CN115722797A - Laser welding signal analysis method based on machine learning - Google Patents
Laser welding signal analysis method based on machine learning Download PDFInfo
- Publication number
- CN115722797A CN115722797A CN202211369035.3A CN202211369035A CN115722797A CN 115722797 A CN115722797 A CN 115722797A CN 202211369035 A CN202211369035 A CN 202211369035A CN 115722797 A CN115722797 A CN 115722797A
- Authority
- CN
- China
- Prior art keywords
- laser welding
- analysis method
- sample signals
- signal analysis
- welding 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
- 238000003466 welding Methods 0.000 title claims abstract description 54
- 238000004458 analytical method Methods 0.000 title claims abstract description 23
- 238000010801 machine learning Methods 0.000 title claims abstract description 19
- 238000000034 method Methods 0.000 claims abstract description 46
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000005070 sampling Methods 0.000 claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000009826 distribution Methods 0.000 claims abstract description 13
- 238000005457 optimization Methods 0.000 claims abstract description 12
- 238000010606 normalization Methods 0.000 claims abstract description 5
- 238000000638 solvent extraction Methods 0.000 claims abstract description 4
- 230000003321 amplification Effects 0.000 claims description 6
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000000203 mixture Substances 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 claims description 2
- 238000012804 iterative process Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 5
- 239000000463 material Substances 0.000 description 6
- 230000007547 defect Effects 0.000 description 5
- 238000012512 characterization method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000002844 melting Methods 0.000 description 2
- 230000008018 melting Effects 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011897 real-time detection Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001311 chemical methods and process Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Landscapes
- Laser Beam Processing (AREA)
Abstract
The invention belongs to the technical field of computers, and discloses a laser welding signal analysis method based on machine learning, which comprises the following steps: carrying out data normalization on the sampling parameters, setting acquisition parameters, and acquiring positive sample signals and negative sample signals; partitioning the whole signal according to the data distribution characteristics of each point, and removing a detection area from an area with a large fluctuation range; selecting an optimal standard area generation method and a detection model based on machine learning, and carrying out hyper-parameter optimization on the detection model; and training the detection model to obtain a target model, and inputting the laser welding signal into the target model to obtain an analysis result. The laser welding signal analysis method provided by the invention adopts the machine learning technology to construct the detection model, provides feedback and data support for the laser system, and analyzes and processes the laser welding signal through the detection model, thereby judging the quality of the welding quality and avoiding the influence of personnel on the welding quality judgment standard.
Description
Technical Field
The invention belongs to the technical field of computers, and particularly relates to a laser welding signal analysis method based on machine learning.
Background
Laser welding is a new processing technology, has the advantages of high welding point quality, small deformation of a welding area, high production efficiency, easy realization of automation, flexibility, non-contact processing and the like, and is widely applied to the industrial manufacturing fields of 3C electronics, automobile industry, new energy and the like. The principle of laser welding is that a laser generates laser, the laser is focused on the surface of a material to be processed through optical fiber conduction and a lens, the material is heated, melted and gasified, the welding process is an extremely complex physical and chemical process, a molten pool is formed by melting the material, a keyhole is generated by melting the material, metal vapor is formed by gasifying the metal, and plasma is generated.
The stability of the laser welding process is also affected by many factors, such as the change of the material surface, the change of the laser optical system, the deformation or deviation of the welding mechanism, etc., which may cause welding defects. However, unlike the traditional welding method which can monitor the current and voltage or the penetration to monitor the welding process, the welding quality can only be judged by vision and spot check, and the welding quality can not be confirmed or real-time online monitoring can not be realized by the method, so that the welding quality can not be guaranteed in percentage, and the monitoring effect is not ideal.
Many researches show that various signals such as optical signals, acoustic signals, electric field signals, heat radiation and the like can be generated in the laser welding process, and the judgment of welding defects can be realized through the change of the signals. Along with the development of detection technology, the real-time detection technology of the laser welding process makes great progress and development, but the existing real-time detection product of the laser welding process is low in integration level, complex in system, incapable of generating feedback to the laser welding process, weak in autonomous learning capacity of the system, and large in influence of human factors on monitoring results.
Disclosure of Invention
The present invention aims to solve the above technical problem at least to some extent. Therefore, the invention aims to provide a laser welding signal analysis method based on machine learning.
The technical scheme adopted by the invention is as follows:
a laser welding signal analysis method based on machine learning comprises the following steps:
s1, carrying out data normalization on sampling parameters; initializing acquisition parameters, acquiring multiple groups of positive sample signals, judging whether the acquired positive sample signals are in an optimal linear region according to standard deviations of the acquired positive sample signals, if so, saving the acquisition parameters, otherwise, setting new acquisition parameters, and acquiring multiple groups of positive sample signals again and judging;
s2, collecting a positive sample signal and a negative sample signal;
s3, partitioning the whole signal through the data distribution characteristics of each point, and removing a detection area from an area with a large fluctuation range;
s4, selecting an optimal standard region generation method and a detection model based on machine learning, and carrying out hyper-parameter optimization on the detection model;
and S5, training the detection model to obtain a target model, and inputting a laser welding signal into the target model to obtain an analysis result.
Preferably, in step S1, the initialized acquisition parameters include a circuit analog amplification coefficient, a depolarizer angle in a light intensity continuously adjustable structure, a digital amplification factor, a sampling frequency, and a sampling time.
Preferably, the continuously adjustable structure for light intensity comprises a polarizer and an analyzer.
Preferably, in step S1, when the acquired normal sample signal is not in the optimal linear region, a new acquisition parameter is set according to a deviation value of the real data from the optimal linear region.
Preferably, in step S2, more than 50 positive-sample signals and more than 30 negative-sample signals are collected.
Preferably, step S3 further comprises: and (4) counting the average value of the normal sample signals, and if a certain area of the normal sample signals seriously deviates from the average value, rejecting the normal sample signals.
Preferably, in step S4, the standard region is generated by using the positive sample signal, where the time t is an X axis and the sampling value is a Y axis, and the standard region generation method on the Y axis may adopt a gaussian mixture method, a percentage method, an average value method, a median method or a maximum/minimum value method.
The invention has the beneficial effects that:
the laser welding signal analysis method based on machine learning provided by the invention adopts the machine learning technology to construct the detection model, provides feedback and data support for the laser system, and analyzes and processes the laser welding signal through the detection model, so that the quality of welding is judged, the influence of personnel on the welding quality judgment standard is avoided, and the stability of the welding system is tracked and fed back for a long time.
Drawings
Fig. 1 is a flow chart of a laser welding signal analysis method of the present invention.
Fig. 2 is a schematic diagram of a light intensity continuously adjustable structure according to the present invention.
Fig. 3 is a flow chart of the present invention for setting acquisition parameters.
Fig. 4 is a diagram of the general format of the signal of the present invention.
Fig. 5 is a flow chart of step S4 of the present invention.
FIG. 6 is a schematic diagram of the hyper-parametric optimization of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it is to be understood 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.
As shown in fig. 1, the laser welding signal analysis method based on machine learning of this embodiment specifically includes the following steps:
s1, data normalization is carried out on sampling parameters, continuous adjustment of signal intensity is achieved, the signal distribution ranges of different processes are consistent, standardization of reference data is achieved, and effective and rapid learning of a model (linear discriminant, SVM) in the subsequent step S5 can be promoted based on a small number of samples.
The data normalization of the sampling parameters adopts a formula:
setting acquisition parameters, as shown in fig. 3, specifically includes: initializing acquisition parameters, acquiring multiple groups of positive sample signals, judging whether the acquired positive sample signals are in an optimal linear region according to standard deviation of the acquired positive sample signals, if so, saving the acquisition parameters, otherwise, setting new acquisition parameters according to deviation values of real data and the optimal linear region, and acquiring multiple groups of positive sample signals again and judging.
The initialized acquisition parameters comprise circuit analog amplification coefficients, the angle of the depolarizer in the light intensity continuous adjustable structure, digital amplification times, sampling frequency and sampling time. As shown in fig. 2, the continuously adjustable light intensity structure includes a polarizer and an analyzer, and the continuously adjustable light intensity structure is used for realizing continuous adjustment of light intensity. The same data distribution can be obtained under different processes.
And S2, collecting more than 50 positive sample signals and more than 30 negative sample signals, wherein the negative sample signals are not necessarily equal to the positive sample signals due to the high cost of processing materials. When the cost is not considered, the same number of positive-sample signals and negative-sample signals can be collected.
And S3, counting the average value of the normal sample signals, and rejecting the normal sample signals if a certain region of the normal sample signals deviates from the average value seriously. Due to the various complexity of laser welding, the center of the normal signal also has data that the signal distribution deviates greatly from the normal signal distribution. The elimination of abnormal signals is realized by counting the average value of the positive signals: if a certain area of the positive sample signal is seriously deviated from the average value, the signal is not taken as the positive sample signal. Through the data cleaning, the problem of model overfitting in the step S5 is solved under the condition of less sample data, namely, the risk of model overfitting is reduced by ensuring the correctness of the sample.
Partitioning the whole signal according to the data distribution characteristics of each point, and removing a detection area from an area with a large fluctuation range; as shown in fig. 4, if the fluctuation value of the edge portion on the left side of the signal is large, the edge portion is removed from the detection area.
S4, as shown in FIG. 5, selecting an optimal standard area generation method and a detection model based on machine learning, and carrying out hyper-parameter optimization on the detection model;
the method specifically comprises the following steps:
and generating a standard area by using the positive sample signal, wherein the time t is an X axis, the sampling value is a Y axis, and the standard area generation method on the Y axis can adopt a Gaussian mixture method, a percentage method, an average value method, a median method or a maximum/minimum value method to realize the selection of the optimal standard area generation method.
Gaussian mixture method:
assuming that the sampling value of each signal at a certain time is in normal distribution (time t is X axis, the sampling value is Y axis), the value within 3 standard deviations is taken as the standard range.
The percentage method comprises the following steps:
the Y-axis calibration area is obtained by subtracting the upper and lower percentages considered to be defined from the original sample value.
Method of averaging
The Y-axis standard area is given by the mean value as the center, plus the artificially defined upper and lower fluctuation values.
Median method
The Y-axis is obtained by taking the median as the center of the standard area and adding the artificially defined upper and lower fluctuation values.
Maximum/minimum method
The Y-axis acquisition criteria area is given by the raw acquisition data.
The detection method of the detection model can adopt the following methods:
local area characterization method
Defect definition: the actual signal exceeds a certain area of the standard region and the duration exceeds the specified time
Learning the target: defining top/bottom area and time by hyper-parametric optimization automated machine learning
Signal value characterization method
Defect definition: the actual signal exceeding the limit value and the permissible time given according to the standard range
Learning the target: the upper/lower limit values and the tolerance times are defined by hyperparametric optimization automated machine learning.
Average offset characterization method
Defect definition: the signal mean shift exceeds a limit value which is considered to be defined
Learning the target: defining average offset limit values by hyper-parametric optimization automated machine learning
As shown in fig. 6, the iterative process of hyper-parameter optimization includes:
initializing and training a Bayesian agent model;
selecting a new super parameter group:
and updating the agent model.
S5, training the detection model to obtain a target model:
training set: t = { (x) 1 ,y 1 ),(x 2 ,y 2 ),…,(x N ,y N )}，
y i ∈Y＝{-1,1}；
A classifier: detection model G after hyper-parameter optimization 0 (x) G' (x) is SVM;
training process:
1) Initializing training data weight distribution:
(a) The training is carried out by the user,
(b) Calculation of G 0 (x) Classification error rate of (2):
(c) Calculation of G 0 (x) Coefficient (c):
(d) Updating the weight of the training data, increasing the error class weight, decreasing the correct class weight, focusing on the error class attention,
2) Using G' (x) as classifier, D 1 Repeating the steps (a), (b), (c) and (d) for the initial data weight distribution;
3) Constructing a linear combination of basic classifiers:
and (4) a final classifier:
and inputting the laser welding signal into a target model to obtain an analysis result. And the target model can calculate and compare the laser welding signal with the good product signal so as to judge the welding quality. And according to the signal trend of welding, the laser controller is fed back, and the special reasons such as defocusing and power change are corrected to be deteriorated, so that the welding system is always in a stable state. Based on the characteristic of machine learning, the full automation can be realized, the influence of personnel on the welding quality judgment standard is avoided, and great convenience is provided for a production line.
The invention is not limited to the above alternative embodiments, and any other various forms of products can be obtained by anyone in the light of the present invention, but any changes in shape or structure thereof, which fall within the scope of the present invention as defined in the claims, fall within the scope of the present invention.
Claims (9)
1. A laser welding signal analysis method based on machine learning is characterized by comprising the following steps:
s1, carrying out data normalization on sampling parameters; initializing acquisition parameters, acquiring multiple groups of positive sample signals, judging whether the acquired positive sample signals are in an optimal linear region according to standard deviations of the acquired positive sample signals, if so, saving the acquisition parameters, otherwise, setting new acquisition parameters, and acquiring multiple groups of positive sample signals again and judging;
s2, collecting a positive sample signal and a negative sample signal;
s3, partitioning the whole signal through the data distribution characteristics of each point, and removing a detection area from an area with a large fluctuation range;
s4, selecting an optimal standard region generation method and a detection model based on machine learning, and carrying out hyper-parameter optimization on the detection model;
and S5, training the detection model to obtain a target model, and inputting a laser welding signal into the target model to obtain an analysis result.
2. The laser welding signal analysis method according to claim 1, characterized in that: in step S1, the initialized acquisition parameters include a circuit analog amplification coefficient, a depolarizer angle in a light intensity continuously adjustable structure, a digital amplification factor, a sampling frequency, and a sampling time.
3. The laser welding signal analysis method according to claim 2, characterized in that: the light intensity continuously adjustable structure comprises a polarizer and an analyzer.
4. The laser welding signal analysis method according to claim 1, characterized in that: in step S1, when the collected normal sample signal is not in the optimal linear area, new collection parameters are set according to the deviation value of the real data and the optimal linear area.
5. The laser welding signal analysis method according to claim 1, characterized in that: in step S2, more than 50 positive-sample signals and more than 30 negative-sample signals are collected.
6. The laser welding signal analysis method according to claim 1, characterized in that: step S3 further includes: and (4) counting the average value of the positive sample signals, and if a certain region of the positive sample signals deviates from the average value seriously, rejecting the positive sample signals.
7. The laser welding signal analysis method according to claim 1, characterized in that: in step S4, a standard region is generated by using the sample signal, where time t is an X axis, a sampling value is a Y axis, and a gaussian mixture method, a percentage method, an average value method, a median method, or a maximum/minimum value method may be used as a method for generating the standard region on the Y axis.
8. The laser welding signal analysis method according to claim 1, characterized in that: in step S4, the iterative process of the hyper-parameter optimization includes:
initializing and training a Bayesian agent model;
selecting a new super-parameter group:
and updating the agent model.
9. The laser welding signal analysis method according to claim 1, characterized in that: in the step S5, the process is carried out,
training set:
a classifier: detection model G after hyper-parameter optimization 0 (x) G' (x) is SVM;
a training process:
1) Initializing training data weight distribution:
(a) The training is carried out by the user,
(b) Calculation of G 0 (x) Classification error rate of (2):
(c) Calculation of G 0 (x) Coefficient (c):
(d) Updating the weight of the training data, increasing the error weight, decreasing the correct weight, focusing on the error attention,
2) Using G' (x) as a classifier, D 1 Repeating the steps (a), (b), (c) and (d) for the initial data weight distribution;
3) Constructing a linear combination of basic classifiers:
and (4) a final classifier:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211369035.3A CN115722797A (en) | 2022-11-03 | 2022-11-03 | Laser welding signal analysis method based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211369035.3A CN115722797A (en) | 2022-11-03 | 2022-11-03 | Laser welding signal analysis method based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115722797A true CN115722797A (en) | 2023-03-03 |
Family
ID=85294475
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211369035.3A Pending CN115722797A (en) | 2022-11-03 | 2022-11-03 | Laser welding signal analysis method based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115722797A (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102008058422A1 (en) * | 2008-11-21 | 2010-05-27 | Precitec Itm Gmbh | Monitoring laser machining process to be carried out on workpiece, comprises detecting actual measuring values by sensor, which monitors the laser machining process, and determining actual characteristic values from actual measuring values |
| US20110192825A1 (en) * | 2010-02-08 | 2011-08-11 | Prima Industrie S.P.A. | Method for monitoring the quality of laser-machining processes and corresponding system |
| CN107199397A (en) * | 2016-03-17 | 2017-09-26 | 发那科株式会社 | Machine learning device, laser-processing system and machine learning method |
| CN107256393A (en) * | 2017-06-05 | 2017-10-17 | 四川大学 | The feature extraction and state recognition of one-dimensional physiological signal based on deep learning |
| CN109300116A (en) * | 2018-09-03 | 2019-02-01 | 广东工业大学 | The online defect identification method of laser welding based on machine learning |
| JP2020179406A (en) * | 2019-04-24 | 2020-11-05 | 株式会社ナ・デックス | Welding monitoring system and welding monitoring method for resistance welder |
| CN113221468A (en) * | 2021-05-31 | 2021-08-06 | 福州大学 | Photovoltaic array fault diagnosis method based on ensemble learning |
| EP3885069A1 (en) * | 2020-03-25 | 2021-09-29 | Bystronic Laser AG | Quality control of a laser machining process by means of machine learning |
| CN113536921A (en) * | 2021-06-11 | 2021-10-22 | 五邑大学 | Arc welding quality evaluation method and device and storage medium |
| US20210354234A1 (en) * | 2019-01-21 | 2021-11-18 | Mitsubishi Electric Corporation | Processing state detecting device, laser processing machine, and machine learning device |
| CN113743651A (en) * | 2021-08-04 | 2021-12-03 | 三峡大学 | Electronic voltage transformer error state prediction method based on machine learning and SVM algorithm |
| CN113762410A (en) * | 2021-09-17 | 2021-12-07 | 哈尔滨工业大学 | Key welding process parameter prediction method based on machine learning |
| WO2022053526A1 (en) * | 2020-09-09 | 2022-03-17 | Precitec Gmbh & Co. Kg | Method for monitoring the condition of a laser machining head, and laser machining system for carrying out said method |
| CN115115192A (en) * | 2022-06-15 | 2022-09-27 | 浙江工业大学 | Method for detecting welding quality in manual welding process |
| CN115146530A (en) * | 2022-06-15 | 2022-10-04 | 蕴硕物联技术(上海)有限公司 | Method, apparatus, medium, and program product for constructing welding quality detection model |
-
2022
- 2022-11-03 CN CN202211369035.3A patent/CN115722797A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102008058422A1 (en) * | 2008-11-21 | 2010-05-27 | Precitec Itm Gmbh | Monitoring laser machining process to be carried out on workpiece, comprises detecting actual measuring values by sensor, which monitors the laser machining process, and determining actual characteristic values from actual measuring values |
| US20110192825A1 (en) * | 2010-02-08 | 2011-08-11 | Prima Industrie S.P.A. | Method for monitoring the quality of laser-machining processes and corresponding system |
| CN107199397A (en) * | 2016-03-17 | 2017-09-26 | 发那科株式会社 | Machine learning device, laser-processing system and machine learning method |
| CN107256393A (en) * | 2017-06-05 | 2017-10-17 | 四川大学 | The feature extraction and state recognition of one-dimensional physiological signal based on deep learning |
| CN109300116A (en) * | 2018-09-03 | 2019-02-01 | 广东工业大学 | The online defect identification method of laser welding based on machine learning |
| US20210354234A1 (en) * | 2019-01-21 | 2021-11-18 | Mitsubishi Electric Corporation | Processing state detecting device, laser processing machine, and machine learning device |
| JP2020179406A (en) * | 2019-04-24 | 2020-11-05 | 株式会社ナ・デックス | Welding monitoring system and welding monitoring method for resistance welder |
| EP3885069A1 (en) * | 2020-03-25 | 2021-09-29 | Bystronic Laser AG | Quality control of a laser machining process by means of machine learning |
| WO2022053526A1 (en) * | 2020-09-09 | 2022-03-17 | Precitec Gmbh & Co. Kg | Method for monitoring the condition of a laser machining head, and laser machining system for carrying out said method |
| CN113221468A (en) * | 2021-05-31 | 2021-08-06 | 福州大学 | Photovoltaic array fault diagnosis method based on ensemble learning |
| CN113536921A (en) * | 2021-06-11 | 2021-10-22 | 五邑大学 | Arc welding quality evaluation method and device and storage medium |
| CN113743651A (en) * | 2021-08-04 | 2021-12-03 | 三峡大学 | Electronic voltage transformer error state prediction method based on machine learning and SVM algorithm |
| CN113762410A (en) * | 2021-09-17 | 2021-12-07 | 哈尔滨工业大学 | Key welding process parameter prediction method based on machine learning |
| CN115115192A (en) * | 2022-06-15 | 2022-09-27 | 浙江工业大学 | Method for detecting welding quality in manual welding process |
| CN115146530A (en) * | 2022-06-15 | 2022-10-04 | 蕴硕物联技术(上海)有限公司 | Method, apparatus, medium, and program product for constructing welding quality detection model |
Non-Patent Citations (2)
| Title |
|---|
| 李驰;李广军;: "机器学习算法在焊接领域中的应用", 焊管, no. 05, 28 May 2016 (2016-05-28), pages 30 - 35 * |
| 程力勇;米高阳;黎硕;胡席远;王春明;: "基于主成分分析-支持向量机模型的激光钎焊接头质量诊断", 中国激光, no. 03, 10 March 2017 (2017-03-10), pages 82 - 89 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xiong et al. | Deposition height detection and feature point extraction in robotic GTA-based additive manufacturing using passive vision sensing | |
| Kao et al. | Laser cladding quality monitoring using coaxial image based on machine learning | |
| Yu et al. | Application of arc plasma spectral information in the monitor of Al–Mg alloy pulsed GTAW penetration status based on fuzzy logic system | |
| CN115722797A (en) | Laser welding signal analysis method based on machine learning | |
| CN110539054A (en) | Electric arc increase material stacking operation monitoring system | |
| CN115305526A (en) | Self-adaptive control method for consistency of copper foil thickness and surface density based on X-ray measurement | |
| CN109816020A (en) | The laser melting coating optimization technique of mahalanobis distance based on Principal Component Analysis | |
| CN112183665A (en) | CO based on convolutional neural network2Welding pool image identification method | |
| Pacher et al. | Real-time continuous estimation of dross attachment in the laser cutting process based on process emission images | |
| CN115908410A (en) | Machine vision-based pressure container laser welding control method | |
| Hong et al. | Filter-PCA-based process monitoring and defect identification during climbing helium arc welding process using DE-SVM | |
| CN115601313A (en) | Visual monitoring management system for tempered glass production process | |
| Ai et al. | The characteristics extraction of weld seam in the laser welding of dissimilar materials by different image segmentation methods | |
| CN113182709B (en) | Plate laser cutting process parameter optimization method considering sharp corner ablation | |
| CN116228798B (en) | Intelligent iron tower cutting detection method based on machine vision | |
| CN110232399A (en) | The transmission facility defect analysis method and system clustered based on Set Pair Analysis and K-means | |
| Lei et al. | Electric arc length control of circular seam in welding robot based on arc voltage sensing | |
| CN113902955A (en) | Prediction model training method, electronic device, laser diagnosis device and method | |
| CN116563277B (en) | Machine vision-based copper-clad aluminum surface defect detection method and system | |
| Muhammad et al. | control schemes for low density polyethylene reactor | |
| CN116765427A (en) | Real-time prediction method for size of molten pool in selective laser melting process | |
| CN113656372B (en) | Standard index database data mart architecture device and method | |
| Fei et al. | The research of classification method of arc welding pool image based on invariant moments | |
| CN109960307B (en) | MPPT active disturbance rejection control method for photovoltaic off-grid inverter | |
| Guan et al. | Single-crystal quality detection and evaluation algorithms applied to roller mill orientation instrument |
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 |