CN113043073A - Cutter abrasion and service life prediction method and device - Google Patents
Cutter abrasion and service life prediction method and device Download PDFInfo
- Publication number
- CN113043073A CN113043073A CN202110302167.3A CN202110302167A CN113043073A CN 113043073 A CN113043073 A CN 113043073A CN 202110302167 A CN202110302167 A CN 202110302167A CN 113043073 A CN113043073 A CN 113043073A
- Authority
- CN
- China
- Prior art keywords
- cutter
- tool
- wear
- various signals
- signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
- B23Q17/00—Arrangements for observing, indicating or measuring on machine tools
- B23Q17/09—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
- B23Q17/00—Arrangements for observing, indicating or measuring on machine tools
- B23Q17/09—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool
- B23Q17/0952—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool during machining
- B23Q17/0957—Detection of tool breakage
Abstract
The invention provides a cutter wear and service life prediction method, which comprises the steps of acquiring various signals of a cutter in a real-time state, preprocessing the signals, further extracting the characteristics of the preprocessed signals, and acquiring characteristic samples of the signals of the cutter; predicting by using a cutter wear prediction model trained in advance according to the obtained characteristic samples of various signals of the cutter to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy; and combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result. By implementing the method, the real-time wear state and the service life of the cutter can be accurately predicted, and the problem of reduced production benefit caused by cutter damage is solved to a certain extent.
Description
Technical Field
The invention relates to the technical field of computers and cutters, in particular to a cutter abrasion and service life prediction method and device.
Background
In the manufacturing industry nowadays, the technologies such as artificial intelligence and big data processing, analyzing and mining are rapidly developed, and the health status monitoring of mechanical equipment is gradually developed from the after-the-fact fault diagnosis of processing equipment to early warning maintenance and life prediction management of the fault of the processing equipment.
At present, the real-time working state of a cutter of a numerical control machine tool also needs to realize intelligent early warning maintenance and life prediction management, and the working state mainly in cutter production plays a decisive role in the quality of a workpiece. Once the tool is damaged, the workpiece may be damaged, and even a machine tool accident may be caused, thereby affecting the production efficiency of the product. Therefore, the monitoring of the real-time working state of the cutter can improve the stability of product production and ensure the production efficiency of the product.
Conventional monitoring methods include direct monitoring (optical spectroscopy, contact methods, electrical discharge techniques) and indirect monitoring (cutting temperature measurement, acoustic emission detection, vibration monitoring, cutting force monitoring). Although the direct monitoring has strong operability and can obtain visual data, the data has errors and insufficient precision due to environmental interference. Although indirect monitoring can collect relatively accurate data, real-time monitoring is not enough, and the wear degree and the service life condition of the tool in a future working period are required to be intelligently predicted.
Therefore, it is necessary to use the technologies of artificial intelligence and big data processing, analyzing, mining, etc. to predict the real-time wear state and service life of the tool accurately through the real-time signal generated during the tool production, and to change and remove the fault of the tool in real time, thereby reducing the production loss to a greater extent and realizing the production automation and intellectualization.
Disclosure of Invention
The technical problem to be solved by the embodiments of the present invention is to provide a method and an apparatus for predicting wear and life of a tool, which can accurately predict the real-time wear state and life of the tool, and reduce the problem of reduced production efficiency caused by tool damage to a certain extent.
In order to solve the above technical problem, an embodiment of the present invention provides a method for predicting wear and life of a tool, where the method includes:
acquiring various signals of a cutter in a real-time state, preprocessing the various signals, and further extracting the characteristics of the preprocessed various signals to obtain characteristic samples of the various signals of the cutter;
predicting by using a cutter wear prediction model trained in advance according to the obtained characteristic samples of various signals of the cutter to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result.
The preprocessing steps of various signals comprise removal of invalid values, processing of abnormal values and noise reduction operation.
The characteristic samples of various signals are composed of time domain characteristics, frequency domain characteristics and a mixture of the time domain characteristics and the frequency domain characteristics, wherein the time domain characteristics and the frequency domain characteristics are extracted through an embedding method.
The various signals comprise a cutter cutting speed signal, a cutter back draft signal, a cutter cutting width signal, a cutter diameter signal, a cutter per-tooth feeding amount signal and a cutter tooth number signal.
The embodiment of the invention also provides a device for predicting the abrasion and the service life of the cutter, which comprises:
the sample acquisition unit is used for acquiring various signals of the real-time state of the cutter, preprocessing the various signals and further extracting the characteristics of the preprocessed various signals to obtain characteristic samples of the various signals of the cutter;
the cutter wear prediction unit is used for predicting by using a cutter wear prediction model trained in advance according to the obtained characteristic samples of various signals of the cutter to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and the cutter life prediction unit is used for combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result.
The characteristic samples of various signals are composed of time domain characteristics, frequency domain characteristics and a mixture of the time domain characteristics and the frequency domain characteristics, wherein the time domain characteristics and the frequency domain characteristics are extracted through an embedding method.
The various signals comprise a cutter cutting speed signal, a cutter back draft signal, a cutter cutting width signal, a cutter diameter signal, a cutter per-tooth feeding amount signal and a cutter tooth number signal.
The embodiment of the invention has the following beneficial effects:
the invention can accurately predict the real-time wear state and service life of the cutter based on various signals of the real-time state of the cutter, and reduces the problem of reduced production benefit caused by cutter damage to a certain extent.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is within the scope of the present invention for those skilled in the art to obtain other drawings based on the drawings without inventive exercise.
FIG. 1 is a flow chart of a tool wear and life prediction method according to an embodiment of the present invention;
fig. 2 is a BP neural network structure diagram adopted by a tool life prediction model in an application scenario of the tool wear and life prediction method provided in the embodiment of the present invention;
fig. 3 is a schematic structural diagram of a device for predicting tool wear and life according to an 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 will be described in further detail with reference to the accompanying drawings.
As shown in fig. 1, a method for predicting wear and life of a tool according to an embodiment of the present invention includes the following steps:
s1, acquiring various signals of the real-time state of the cutter, preprocessing the various signals, and further extracting the characteristics of the preprocessed various signals to obtain characteristic samples of the various signals of the cutter;
step S2, according to the obtained characteristic samples of various signals of the cutter, predicting by using a cutter wear prediction model trained in advance to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and step S3, combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result.
In step S1, various signals of the tool, including but not limited to a tool cutting speed signal, a tool back draft signal, a tool cutting width signal, a tool diameter signal, a tool per tooth feed signal, a tool tooth number signal, etc., acquired by each sensor in the tool holder in real time are received. And then, preprocessing is carried out on each type of signal, including removing invalid values, processing abnormal values, carrying out noise reduction operation and the like. And then, extracting characteristics on the basis of wavelet denoising, wherein the extraction comprises the extraction of time domain characteristics, frequency domain characteristics and time domain characteristics, so that characteristic samples of various signals are formed by mixing the time domain characteristics, the frequency domain characteristics and the characteristics extracted by the embedding method.
It should be noted that the time domain characteristics: and eliminating the influence of the signal amplitude by the dimensional characteristic parameters and the dimensionless parameters to reflect the tool wear information. Frequency domain characteristics: and the time domain signal reflects the key information of the wear condition of the cutter through key parameters in the frequency domain information extracted after Fourier transform. Time-frequency domain characteristics: feature extraction is performed by an embedding method. The embedding method obtains importance weights of all the features through model training results, and selects the optimal features according to the weight sorting.
In step S2, a tool wear prediction model is first constructed, which is formed by fusing a random forest prediction model and a convolutional neural network prediction model by using a stacking fusion strategy, and a layer of learners is added after integrating weak learners according to a certain method instead of simply performing logic processing on the weak learners. And training a random forest prediction model and a convolutional neural network prediction model by adopting a k-fold cross validation method, and reducing the dimension of the output of the base classifier by a principal component analysis method so as to obtain the prediction result output of the tool wear state.
It should be noted that the random forest-based model algorithm roughly comprises the following steps: firstly, bootstrap sampling is carried out on a training set output by a signal preprocessing and feature extraction module, and the sampling is random sampling with a return. We then train each decision tree with the training set extracted by the sampling method. And then selecting n optimal features from the feature set extracted by the signal preprocessing and feature extraction module. And repeating the steps for k times to obtain a random forest consisting of k decision trees. In the invention, a CART algorithm is adopted to process a regression problem, and finally the average value of k decision trees is used as a final result to be output. The output of the invention is the predicted value of the tool wear. And reasonably selecting and setting the number k of decision trees, the depth d of the decision trees and the number n of selected features in the feature set extracted by the signal preprocessing and feature extraction module according to the actual feature condition of a specific cutter during training.
It should be noted that the convolutional neural network is a deep learning model, and is essentially a multilayer perceptron, and the convolutional neural network adopts a local connection and weight sharing mode, so that on one hand, the number of weights is reduced, so that the network is easier to train and optimize compared with a fully-connected network, on the other hand, the complexity of the model is reduced, and the risk of network overfitting is reduced. The algorithm adopting the convolutional neural network base model roughly comprises the following steps: firstly, the convolution layer carries out secondary feature extraction on training set samples output by a signal preprocessing and feature extraction module, and enhances and extracts key features of a cutter signal through convolution operation, so that noise is reduced. However, these operations are already greatly optimized in the signal preprocessing and feature extraction modules, so this improves the performance of the algorithm to some extent for the multiple convolution processes of the convolutional layer. Wherein the content of the first and second substances,
the pooling layer performs down-sampling on the features in the matrix, compresses the tool information features and reserves useful features as far as possible, and the step also performs related preprocessing and feature extraction in a signal preprocessing and feature extraction module, so that the model operation speed is accelerated to a certain extent, and the method has a certain promotion effect on controlling overfitting and enhancing the robustness of the convolutional neural network.
And fully connecting the neurons of the full connection layer with the activated neurons, and performing weighting operation on the tool wear characteristics in the tool wear prediction model according to the properties of the full connection layer. The Dropout layer deletes part of units of the hidden layer, so that the overfitting degree is reduced, and finally the normalization layer solves the convergence problem in model training. And after convolution for many times, outputting the tool wear prediction result by the full connection layer.
And then, training the tool wear prediction model by adopting an off-line sample to obtain the trained tool wear prediction model. The offline sample is a feature sample formed by preprocessing and feature extracting historical data of various signals of the tool, and the preprocessing and feature extracting are the same as those in step S1, and are not described herein again. On one hand, the off-line samples form a training set and a testing set which are used as the input of a tool wear prediction off-line training module, and a tool wear prediction model is trained, so that the prediction result of tool wear is output; on the other hand, the off-line sample and the tool wear prediction result output by the tool wear off-line training module are used as the input of the tool life prediction off-line training module together, and the tool life prediction model is trained, so that a tool life prediction result is generated.
And finally, predicting the characteristic samples of various signals of the cutter, namely the real-time online samples, obtained in the step S1 by using the trained cutter wear prediction model to obtain a cutter wear prediction result.
In step S3, the tool life, i.e., the time from the time the tool is put into use to the time it is scrapped, is determined by various factors affecting the use of the tool. The service life of the cutter is generally considered to be calculated by the formula
Wherein, CrIs a tool life factor, D0V is the diameter of the toolCAs cutting speed, aPFor the amount of cutting, awF is the cutting width, f is the feed amount, z is the cutter tooth number, wherein x, y, u, p and q are index values of all corresponding parameters, and the influence degree of all corresponding parameters on the cutter service life is represented by the size table.
Therefore, as can be seen from the tool life calculation formula, the main parameters affecting the tool life include cutting speed, back-cut amount, cutting width, feed per tooth, number of teeth of the tool, and tool diameter, and show a corresponding functional relationship with the tool life.
The traditional tool life prediction algorithm is difficult to accurately reflect the nonlinear relation between the tool life influence factor and the service life of the tool, but considering that the BP neural network algorithm has the advantage of processing the nonlinear system problem, the invention selectively utilizes the BP neural network algorithm to construct a tool life prediction model.
Because the BP neural network is a three-layer feedforward neural network comprising an input layer, a hidden layer and an output layer, the BP neural network can approach any continuous function with any precision, and the good error feedback learning capability of the BP neural network provides a theoretical basis for the tool life prediction. According to the life calculation formula of the cutter, the life influence factors of the cutter are numerous, and the machining condition of the cutter is constantly changed.
In one embodiment, as shown in fig. 2, when a tool life prediction model is established, according to the specific tool usage to be monitored and life predicted in actual working production, 6 influence indexes, namely cutting speed, back cut amount, cutting width, tool diameter, feed per tooth and tool tooth number, are selected as input layer neurons of the tool life prediction model of the present invention. And selecting an output layer neuron of which the service life of the cutter is the cutter service life prediction model. The determination of the number of nodes of the hidden layer of the network depends on an empirical formula
Where n is the number of nodes of the input layer, m is the number of nodes of the output layer, and a is a constant. Considering the learning duration problem of the BP neural network and the reason that the number of the nodes is too small and the fault tolerance is poor, the number of the hidden layer nodes is selected to be 8.
Next, the flank wear amount measured on the flank face at half the cutting depth was defined as a standard for tool dull standard tool wear according to the international standard ISO 3685-1977. Selecting a corresponding wear standard according to a specific prop type involved in actual work production, then calculating a residual life value corresponding to each sample in a historical data set, then using a new tool life training set and a new test set which are composed of characteristic values and residual life values, and then training a tool life prediction model by using the tool life training set and the test set, thereby outputting a tool life prediction result. Through a large amount of training of the service life prediction model in the tool wear state offline training module, after performance evaluation, the model can predict a corresponding service life result for new tool information data.
And finally, combining the characteristic samples of various types of signals of the cutter obtained in the step S1 with the cutter wear prediction result obtained in the step S2, and predicting by using the trained cutter life prediction model to obtain a cutter life prediction result.
It can be understood that the tool wear prediction result obtained in the step S2 may be compared with a corresponding threshold value to give an alarm; or comparing the tool life prediction result obtained in the step S2 with a corresponding threshold value, and giving an alarm.
As shown in fig. 3, an embodiment of the present invention provides a device for predicting tool wear and life, including:
the sample acquisition unit 110 is configured to acquire various signals of the real-time state of the tool, preprocess the various signals, and further perform feature extraction on the preprocessed various signals to obtain feature samples of the various signals of the tool;
the cutter wear prediction unit 120 is configured to predict, according to the obtained feature samples of various types of signals of the cutter, a cutter wear prediction model trained in advance, so as to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and the tool life prediction unit 130 is configured to combine the obtained feature samples of the various types of signals of the tool with the obtained tool wear prediction result, and predict the tool wear using a pre-trained tool life prediction model to obtain a tool life prediction result.
The characteristic samples of various signals are composed of time domain characteristics, frequency domain characteristics and a mixture of the time domain characteristics and the frequency domain characteristics, wherein the time domain characteristics and the frequency domain characteristics are extracted through an embedding method.
The various signals comprise a cutter cutting speed signal, a cutter back draft signal, a cutter cutting width signal, a cutter diameter signal, a cutter per-tooth feeding amount signal and a cutter tooth number signal.
The embodiment of the invention has the following beneficial effects:
the invention can accurately predict the real-time wear state and service life of the cutter based on various signals of the real-time state of the cutter, and reduces the problem of reduced production benefit caused by cutter damage to a certain extent.
It should be noted that, in the above device embodiment, each included unit is only divided according to functional logic, but is not limited to the above division as long as the corresponding function can be achieved; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It will be understood by those skilled in the art that all or part of the steps in the method for implementing the above embodiments may be implemented by relevant hardware instructed by a program, and the program may be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc.
The above disclosure is only for the purpose of illustrating the preferred embodiments of the present invention, and it is therefore to be understood that the invention is not limited by the scope of the appended claims.
Claims (7)
1. A method for predicting wear and life of a tool, said method comprising the steps of:
acquiring various signals of a cutter in a real-time state, preprocessing the various signals, and further extracting the characteristics of the preprocessed various signals to obtain characteristic samples of the various signals of the cutter;
predicting by using a cutter wear prediction model trained in advance according to the obtained characteristic samples of various signals of the cutter to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result.
2. The tool wear and life prediction method of claim 1 wherein each type of signal preprocessing step includes removal of invalid values, processing of outliers, and noise reduction.
3. The tool wear and life prediction method of claim 2, wherein the feature samples of each type of signal are composed of time domain features, frequency domain features, and a mixture of both features extracted by the embedding method.
4. The tool wear and life prediction method of claim 3 wherein the types of signals include a tool cutting speed signal, a tool back bite signal, a tool cutting width signal, a tool diameter signal, a tool feed per tooth signal, and a tool tooth count signal.
5. A tool wear and life prediction device, comprising:
the sample acquisition unit is used for acquiring various signals of the real-time state of the cutter, preprocessing the various signals and further extracting the characteristics of the preprocessed various signals to obtain characteristic samples of the various signals of the cutter;
the cutter wear prediction unit is used for predicting by using a cutter wear prediction model trained in advance according to the obtained characteristic samples of various signals of the cutter to obtain a cutter wear prediction result; the tool wear prediction model is formed by fusing a random forest prediction model and a convolutional neural network prediction model by adopting a stacking fusion strategy;
and the cutter life prediction unit is used for combining the obtained characteristic samples of various signals of the cutter with the obtained cutter wear prediction result, and predicting by using a cutter life prediction model trained in advance to obtain a cutter life prediction result.
6. The tool wear and life prediction device of claim 5, wherein the feature samples of each type of signal are composed of time domain features, frequency domain features, and a mixture of features extracted from both by an embedding method.
7. The apparatus for predicting tool wear and life according to claim 6, wherein the signals of the respective types include a tool cutting speed signal, a tool back bite signal, a tool cutting width signal, a tool diameter signal, a tool feed per tooth signal, and a tool tooth number signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110302167.3A CN113043073A (en) | 2021-03-22 | 2021-03-22 | Cutter abrasion and service life prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110302167.3A CN113043073A (en) | 2021-03-22 | 2021-03-22 | Cutter abrasion and service life prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113043073A true CN113043073A (en) | 2021-06-29 |
Family
ID=76514139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110302167.3A Pending CN113043073A (en) | 2021-03-22 | 2021-03-22 | Cutter abrasion and service life prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113043073A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113458873A (en) * | 2021-07-01 | 2021-10-01 | 太原科技大学 | Method for predicting wear loss and residual life of cutter |
| CN114749996A (en) * | 2022-05-25 | 2022-07-15 | 哈尔滨工业大学 | Tool residual life prediction method based on deep learning and time sequence regression model |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000107986A (en) * | 1998-09-29 | 2000-04-18 | Agency Of Ind Science & Technol | Automatic detecting method for cutting tool life |
| CN109015111A (en) * | 2018-07-06 | 2018-12-18 | 华中科技大学 | A kind of cutting tool state on-line monitoring method based on information fusion and support vector machines |
| CN109514349A (en) * | 2018-11-12 | 2019-03-26 | 西安交通大学 | Monitoring Tool Wear States in Turning based on vibration signal and Stacking integrated model |
| CN110153801A (en) * | 2019-07-04 | 2019-08-23 | 西南交通大学 | A kind of cutting-tool wear state discrimination method based on multi-feature fusion |
| CN110561193A (en) * | 2019-09-18 | 2019-12-13 | 杭州友机技术有限公司 | Cutter wear assessment and monitoring method and system based on feature fusion |
| CN110900307A (en) * | 2019-11-22 | 2020-03-24 | 北京航空航天大学 | Numerical control machine tool cutter monitoring system driven by digital twin |
| CN111085898A (en) * | 2019-12-30 | 2020-05-01 | 南京航空航天大学 | Working condition self-adaptive high-speed milling process cutter monitoring method and system |
| CN111890124A (en) * | 2019-05-05 | 2020-11-06 | 深圳市玄羽科技有限公司 | On-line cutter monitoring system and method |
-
2021
- 2021-03-22 CN CN202110302167.3A patent/CN113043073A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000107986A (en) * | 1998-09-29 | 2000-04-18 | Agency Of Ind Science & Technol | Automatic detecting method for cutting tool life |
| CN109015111A (en) * | 2018-07-06 | 2018-12-18 | 华中科技大学 | A kind of cutting tool state on-line monitoring method based on information fusion and support vector machines |
| CN109514349A (en) * | 2018-11-12 | 2019-03-26 | 西安交通大学 | Monitoring Tool Wear States in Turning based on vibration signal and Stacking integrated model |
| CN111890124A (en) * | 2019-05-05 | 2020-11-06 | 深圳市玄羽科技有限公司 | On-line cutter monitoring system and method |
| CN110153801A (en) * | 2019-07-04 | 2019-08-23 | 西南交通大学 | A kind of cutting-tool wear state discrimination method based on multi-feature fusion |
| CN110561193A (en) * | 2019-09-18 | 2019-12-13 | 杭州友机技术有限公司 | Cutter wear assessment and monitoring method and system based on feature fusion |
| CN110900307A (en) * | 2019-11-22 | 2020-03-24 | 北京航空航天大学 | Numerical control machine tool cutter monitoring system driven by digital twin |
| CN111085898A (en) * | 2019-12-30 | 2020-05-01 | 南京航空航天大学 | Working condition self-adaptive high-speed milling process cutter monitoring method and system |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113458873A (en) * | 2021-07-01 | 2021-10-01 | 太原科技大学 | Method for predicting wear loss and residual life of cutter |
| CN113458873B (en) * | 2021-07-01 | 2022-03-11 | 太原科技大学 | Method for predicting wear loss and residual life of cutter |
| CN114749996A (en) * | 2022-05-25 | 2022-07-15 | 哈尔滨工业大学 | Tool residual life prediction method based on deep learning and time sequence regression model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Madhusudana et al. | Condition monitoring of face milling tool using K-star algorithm and histogram features of vibration signal | |
| CN113664612A (en) | Numerical control machine tool milling cutter abrasion real-time monitoring method based on deep convolutional neural network | |
| CN106181579B (en) | A kind of Tool Wear Monitoring method based on multisensor current signal | |
| CN109396953B (en) | Machine tool working state intelligent identification system based on signal fusion | |
| CN111716150B (en) | Evolution learning method for intelligently monitoring cutter state | |
| Madhusudana et al. | Fault diagnosis of face milling tool using decision tree and sound signal | |
| CN113255848B (en) | Water turbine cavitation sound signal identification method based on big data learning | |
| CN113305645B (en) | Numerical control machine tool cutter residual life prediction method based on hybrid neural model | |
| CN113043073A (en) | Cutter abrasion and service life prediction method and device | |
| CN113798920B (en) | Cutter wear state monitoring method based on variational automatic encoder and extreme learning machine | |
| CN106112697A (en) | A kind of milling parameter automatic alarm threshold setting method based on 3 σ criterions | |
| CN112692646B (en) | Intelligent assessment method and device for tool wear state | |
| CN111136509A (en) | Cutter state monitoring method based on ELM-SDAE algorithm | |
| CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| CN114273977A (en) | MES-based cutter wear detection method and system | |
| CN111390648B (en) | Turning tool abrasion judging method based on antagonistic neural network | |
| CN111126255A (en) | Numerical control machine tool cutter wear value prediction method based on deep learning regression algorithm | |
| CN112434613A (en) | Cutter state monitoring method and device, equipment and storage medium | |
| Martinez Arellano et al. | Towards an active learning approach to tool condition monitoring with bayesian deep learning | |
| CN116070527A (en) | Milling cutter residual life prediction method based on degradation model | |
| CN114749996A (en) | Tool residual life prediction method based on deep learning and time sequence regression model | |
| CN112192318A (en) | Machining tool state monitoring method and system | |
| CN114871850A (en) | Cutter wear state evaluation method based on vibration signal and BP neural network | |
| Massol et al. | An exTS based neuro-fuzzy algorithm for prognostics and tool condition monitoring | |
| Dey et al. | Performance improvement techniques for neural networks in tool condition monitoring |
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 |