CN115741235A - Wear prediction and health management method based on five-axis machining center cutter - Google Patents
Wear prediction and health management method based on five-axis machining center cutter Download PDFInfo
- Publication number
- CN115741235A CN115741235A CN202211502529.4A CN202211502529A CN115741235A CN 115741235 A CN115741235 A CN 115741235A CN 202211502529 A CN202211502529 A CN 202211502529A CN 115741235 A CN115741235 A CN 115741235A
- Authority
- CN
- China
- Prior art keywords
- cutter
- health management
- tool
- machining center
- axis machining
- 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.)
- Withdrawn
Links
- 230000036541 health Effects 0.000 title claims abstract description 25
- 238000007726 management method Methods 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000004458 analytical method Methods 0.000 claims abstract description 20
- 238000005520 cutting process Methods 0.000 claims abstract description 19
- 238000003754 machining Methods 0.000 claims abstract description 18
- 238000003801 milling Methods 0.000 claims abstract description 9
- 238000012544 monitoring process Methods 0.000 claims abstract description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 11
- 230000001133 acceleration Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000011176 pooling Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 230000005484 gravity Effects 0.000 claims description 3
- 230000000737 periodic effect Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000005299 abrasion Methods 0.000 abstract description 14
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000004519 manufacturing process Methods 0.000 abstract description 7
- 230000004927 fusion Effects 0.000 abstract description 3
- 238000012423 maintenance Methods 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 238000012098 association analyses Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Numerical Control (AREA)
Abstract
The invention belongs to the technical field of monitoring of numerical control machine tools, and discloses a wear prediction and health management method based on a five-axis machining center tool, which comprises the following steps: the method comprises the steps that a sensor is used for collecting cutting vibration signals and spindle current signals in the milling process of a five-axis machining center in real time, relevant data of the abrasion condition of a cutter are indirectly monitored, and a basis is provided for achieving health management, mode recognition and service life prediction of the cutter; the method comprises the steps of collecting cutting vibration signals and spindle current signals in real time, extracting data characteristics reflecting tool abrasion through time domain and frequency domain analysis and wavelet packet analysis, and constructing a tool abrasion and residual life prediction model by adopting a multi-source information fusion technology, so that predictive maintenance of the numerical control machining tool is realized, intelligent tool changing can be carried out before the tool abrasion is in a critical threshold value, and the method has important significance for actual production.
Description
Technical Field
The invention belongs to the technical field of monitoring of numerical control machine tools, and particularly relates to a wear prediction and health management method based on a five-axis machining center tool.
Background
The five-axis numerical control machining center is high-precision end equipment integrating high technology, high precision and high efficiency, is specially used for machining complex curved surface parts, and the promotion of the key technology has great significance for improving the equipment manufacturing level. In a five-axis numerical control machining center, due to the complexity of machining objects, the degree of tool wear is more serious, according to investigation, more than 90% of machining is realized by cutting machining, and 20% of machine tool downtime is caused by tool wear.
When the abrasion of the cutter exceeds a given threshold value, the machining precision of a workpiece is greatly influenced, so that the quality of a machined product does not reach the standard, the machining input time is wasted, the economic loss is caused, and even the occurrence of machine tool accidents can be caused; on the other hand, in order to ensure the machining precision, if the cutter is replaced when the cutter has a longer residual service life, the use economy of the cutter is affected, the production cost of an enterprise is improved, particularly, the production beat interruption and the production efficiency reduction can be caused in the batch machining process, and therefore how to wear the cutter to perform intelligent cutter changing before the cutter is in a critical threshold value is a problem to be solved urgently in future high-end manufacturing industry.
Disclosure of Invention
The invention aims to provide a wear prediction and health management method based on a five-axis machining center cutter, so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: the wear prediction and health management method based on the five-axis machining center cutter comprises the following steps:
s1: the method comprises the steps that a sensor is used for collecting cutting vibration signals and spindle current signals in the milling process of a five-axis machining center in real time, relevant data of the abrasion condition of a cutter are indirectly monitored, and a basis is provided for achieving health management, mode recognition and service life prediction of the cutter;
s2: preprocessing the various collected original signals, eliminating noise influence, and further extracting characteristic quantities related to the wear state of the cutter in a time domain and a frequency domain;
s3: meanwhile, wavelet packet analysis is carried out, time-frequency domain information of the signals is locally analyzed through multi-scale and different resolutions, and local time-frequency characteristics required by tool wear state identification are researched, so that the focusing on any details of the signals is realized;
s4: extracting all characteristics related to the wear state of the cutter, and then performing further data processing to form a sample set for model training;
s5: the tool wear state is detected based on the convolutional neural network by adopting an efficient deep learning algorithm considering model training time and precision, a classifier is constructed on an output layer after a series of operations such as convolutional layer and pooling layer of the convolutional neural network, tool state information is output, and therefore health management of the tool wear state of the numerical control machining of the complex curved surface part is achieved.
Preferably, the vibration signal is generated by periodic vibration of a cutting system formed by a workpiece or a tool of the machine tool.
Preferably, the sensor for collecting vibration signals includes, but is not limited to, a piezoresistive acceleration sensor, a piezoelectric acceleration sensor, and a capacitive acceleration sensor.
Preferably, the spindle current signal is a working current generated by the spindle during the machining of the part.
Preferably, the sensor for acquiring the spindle current signal is a current sensor.
Preferably, the time domain features are mainly classified into dimensional features and dimensionless features.
Preferably, the dimensional features include absolute mean, variance, significance, peak-to-peak.
Preferably, the dimensionless feature includes a skewness index, a kurtosis index, a peak factor, a coefficient of variation, and a form factor.
Preferably, the frequency domain features include frequency mean square, frequency center of gravity, frequency variance, and peak frequency.
Preferably, the calculation formula of the time-frequency domain features is as follows:
wherein En represents the total energy of the original signal, j represents the number of layers of wavelet packet decomposition, and x k,m (i) Is represented in a subspaceOf (2) a signalThe decomposed signal of (a).
Compared with the prior art, the invention has the beneficial effects that:
the method comprises the steps of firstly collecting cutting vibration signals and spindle current signals in real time, then extracting data characteristics reflecting tool abrasion through time domain and frequency domain analysis and wavelet packet analysis, further establishing a tool abrasion state and characteristics such as the cutting vibration signals and the spindle current signals based on a state recognition technology to establish a physical association analysis model, identifying the tool state based on a convolutional neural network, establishing a tool abrasion and residual life prediction model by adopting a multi-source information fusion technology, taking the measured tool flank abrasion as a training label, and performing model training, wherein the obtained model has the capability of predicting the tool abrasion value with high precision and the residual life prediction capability, so that the predictive maintenance of the numerical control machining tool is realized, the tool can be intelligently changed before the tool abrasion is in a critical threshold value, and the method has important significance for actual production.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram illustrating a method for extracting original signal features according to the present invention;
FIG. 3 is a schematic view of the flank wear of the inventive tool;
FIG. 4 is an exemplary graph of a wear curve for a cubic polynomial interpolation tool of the present invention;
FIG. 5 is a diagram of a CNN convolutional neural network structure of the present invention;
FIG. 6 is a diagram of the tool state recognition convolutional neural network architecture of the present invention;
FIG. 7 is a graph of accuracy and loss function of the present invention;
FIG. 8 is a comparison of test set prediction results for the present invention;
FIG. 9 is a schematic diagram of a confusion matrix according to the present invention;
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides the following technical scheme:
referring to fig. 1 to 9, a wear prediction and health management method based on a five-axis machining center tool includes the following steps:
s1: the method comprises the steps that a sensor is used for collecting cutting vibration signals and spindle current signals in the milling process of a five-axis machining center in real time, relevant data of the abrasion condition of a cutter are indirectly monitored, and a basis is provided for achieving health management, mode recognition and service life prediction of the cutter;
the vibration signal is caused by the periodic vibration of a cutting system formed by a machine tool workpiece or a cutter, and the strength of the vibration among the systems is closely related to the wear state of the cutter. The vibration signal is collected by adopting an acceleration sensor, and the method is roughly divided into three modes according to different measurement principles: the invention adopts a piezoelectric acceleration sensor to collect vibration signals, and when the sensor is installed, the sensor can be adsorbed on the surface of a part to be processed by magnetic force for detection, but the result of measuring the vibration signals of a cutter is influenced by the installation position, and the vibration strength of a machine tool system and the interference of external environmental factors can influence the collection of the vibration signals, and the factors belong to noise;
the spindle current in the milling process refers to the working current generated by the spindle in the process of machining parts. The relevant data show that the more serious the tool wear, the larger the machine tool spindle current is, and the linear proportional relation is almost formed, so that the machine tool spindle current signal can indirectly reflect the tool wear state. The invention uses the current sensor to collect the current signal of the main shaft, and the sensor has the characteristics of simple installation and no limitation of processing environment, so the application range is relatively wide. Meanwhile, the acquisition of a machine tool spindle current signal is very convenient and fast, and the current signal can be directly acquired from the interior of a machine tool, but the spindle motor interferes the acquired data at the moment of starting and braking, and the factors belong to noise;
s2: preprocessing the various collected original signals, eliminating noise influence, and further extracting characteristic quantities related to the wear state of the cutter in a time domain and a frequency domain;
the time domain characteristic of the signal is to expand infinitely for a certain time period of the milling process, and discover and analyze the law of the change of the relevant variable along with the time. Although the acquired signal has a continuously changing waveform, the sampling frequency is high, the noise is limited by frequent interference, and the characteristic related to the tool wear is difficult to directly extract from the original signal, so that time domain analysis is required. The time domain analysis is to perform processing such as related parameter calculation and data analysis on the original signal, so that the extracted time domain features are more representative. The time domain characteristics of the original signal extracted for realizing intelligent cutter wear prediction and health management are mainly divided into dimensional characteristics and dimensionless characteristics, the dimensional time domain characteristics can directly reflect various changes of the milling cutter machining process, and the time domain characteristics mainly comprise 5 time domain characteristics which are respectively an absolute average value, a variance, an effective value, a peak value and a peak-to-peak value; the dimensionless parameters are obtained through the division of the same dimension, so that the interference of factors such as signal amplitude and the like can be avoided, and other information of tool wear can be reflected. The dimensionless features mainly include 5 time domain features, which are skewness index, kurtosis index, peak factor, variation coefficient and wave form factor respectively;
the frequency domain characteristics of the signal describe the law of observing the related variables of the signal on the aspect of frequency, and are more profound and convenient than time domain analysis. Fourier transform (fourier transform) is the most commonly used method for frequency domain analysis, and its essence is to convert the time domain signal into the frequency domain, and perform tool life prediction by extracting the spectral features of the sample signal. When the abrasion degree of the cutter is changed in the milling process, the frequency components in the signal frequency spectrum are changed, so that the signal frequency spectrum information can be accurately represented by analyzing the frequency domain characteristics, and whether the cutter is in a healthy state or not can be known. The frequency domain characteristics mainly comprise 4 frequency domain characteristics of frequency mean square, frequency gravity center, frequency variance and peak frequency;
s3: meanwhile, wavelet packet analysis is carried out, time-frequency domain information of the signals is locally analyzed through multi-scale and different resolutions, and local time-frequency characteristics required by tool wear state identification are researched, so that the focusing on any details of the signals is realized;
due to the change of the geometric characteristics or the process parameters of the machined part, in the process of monitoring the tool wear signal, the signal acquired by the sensor is subjected to instantaneous sudden change, so that the signal on a time-frequency domain needs to be analyzed. Statistical short-time fourier analysis and wavelet analysis are the most common methods for time-frequency domain analysis. The project is further researched on the basis of wavelet analysis, and time-frequency domain characteristics are extracted by utilizing wavelet packet analysis. Wavelet analysis is to decompose the primitive signal gathered layer by layer, only consider the low frequency part of the time domain signal generally while decomposing in each layer, the signal of the high frequency part can be ignored, and the wavelet packet analysis that the invention utilizes needs to consider high frequency signal and low frequency signal separately in decomposing the course layer by layer, after the high low frequency signal decomposes, make low frequency and high frequency part have the same resolution, the signal is subdivided into different frequency bands, the frequency band structure of the monitoring signal will change too along with the change of the wear state of the cutting tool, cause the energy parameter of different frequency bands to change, therefore the energy size of each frequency band can accurately represent the wear degree of the cutting tool.
Wherein En represents the total energy of the original signal, j represents the number of layers of wavelet packet decomposition, and x k,m (i) Is represented in a subspaceOf (2) a signalThe decomposed signal of (a).
The time domain signal is decomposed into wavelet packet according to the principle, the decomposed layers are 3 layers and are all completed by db5 wavelet basis, and then the decomposed signals of each layer are reconstructed according to wavelet coefficients so as to carry out more accurate analysis. Because the wavelet packet basis has orthogonality, the energy of the frequency band can be represented by the wavelet packet coefficient of each frequency band, and through 3-layer decomposition, the frequency domain is divided into 8 frequency bands, so that 8 time-frequency domain characteristics are extracted.
S4: extracting all characteristics related to the wear state of the cutter, and then performing further data processing to form a sample set for model training;
the invention respectively extracts the time domain, the frequency domain and the time-frequency domain characteristics of the original data, wherein the characteristic extraction is carried out once every delta t time, namely, the cutting vibration original signal and the main shaft current original signal are respectively extracted in every delta t time, 10 time domain characteristics, 4 frequency domain characteristics and 8 time-frequency domain characteristics are extracted through the analysis, and 22 characteristic values are summed, so that a sample characteristic is formed, namely: in the time from t to t + Δ t, let a = { A1, A2, …, AN } and B = { B1, B2, …, BN } for the cutting vibration raw data set, and assume the above 22 features calculation formula is Fi, where i =1,2, …,22; extracting the feature of the cutting vibration as Xi = Fi (a); the spindle current is characterized by Yi = Fi (B), where i =1,2, …,22; repeating the above steps for the next Δ t time, i.e. from t + Δ t to t +2 Δ t, calculating the next sample feature value X, Y until all the original data are feature-extracted, and the original signal feature extraction mode is shown in fig. 2. Because partial data is invalid values when the characteristics of the time domain, the frequency domain and the time-frequency domain are extracted, corresponding processing is needed, otherwise, the model training is negatively influenced. For example, in actual processing, the spindle stops, which is similar to no-load, or the spindle motor has great influence on the acquired data at the moment of starting and braking, belongs to noise, and should be identified and deleted. Therefore, for the sample set after feature extraction, threshold judgment is carried out on the absolute average feature in each sample, if the absolute average feature is idle data, stall data or pulse data, the sample is deleted as a whole, and if the absolute average feature is not invalid data, the sample is retained. Thus, after all samples are screened, the remaining samples are the data-processed sample set, and each sample is the sample generated when the tool is effectively cutting. At this time, the sample characteristic value is a matrix of N × 22, the number of rows N of the matrix is the number of samples, and the structures of the cutting vibration sample characteristic value X and the spindle current characteristic value Y are as follows:
initial wear, normal wear, and extreme wear are three major stages that characterize the extent of tool wear during milling, as shown in fig. 3. The method mainly measures the wear value VB of the region of the rear cutter face of the cutter, and the wear value VB is used as the basis for judging the wear degree of the cutter;
the sample target value, i.e., the actually measured tool flank wear value, is measured at regular intervals. Since the measured wear is several moments and the wear value is a continuous curve, interpolation of the wear values between the actual wear value coordinates is required. The method adopts cubic polynomial to carry out interpolation, and for the cubic polynomial
ω j Is t j Y is the interpolated tool wear value and t is time. For the tool wear value yi acquired at xi moment, N times are acquired in total, and then the loss function of the cubic polynomial interpolation curve is:
the coefficient omega of the difference curve can be obtained by solving the following equation j ,
The results of fitting the curve with a cubic polynomial are shown in FIG. 4. The fitting curve shows that the tool is quickly worn in the initial stage, the tool is slowly worn in the middle stage, the tool is quickly worn in the later stage, and the tool wear rate is consistent with the tool wear curve condition, so that the tool wear curve fitted by using a cubic polynomial has high fitting precision;
through cubic polynomial curve fitting, a tool wear value at each moment, that is, a sample target value Q can be obtained. The target value Q is an N × 1 matrix. And (3) fusing the time domain, frequency domain and time-frequency domain characteristics of the cutting vibration signal and the spindle current signal extracted under different wear states of the cutter by using a multi-source information fusion technology, so as to obtain a final sample set (X, Q) & (Y, Q).
S5: the tool wear state is detected based on the convolutional neural network by adopting an efficient deep learning algorithm considering model training time and precision, a classifier is constructed on an output layer after a series of operations such as convolutional layer and pooling layer of the convolutional neural network, and tool state information is output, so that the tool wear state of the numerical control machining of the complex curved surface part is healthily managed;
the efficient deep learning algorithm is mainly used for image recognition and classification tasks and is a Convolutional Neural Network (CNN), which is a weight-sharing neural network structure composed of an input layer, a convolutional layer, a pooling layer, a full-link layer and an output, and the structure of the efficient deep learning algorithm is shown in FIG. 5;
the invention reforms the traditional neural network structure, and designs a tool wear state recognition convolution neural network architecture independently, thereby realizing the health management of the tool wear state of the numerical control machining of complex curved surface parts, and the network is composed of 22 characteristics as an input layer, 2 convolution layers, 2 pooling layers and 3 full-connection layers, as shown in fig. 6;
training a self-designed tool state recognition convolution neural network architecture, wherein the training process mainly comprises two stages of forward propagation and backward propagation. The forward propagation is to input the extracted sample features into the CNN, and obtain the output of the network, i.e. the probability distribution of the tool wear category, through a series of operations such as convolution, pooling, full connection, and the like. And the back propagation is to calculate the error between the output probability value of the CNN network and the standard answer, then to reversely propagate the calculated error to obtain the error of each layer, and finally to adjust the parameters of the whole network by using a gradient descent method, thereby perfecting the whole CNN system model. Observing the accuracy and the loss function of the model in fig. 7, it can be seen that the accuracy of the model shows an ascending trend in the first 50 iteration processes, and then the accuracy is gradually improved;
respectively extracting 117 samples from slightly worn (label 1), normally worn (label 2) and rapidly worn (label 3) labels in a test set to test the model, wherein the prediction result of the test set is shown in fig. 8, and the accuracy of the test set can reach 97.44%; the confusion matrix is shown in fig. 9, when the model tests slightly worn samples, 1 sample is wrongly classified to be normally worn, and the test accuracy is 97.1%; when normal wear samples are tested, 1 sample is misclassified to be worn rapidly, and the accuracy rate is 96.2%; when testing sharply worn samples, 1 sample was misclassified to normal wear with an accuracy of 98.2%.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. The wear prediction and health management method based on the five-axis machining center cutter is characterized by comprising the following steps of: the method comprises the following steps:
s1: the method comprises the following steps of acquiring a cutting vibration signal and a spindle current signal in the milling process of a five-axis machining center in real time by using a sensor, and indirectly monitoring relevant data of the wear condition of a cutter, so as to provide a basis for realizing health management, mode identification and service life prediction of the cutter;
s2: preprocessing the various collected original signals, eliminating noise influence, and further extracting characteristic quantities related to the wear state of the cutter in a time domain and a frequency domain;
s3: meanwhile, wavelet packet analysis is carried out, time-frequency domain information of the signals is locally analyzed through multi-scale and different resolutions, and local time-frequency characteristics required by tool wear state identification are researched, so that the focusing on any details of the signals is realized;
s4: extracting all characteristics related to the wear state of the cutter, and then performing further data processing to form a sample set for model training;
s5: the tool wear state is detected based on the convolutional neural network by adopting an efficient deep learning algorithm considering model training time and precision, a classifier is constructed on an output layer after a series of operations such as convolutional layer and pooling layer of the convolutional neural network, tool state information is output, and therefore health management of the tool wear state of the numerical control machining of the complex curved surface part is achieved.
2. The five-axis machining center cutter based wear prediction and health management method of claim 1, comprising: the vibration signal is generated by the periodic vibration of a cutting system consisting of a machine tool workpiece or a cutter.
3. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the sensor for collecting vibration signals includes, but is not limited to, a piezoresistive acceleration sensor, a piezoelectric acceleration sensor and a capacitive acceleration sensor.
4. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the spindle current signal is the working current generated by the spindle in the process of machining parts.
5. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the sensor for collecting the main shaft current signal is a current sensor.
6. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the time domain features are mainly divided into dimensional features and dimensionless features.
7. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the dimensional features include absolute mean, variance, effective value, peak-to-peak value.
8. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the dimensionless features include skewness index, kurtosis index, peak factor, variation coefficient, and form factor.
9. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the frequency domain features comprise frequency mean square, frequency center of gravity, frequency variance and peak frequency.
10. The five-axis machining center cutter-based wear prediction and health management method of claim 1, wherein: the calculation formula of the time-frequency domain characteristics is as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211502529.4A CN115741235A (en) | 2022-11-28 | 2022-11-28 | Wear prediction and health management method based on five-axis machining center cutter |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211502529.4A CN115741235A (en) | 2022-11-28 | 2022-11-28 | Wear prediction and health management method based on five-axis machining center cutter |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115741235A true CN115741235A (en) | 2023-03-07 |
Family
ID=85339405
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211502529.4A Withdrawn CN115741235A (en) | 2022-11-28 | 2022-11-28 | Wear prediction and health management method based on five-axis machining center cutter |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115741235A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN117708508A (en) * | 2024-02-06 | 2024-03-15 | 深圳宇翊技术股份有限公司 | Method, device, equipment and storage medium for predicting service life of disconnecting switch |
CN117826715A (en) * | 2023-11-29 | 2024-04-05 | 苏州晶软信息科技有限公司 | Intelligent device monitoring and early warning system |
-
2022
- 2022-11-28 CN CN202211502529.4A patent/CN115741235A/en not_active Withdrawn
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN117826715A (en) * | 2023-11-29 | 2024-04-05 | 苏州晶软信息科技有限公司 | Intelligent device monitoring and early warning system |
CN117708508A (en) * | 2024-02-06 | 2024-03-15 | 深圳宇翊技术股份有限公司 | Method, device, equipment and storage medium for predicting service life of disconnecting switch |
CN117708508B (en) * | 2024-02-06 | 2024-04-30 | 深圳宇翊技术股份有限公司 | Method, device, equipment and storage medium for predicting service life of disconnecting switch |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115741235A (en) | Wear prediction and health management method based on five-axis machining center cutter | |
CN105834835B (en) | A kind of tool wear on-line monitoring method based on Multiscale Principal Component Analysis | |
CN114619292B (en) | Milling cutter wear monitoring method based on fusion of wavelet denoising and attention mechanism with GRU network | |
TWI481978B (en) | Method for predicting machining quality of machine tool | |
CN110647943A (en) | Cutting tool wear monitoring method based on evolutionary data cluster analysis | |
CN113496312A (en) | Cutter residual life prediction method and equipment based on multi-dimensional feature extraction fusion and long-term and short-term memory network and storage medium | |
Niu et al. | Remaining useful life prediction of machining tools by 1D-CNN LSTM network | |
CN113579851B (en) | Non-stationary drilling process monitoring method based on adaptive segmented PCA | |
CN112766182A (en) | Method for identifying wear state of broaching tool | |
CN114871850B (en) | Tool wear state assessment method based on vibration signals and BP neural network | |
CN106198020A (en) | Wind turbines bearing failure diagnosis method based on subspace and fuzzy C-means clustering | |
CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
Du et al. | Intelligent turning tool monitoring with neural network adaptive learning | |
CN114523338A (en) | Cutting tool wear state monitoring method based on noise analysis | |
CN112781820A (en) | Hob performance degradation trend evaluation method | |
CN114611745A (en) | Machine tool evaluation method, machine tool evaluation system, and medium | |
CN114326593B (en) | Cutter life prediction system and method | |
Yu et al. | Fault feature of gearbox vibration signals based on morphological filter dynamic convolution autoencoder | |
Wang et al. | Research on tool remaining life prediction method based on CNN-LSTM-PSO | |
CN112801139B (en) | Intelligent cutter wear state identification method based on heterogeneous domain self-adaptive transfer learning | |
CN118060972A (en) | Cutter wear monitoring method based on discrete wavelet decomposition | |
CN114279704B (en) | Method and system for monitoring planet-level running state of wind power gear box | |
CN115526424A (en) | Machine tool spindle Z-direction thermal deformation prediction method based on multi-source heterogeneous information fusion | |
Lorenti et al. | Cuad-mo: Continuos unsupervised anomaly detection on machining operations | |
Li et al. | Discrimination and prediction of tool wear state based on gray theory |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20230307 |