequency domain.
5. The method for predicting the remaining service life of the numerical control machine tool cutter according to claim 2, wherein in the third step, the signal characteristic selection comprises:
(1) and (3) screening the signal characteristics obtained in the second step by using two evaluation criteria of monotonicity and tendency, wherein the two evaluation criteria comprise:
calculating the monotonicity value and the trend value of each signal feature obtained in the step two according to the following formula:
wherein S ismonIs the monotonicity value of the single signal feature, T is the sample length of the single signal feature, and dH represents the difference value of each sample in the single feature and the previous sample;
wherein S istredIs the trend value of a single signal feature, T is the sample length of the single signal feature, xiFor the ith sample of a single signal feature,
is the mean of the features of the individual signals, tiIs xiThe corresponding accumulated working time is calculated according to the working time,
is the average value of the accumulated working time sequence;
secondly, taking the mean value of the monotonicity value and the trend value of each signal feature as a comprehensive evaluation value, and forming a comprehensive evaluation matrix by the comprehensive evaluation values of all the signal features;
normalizing the comprehensive evaluation matrix, selecting signal characteristics with the comprehensive evaluation value larger than 0.5 to form a matrix, and obtaining screened signal characteristics;
(2) performing dimensionality reduction processing on the signal characteristics obtained by screening by using a kernel principal component analysis algorithm, wherein the dimensionality reduction processing comprises the following steps:
firstly, standardizing the signal characteristics obtained by screening;
selecting a kernel function and calculating a kernel matrix;
thirdly, centralizing the kernel matrix to obtain a centralized kernel matrix;
fourthly, calculating the eigenvalue and the eigenvector of the centralized kernel matrix, and ordering the eigenvalue according to the power reduction;
setting the threshold value of the accumulated contribution rate to be 90%, determining the number p of the principal elements, selecting eigenvectors corresponding to the previous p eigenvalues to form a matrix, and obtaining the signal characteristics after dimensionality reduction.
6. The method for predicting the remaining service life of the numerical control machine tool cutter according to claim 2, wherein in the fourth step, the modeling prediction comprises:
and (4) carrying out normalization processing on the signal characteristics obtained in the step three, taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, constructing a training set, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
7. The method for predicting the remaining service life of the numerical control machine tool cutter according to claim 6, wherein the establishing of the long-time memory network prediction model combined with the attention mechanism comprises the following steps:
the structure of the long-time and short-time memory network combined with the attention mechanism is as follows in sequence: the number of output neurons of the LSTM network layer is set to be 70, the number of neurons of the full connection layer is set to be 1024, and the number of neurons of the output layer is set to be 1.
8. A system for predicting the remaining useful life of a cutter of a numerically controlled machine tool, which performs the method for predicting the remaining useful life of a cutter of a numerically controlled machine tool according to any one of claims 1 to 7, wherein the system for predicting the remaining useful life of a cutter of a numerically controlled machine tool comprises:
the signal acquisition and processing module is used for acquiring signals in the working process of the numerical control machine tool and preprocessing the acquired signals;
the signal characteristic extraction module is used for extracting the characteristics of the acquired signals from the three aspects of time domain, frequency domain and time-frequency domain;
the signal characteristic selection module is used for screening the signal characteristics by utilizing two evaluation standards of monotonicity and tendency, and then performing dimensionality reduction processing on the screened signal characteristics by utilizing a kernel principal component analysis algorithm;
and the modeling prediction module is used for carrying out normalization processing on the signal characteristics, constructing a training set by taking the wear ratio, namely the residual working time divided by the total working time of the cutter as a residual service life label, establishing a long-time memory network prediction model combined with an attention mechanism, and predicting the residual service life of the cutter.
9. A computer device, characterized in that the computer device comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of:
collecting controller signals and sensor signals in the working process of the numerical control machine tool, preprocessing the signals, extracting features and selecting the features, and mining information related to tool abrasion in various signals; and establishing a tool residual service life prediction model by using a long-time memory network and an attention mechanism, so as to realize the prediction of the residual service life of the numerical control machine tool.
10. An information data processing terminal, characterized in that the information data processing terminal is used for implementing the system for predicting the remaining service life of the cutting tool of the numerical control machine tool according to claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110642406.XA CN113560955B (en) | 2021-06-09 | 2021-06-09 | Method, system and application for predicting remaining service life of numerical control machine tool cutter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110642406.XA CN113560955B (en) | 2021-06-09 | 2021-06-09 | Method, system and application for predicting remaining service life of numerical control machine tool cutter |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113560955A true CN113560955A (en) | 2021-10-29 |
| CN113560955B CN113560955B (en) | 2022-09-09 |
Family
ID=78161836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110642406.XA Active CN113560955B (en) | 2021-06-09 | 2021-06-09 | Method, system and application for predicting remaining service life of numerical control machine tool cutter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113560955B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114227381A (en) * | 2021-12-24 | 2022-03-25 | 浙江大学台州研究院 | Machine tool cutter wear detection method |
| CN114417734A (en) * | 2022-03-09 | 2022-04-29 | 深圳市信润富联数字科技有限公司 | Method and device for predicting service life of tool |
| TWI789167B (en) * | 2021-12-16 | 2023-01-01 | 上銀科技股份有限公司 | Method and system for inspecting declining of dynamic response of feeding system |
| CN116681901A (en) * | 2023-07-31 | 2023-09-01 | 山东捷瑞数字科技股份有限公司 | Method for predicting residual life of glass push broach tool bit based on industrial vision |
| CN116825243A (en) * | 2023-05-09 | 2023-09-29 | 安徽工程大学 | Multi-source data-based thermal barrier coating service life prediction method and system |
| CN117464420A (en) * | 2023-12-28 | 2024-01-30 | 江苏新贝斯特智能制造有限公司 | Digital twin control cutter self-adaptive matching system suitable for numerical control machine tool |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110303380A (en) * | 2019-07-05 | 2019-10-08 | 重庆邮电大学 | A kind of cutting tool for CNC machine method for predicting residual useful life |
| CN110355608A (en) * | 2019-07-18 | 2019-10-22 | 浙江大学 | Based on the tool abrasion prediction technique from attention mechanism and deep learning |
| CN110610035A (en) * | 2019-08-28 | 2019-12-24 | 武汉科技大学 | Rolling bearing residual life prediction method based on GRU neural network |
| CN111695521A (en) * | 2020-06-15 | 2020-09-22 | 哈尔滨理工大学 | Attention-LSTM-based rolling bearing performance degradation prediction method |
| CN111687689A (en) * | 2020-06-23 | 2020-09-22 | 重庆大学 | Cutter wear state prediction method and device based on LSTM and CNN |
| CN112034789A (en) * | 2020-08-25 | 2020-12-04 | 国家机床质量监督检验中心 | Health assessment method, system and assessment terminal for key parts and complete machine of numerical control machine tool |
| CN112129534A (en) * | 2020-09-17 | 2020-12-25 | 上海交通大学 | Method and system for predicting service life of main shaft bearing of machine tool |
| WO2021046737A1 (en) * | 2019-09-11 | 2021-03-18 | 大连理工大学 | Method for monitoring tool wear in deep hole machining based on ssae-lstm model |
| CN112784744A (en) * | 2021-01-22 | 2021-05-11 | 北京航空航天大学 | Mechanical component vibration signal preprocessing method with missing value |
-
2021
- 2021-06-09 CN CN202110642406.XA patent/CN113560955B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110303380A (en) * | 2019-07-05 | 2019-10-08 | 重庆邮电大学 | A kind of cutting tool for CNC machine method for predicting residual useful life |
| CN110355608A (en) * | 2019-07-18 | 2019-10-22 | 浙江大学 | Based on the tool abrasion prediction technique from attention mechanism and deep learning |
| CN110610035A (en) * | 2019-08-28 | 2019-12-24 | 武汉科技大学 | Rolling bearing residual life prediction method based on GRU neural network |
| WO2021046737A1 (en) * | 2019-09-11 | 2021-03-18 | 大连理工大学 | Method for monitoring tool wear in deep hole machining based on ssae-lstm model |
| CN111695521A (en) * | 2020-06-15 | 2020-09-22 | 哈尔滨理工大学 | Attention-LSTM-based rolling bearing performance degradation prediction method |
| CN111687689A (en) * | 2020-06-23 | 2020-09-22 | 重庆大学 | Cutter wear state prediction method and device based on LSTM and CNN |
| CN112034789A (en) * | 2020-08-25 | 2020-12-04 | 国家机床质量监督检验中心 | Health assessment method, system and assessment terminal for key parts and complete machine of numerical control machine tool |
| CN112129534A (en) * | 2020-09-17 | 2020-12-25 | 上海交通大学 | Method and system for predicting service life of main shaft bearing of machine tool |
| CN112784744A (en) * | 2021-01-22 | 2021-05-11 | 北京航空航天大学 | Mechanical component vibration signal preprocessing method with missing value |
Non-Patent Citations (3)
| Title |
|---|
| 叶礼伦: "《西安电子科技大学 硕士学位论文》", 11 November 1111 * |
| 赵寄辰: ""基于大数据的盾构机刀具剩余使用寿命预测方法研究"", 《中国优秀硕士学位论文全文数据库(电子期刊) 工程科技II辑》 * |
| 陈柄旭: ""基于深度卷积循环神经网络的刀具状态监测技术研究"", 《中国优秀硕士学位论文全文数据库（电子期刊）工程科技I辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI789167B (en) * | 2021-12-16 | 2023-01-01 | 上銀科技股份有限公司 | Method and system for inspecting declining of dynamic response of feeding system |
| CN114227381A (en) * | 2021-12-24 | 2022-03-25 | 浙江大学台州研究院 | Machine tool cutter wear detection method |
| CN114417734A (en) * | 2022-03-09 | 2022-04-29 | 深圳市信润富联数字科技有限公司 | Method and device for predicting service life of tool |
| CN114417734B (en) * | 2022-03-09 | 2022-07-12 | 深圳市信润富联数字科技有限公司 | Method and device for predicting service life of tool |
| CN116825243A (en) * | 2023-05-09 | 2023-09-29 | 安徽工程大学 | Multi-source data-based thermal barrier coating service life prediction method and system |
| CN116825243B (en) * | 2023-05-09 | 2024-01-16 | 安徽工程大学 | Multi-source data-based thermal barrier coating service life prediction method and system |
| CN116681901A (en) * | 2023-07-31 | 2023-09-01 | 山东捷瑞数字科技股份有限公司 | Method for predicting residual life of glass push broach tool bit based on industrial vision |
| CN116681901B (en) * | 2023-07-31 | 2023-10-31 | 山东捷瑞数字科技股份有限公司 | Method for predicting residual life of glass push broach tool bit based on industrial vision |
| CN117464420A (en) * | 2023-12-28 | 2024-01-30 | 江苏新贝斯特智能制造有限公司 | Digital twin control cutter self-adaptive matching system suitable for numerical control machine tool |
| CN117464420B (en) * | 2023-12-28 | 2024-03-08 | 江苏新贝斯特智能制造有限公司 | Digital twin control cutter self-adaptive matching system suitable for numerical control machine tool |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113560955B (en) | 2022-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113560955B (en) | Method, system and application for predicting remaining service life of numerical control machine tool cutter | |
| CN113569903A (en) | Method, system, equipment, medium and terminal for predicting abrasion of numerical control machine tool cutter | |
| CN103513983A (en) | Method and system for predictive alert threshold determination tool | |
| CN111774931A (en) | On-line monitoring method for abrasion of numerical control turning batch machining tools | |
| US11644812B2 (en) | Machine tool management method, machine tool management system and medium | |
| US20220179393A1 (en) | Machine tool evaluation method, machine tool evaluation system and medium | |
| CN114066196A (en) | Power grid investment strategy optimization system | |
| CN113467375A (en) | Method for predicting remaining usable life of intelligent numerical control machining tool based on virtual measurement | |
| CN112785377A (en) | Data distribution-based order completion period prediction model construction method and prediction method | |
| CN116307215A (en) | Load prediction method, device, equipment and storage medium of power system | |
| Oberle et al. | A use case to implement machine learning for life time prediction of manufacturing tools | |
| CN116383645A (en) | Intelligent system health degree monitoring and evaluating method based on anomaly detection | |
| CN116188745A (en) | Method and device for determining processing technology, electronic equipment and storage medium | |
| CN111614504A (en) | Power grid regulation and control data center service characteristic fault positioning method and system based on time sequence and fault tree analysis | |
| CN116011158A (en) | Topology identification method, system and device for low-voltage transformer area | |
| CN113011325B (en) | Stacker track damage positioning method based on isolated forest algorithm | |
| CN115081551A (en) | RVM line loss model building method and system based on K-Means clustering and optimization | |
| CN115481183A (en) | Real-time data processing method and system | |
| CN114523337A (en) | Cutter wear state identification method and device, electronic equipment and storage medium | |
| CN115509784A (en) | Fault detection method and device for database instance | |
| CN112308338A (en) | Power data processing method and device | |
| CN116467674B (en) | Intelligent fault processing fusion updating system and method for power distribution network | |
| CN117077022B (en) | 3D printer wire feeding mechanism flow monitoring method | |
| CN117574055A (en) | Hydropower unit state monitoring data cleaning method and device and electronic equipment | |
| Di et al. | Tool Fault Diagnosis Method Based on Multiscale-Channel Attention Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |