CN116089818B - Workpiece surface roughness prediction method, system and product in machining process - Google Patents
Workpiece surface roughness prediction method, system and product in machining process Download PDFInfo
- Publication number
- CN116089818B CN116089818B CN202310030459.5A CN202310030459A CN116089818B CN 116089818 B CN116089818 B CN 116089818B CN 202310030459 A CN202310030459 A CN 202310030459A CN 116089818 B CN116089818 B CN 116089818B
- Authority
- CN
- China
- Prior art keywords
- data
- trained
- machining
- feature
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003754 machining Methods 0.000 title claims abstract description 129
- 230000003746 surface roughness Effects 0.000 title claims abstract description 99
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000012545 processing Methods 0.000 claims abstract description 113
- 230000003068 static effect Effects 0.000 claims abstract description 110
- 238000000605 extraction Methods 0.000 claims abstract description 83
- 238000003701 mechanical milling Methods 0.000 claims description 44
- 230000004931 aggregating effect Effects 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 15
- 238000005457 optimization Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 9
- 239000002245 particle Substances 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 6
- 238000003860 storage Methods 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 description 20
- 230000008569 process Effects 0.000 description 17
- 238000005516 engineering process Methods 0.000 description 12
- 238000007418 data mining Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 9
- 238000007781 pre-processing Methods 0.000 description 8
- 238000004220 aggregation Methods 0.000 description 6
- 230000002776 aggregation Effects 0.000 description 6
- 238000011161 development Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000005520 cutting process Methods 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 241000282414 Homo sapiens Species 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000001351 cycling effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 125000006850 spacer group Chemical group 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Numerical Control (AREA)
Abstract
The invention discloses a method, a system and a product for predicting the surface roughness of a workpiece in the machining process, which relate to the technical field of online prediction of machining quality, and comprise the following steps: acquiring a data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted; performing feature extraction on the data set to be predicted by using a vibration time sequence data feature extraction model of the gate control circulation unit to obtain dynamic processing feature data; the dynamic processing characteristic data and the static processing characteristic data are aggregated to obtain aggregated processing characteristic data; and inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model. The invention can improve the predicting precision of the surface roughness of the workpiece in the machining process.
Description
Technical Field
The invention relates to the technical field of on-line prediction of machining quality, in particular to a method, a system and a product for predicting the surface roughness of a workpiece in the machining process.
Background
The human beings develop from the mechanical revolution to the electrified revolution and further to the informatization revolution, and the three revolution are all leaps of social development and bring the human society into a new era. The Germany proposes the concept of industry 4.0, the tide of the fourth industrial revolution is accompanied by great changes in international situation, and the brand new technologies of artificial intelligence, big data, everything interconnection, cloud platform and the like are deeply changed in the development tracks of various countries, various field industries, various enterprises and individuals, so that the world has come to the century. With the rapid development of global economy and science and technology, countries in the world compete strongly in various fields, the aerospace manufacturing industry is an advanced and highly-technical-intensive industry, represents the level of the whole manufacturing industry of one country, and is the core of international competition while serving as the pulse of the national advanced and new technology industry. The strategic status of the aerospace manufacturing technology is strengthened, the development of technology is continuously advanced, and the technology is a serious issue in the development of the national high-precision technology. Under the influence of the large trend of intelligent manufacturing, the aerospace manufacturing industry also faces the pressure and challenges of digital and intelligent transformation, and the intelligent manufacturing obviously becomes one of the main directions of new industrial technological transformation of aerospace in China, and has extremely high requirements on processing equipment, tools, processes and the like because the intelligent manufacturing industry is at the highest end of equipment manufacturing industry. Under the background, how to advance the digital intelligent transformation of the aerospace manufacturing industry and realize high-efficiency and high-precision processing has become an important research problem for various enterprises, universities and scientific research institutions.
In the manufacturing scene of the aerospace industry, the quality of the aerospace product plays a decisive role in the service life of the aerospace product, and the surface roughness can accurately describe the fatigue strength, the wear resistance, the surface hardness and other properties of the product, so the quality evaluation method is often used for evaluating the quality of the aerospace product. The surface roughness has an extremely important influence on the assembly accuracy and the service performance of the product. The overlarge surface roughness not only can accelerate the abrasion of mechanical products, but also can cause poor assembly quality of parts and further cause fatigue fracture, and extremely bad influence is caused on the safety and reliability of the aerospace vehicle. In the fields of aerospace precision manufacturing, which are extremely important and complex, immeasurably huge losses are caused if the surface quality of mechanical parts is problematic. Therefore, surface roughness is considered as a key evaluation index of product quality.
The detection of surface roughness in conventional manufacturing is to detect the product after finishing the machining, which is very inefficient, and the detection results indicate that a large number of work pieces are not acceptable and may result in very high costs. In recent years, breakthroughs in various fields such as sensor technology, communication technology, cloud platform manufacturing technology and the like provide strong support for acquiring aerospace multi-source data. Therefore, in the material processing process, the online prediction of the surface roughness for multi-source time series data has very important significance.
In recent years, the number of researches related to online surface roughness prediction of multi-source time series data is small, and according to the existing algorithm principle of the online surface roughness prediction model for the multi-source time series data, the surface roughness prediction algorithm can be generally divided into three types: surface roughness prediction based on traditional theory, surface roughness prediction based on traditional data mining, surface roughness prediction based on deep learning. Because the aerospace industry manufacturing data has the characteristics of small samples, high cost and the like, and deep learning is generally oriented to a large-sample high-dimension data scene, the surface roughness prediction method based on the deep learning is not suitable for the aerospace industry manufacturing scene. Although the surface roughness prediction method based on the traditional theory and the surface roughness prediction method based on the traditional data mining are suitable for manufacturing scenes of the aerospace industry, the following problems exist:
the surface roughness prediction method based on the traditional theory focuses on the potential mechanism of surface formation, and generally realizes modeling of surface roughness according to physical phenomena observed in the numerical control cutting process, such as periodic repetition of cutter blade profile and cutter blade corrugation, plastic lateral flow, material expansion caused by material rebound and the like. Theoretical modeling methods are critical for understanding surface formation and obtaining better surface quality, but the surface roughness of machined parts is mainly affected by coupling of factors such as machining parameters, workpiece materials, cutters, the whole machining environment and the like. The theoretical modeling method is difficult to process such complex and nonlinear coupling influence relation, and a more effective method is needed to integrate complex influence factors into the surface roughness prediction model so as to improve the prediction performance. The surface roughness modeling method based on data mining can analyze and process high-dimensional nonlinear surface roughness influencing factors, and because of the strong analysis decision capability, the surface roughness modeling method is widely applied to the prediction of the surface roughness of mechanical manufacturing products. The existing data mining modeling method mostly adopts static factors such as process parameters, cutters, workpieces and the like to establish a surface roughness prediction model for machining, and ignores a large amount of dynamic machining information contained in dynamic signals such as vibration, force and the like generated by cutting. Therefore, the existing surface roughness prediction method based on the traditional theory is difficult to process the current complex problem (complex and nonlinear coupling influence relation), while the existing surface roughness prediction method based on the traditional data mining can process the current complex problem, most of the existing surface roughness prediction method based on the traditional data mining only uses static factors such as process parameters, cutters, workpieces and the like to predict the surface roughness, but ignores dynamic signals such as vibration, force and the like, so that the accuracy of predicting the surface roughness of the workpieces in the machining process is lower.
Along with the rapid development of communication technology and sensor technology, dynamic signals are considered in a small part of research at present, in order to improve the precision of workpiece surface roughness prediction in the machining process, a part of students adopt dynamic signals such as vibration, force and the like to acquire time-varying information of machining conditions and environments in the machining process, however, in the characteristic extraction, a mode of manually extracting characteristics (manual characteristic extraction based on signal statistics characteristics) is adopted, namely, in the traditional data mining modeling method, a mechanical machining surface roughness prediction model is established by adopting static factors such as process parameters, cutters and workpieces and dynamic time-varying information such as vibration, force and the like acquired in the manual characteristic extraction mode, and in the manual characteristic extraction mode, only a specific small number of characteristics can be extracted in the time domain, so that a large amount of useful information in the dynamic signals is directly lost, the prediction precision of the model is greatly limited, and therefore, the workpiece surface roughness prediction precision in the machining process is lower.
Disclosure of Invention
The invention aims to provide a method, a system and a product for predicting the surface roughness of a workpiece in the machining process, so as to improve the predicting precision of the surface roughness of the workpiece in the machining process.
In order to achieve the above object, the present invention provides the following solutions:
a method of predicting surface roughness of a workpiece during machining, the method comprising:
acquiring a data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted;
performing feature extraction on the data set to be predicted by using a vibration time sequence data feature extraction model of the gate control circulation unit to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set comprises vibration signals on a main shaft and an on-shaft vice, technological parameter static factors, tool static factors, workpiece static factors and workpiece surface roughness corresponding to the vibration signals on the main shaft and the on-shaft vice, the technological parameter static factors, the tool static factors and the workpiece static factors;
aggregating the dynamic processing characteristic data and the static processing characteristic data to obtain aggregated processing characteristic data; the static tooling characteristic data is provided by the data set to be predicted;
inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using the mechanical milling data set and the gating circulation unit vibration time sequence data characteristic extraction model.
Optionally, the aggregate processing characteristic data to be trained is obtained by aggregating dynamic processing characteristic data to be trained and static processing characteristic data to be trained;
the dynamic processing characteristic data to be trained are obtained by utilizing the trained gating circulation unit vibration time sequence data characteristic extraction model to extract the characteristics of the data set to be trained; the static tooling feature data is provided by the data set to be trained.
Optionally, the gating circulation unit vibration time sequence data feature extraction model takes an aggregate feature spliced by a dynamic feature and a static feature as input, and takes workpiece surface roughness corresponding to the aggregate feature provided by the mechanical milling data set as output training;
the dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set; the static feature is provided by the mechanical milling dataset.
Optionally, the multi-core support vector regression machining roughness prediction model utilizes an improved particle swarm optimization algorithm to optimize model parameters.
The invention also provides the following scheme:
a system for predicting surface roughness of a workpiece during a machining process, the system comprising:
the data set to be predicted acquisition module is used for acquiring the data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted;
the feature extraction module is used for carrying out feature extraction on the data set to be predicted by utilizing a gating circulation unit vibration time sequence data feature extraction model to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set comprises vibration signals on a main shaft and an on-shaft vice, technological parameter static factors, tool static factors, workpiece static factors and workpiece surface roughness corresponding to the vibration signals on the main shaft and the on-shaft vice, the technological parameter static factors, the tool static factors and the workpiece static factors;
the feature fusion module is used for aggregating the dynamic processing feature data and the static processing feature data to obtain aggregated processing feature data; the static tooling characteristic data is provided by the data set to be predicted;
the prediction module is used for inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using the mechanical milling data set and the gating circulation unit vibration time sequence data characteristic extraction model.
Optionally, the aggregate processing characteristic data to be trained is obtained by aggregating dynamic processing characteristic data to be trained and static processing characteristic data to be trained;
the dynamic processing characteristic data to be trained are obtained by utilizing the trained gating circulation unit vibration time sequence data characteristic extraction model to extract the characteristics of the data set to be trained; the static tooling feature data is provided by the data set to be trained.
Optionally, the gating circulation unit vibration time sequence data feature extraction model takes an aggregate feature spliced by a dynamic feature and a static feature as input, and takes workpiece surface roughness corresponding to the aggregate feature provided by the mechanical milling data set as output training;
the dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set; the static feature is provided by the mechanical milling dataset.
Optionally, the multi-core support vector regression machining roughness prediction model utilizes an improved particle swarm optimization algorithm to optimize model parameters.
The invention also provides the following scheme:
an electronic device comprising a memory for storing a computer program and a processor that runs the computer program to cause the electronic device to perform the method of predicting surface roughness of a workpiece during machining.
The invention also provides the following scheme:
a computer readable storage medium storing a computer program which when executed by a processor implements the method of predicting surface roughness of a workpiece during machining.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
according to the method, the system and the product for predicting the surface roughness of the workpiece in the machining process, disclosed by the invention, the characteristic extraction is performed by using the characteristic extraction model of the vibration time sequence data of the gate control circulation unit, the model is suitable for processing time sequence data such as vibration signals, although the small data volume can not fully train the model, the characteristic which is superior to the combination of the coarse granularity and the fine granularity of the artificial characteristic can be extracted, and the characteristic is used for training the multi-core support vector regression machining roughness prediction model, so that the surface roughness of the workpiece in the machining process can be predicted more efficiently and accurately, and the surface roughness prediction precision of the workpiece in the machining process is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a method for predicting the surface roughness of a workpiece during machining according to an embodiment of the present invention;
FIG. 2 is a flowchart of a method for online prediction of machining roughness based on incremental multi-core support vector regression provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of a machining roughness online prediction method based on incremental multi-core support vector regression according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a wavelet denoising process according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a gating cycle unit vibration feature extraction model according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a method, a system and a product for predicting the surface roughness of a workpiece in the machining process, so as to improve the predicting precision of the surface roughness of the workpiece in the machining process.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
Example 1
FIG. 1 is a flowchart of a method for predicting the surface roughness of a workpiece during machining according to an embodiment of the invention. As shown in fig. 1, the present embodiment provides a method for predicting the surface roughness of a workpiece in a machining process, including the following steps:
step S1: acquiring a data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted.
Step S2: performing feature extraction on the data set to be predicted by using a vibration time sequence data feature extraction model of the gate control circulation unit to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set includes vibration signals on the spindle and on-axis vise, process parameter statics, tool statics, workpiece statics, and workpiece surface roughness corresponding to the vibration signals on the spindle and on-axis vise, process parameter statics, tool statics, and workpiece statics.
Step S3: the dynamic processing characteristic data and the static processing characteristic data are aggregated to obtain aggregated processing characteristic data; the static tooling signature data is provided by the data set to be predicted.
Step S4: inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using a mechanical milling data set and a gating circulation unit vibration time sequence data characteristic extraction model.
The to-be-trained aggregation processing characteristic data is obtained by aggregating to-be-trained dynamic processing characteristic data and to-be-trained static processing characteristic data. The dynamic processing characteristic data to be trained is obtained by extracting the characteristics of the data set to be trained by using the trained gating circulating unit vibration time sequence data characteristic extraction model. The static tooling feature data is provided by the data set to be trained.
The vibration time sequence data feature extraction model of the gate control circulation unit takes the aggregate features spliced by the dynamic features and the static features as input and takes the workpiece surface roughness corresponding to the aggregate features provided by the mechanical milling data set as output training. The dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set. The static features are provided by the mechanical milling dataset.
The multi-core support vector regression machining roughness prediction model utilizes an improved particle swarm optimization algorithm to optimize model parameters.
The technical scheme of the invention is described in the following by a specific embodiment:
the invention discloses a method for predicting the surface roughness of a workpiece in the machining process, which is an online method for predicting the roughness of the workpiece in the machining process based on incremental multi-core support vector regression.
Fig. 2 is a flowchart of a method for online predicting machining roughness based on incremental multi-core support vector regression according to an embodiment of the present invention. Fig. 3 is a schematic diagram of a machining roughness online prediction method based on incremental multi-core support vector regression according to an embodiment of the present invention. As shown in fig. 2 and 3, the method for online predicting machining roughness based on incremental multi-core support vector regression of the invention specifically comprises the following steps:
step 101: preprocessing the mechanical processing multisource vibration time sequence data to obtain preprocessed vibration data.
The step 101 specifically includes:
and cleaning, normalizing, wavelet denoising and gray correlation analysis are carried out on the multi-source vibration time sequence data to obtain preprocessed vibration data. Wherein, the wavelet denoising adopts a global unified threshold valueσ=mad/0.6755, MAD is the median of the absolute value of the first layer wavelet decomposition coefficient, 0.6755 is the adjustment coefficient of gaussian noise standard deviation, N is the scale of the signal, σ is the noise standard deviation. The wavelet denoising process is shown in fig. 4.
The mechanical processing multi-source vibration time sequence data are mechanical milling data sets of open sources of Taiwan university of middle school, the data sets comprise vibration signals on a main shaft and an on-shaft vice, and the vibration signals are measured through an accelerometer; and static factors such as process parameters, cutters, workpieces and the like and corresponding roughness.
Step 102: training a gating cycle unit vibration time sequence data feature extraction model (a gating cycle unit vibration feature extraction model) through the preprocessing data (vibration data) in the step 101, and establishing a trained feature extraction model to obtain a trained feature extraction model.
The gating circulation unit vibration time sequence data feature extraction model comprises a source feature input layer (vibration data input layer), a feature amplification layer (vibration feature amplification layer) and a feature aggregation output layer (feature extraction output layer). The characteristic extraction model of the vibration time sequence data of the gating circulation unit adopts a bidirectional multi-head gating circulation unit to extract the characteristic, and a schematic diagram of the vibration characteristic extraction model of the gating circulation unit is shown in fig. 5. Wherein p is t And q t Representing an update gate and a reset gate, respectively. The update gate is used to control the extent to which the state information at the previous time is brought into the current state, and a larger value of the update gate indicates that the state information at the previous time is brought more. The reset gate controls how much information from the previous state is to be written to the current candidate setThe smaller the reset gate, the less information of the previous state is written. The GRU combines the forget gate and the input gate into a single update gate, and also combines the cell state C and the hidden state g, whose mathematical expression is:
p t ＝σ(W p ·[g t-1 ,w t ])
q t ＝σ(W q ·[g t-1 ,w t ])
wherein σ represents a sigmoid function, W p Weight matrix for controlling new and old information retention degree at t moment, W q Weight matrix g representing the reservation degree of each input at time t-1 t-1 Represents the hidden state transferred from the previous moment, g t Represents the hidden state, w, at the current moment t An input vector representing the current time, W representing a weight matrix controlling the state of the current time, and a gating signal p t The range of (2) is 0 to 1. The closer the gating signal is to 1, the more data is represented as "remembered"; while a closer to 0 represents more "forget".
And when the gating circulation unit vibration time sequence data characteristic extraction model is trained, the aggregation characteristics of the dynamic characteristics and the static characteristics are taken as input, and the real roughness training model provided by the mechanical milling data set, namely the gating circulation unit vibration time sequence data characteristic extraction model is trained by taking the aggregation characteristics of the dynamic characteristics and the static characteristics as input and the workpiece surface roughness corresponding to the aggregation characteristics as output. The workpiece surface roughness corresponding to the aggregated features is provided by a mechanical milling dataset. The dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set. The static features are provided by the mechanical milling dataset.
Step 103: and (3) performing feature extraction on the vibration time sequence data by using the feature extraction model (the trained feature extraction model) trained in the step (102) to obtain dynamic processing feature data (vibration data on a main shaft and an on-shaft vice) to be trained. Compared with the static processing characteristic data to be trained, the dynamic processing characteristic data to be trained more shows real-time information in the processing process, including unknown factors such as processing environment and the like, and is closer to an actual processing scene. The feature extraction data (dynamic machining feature data to be trained) includes dynamic machining feature extraction data of vibration time series data of a plurality of sets of different experimental parameter settings (spindle rotation speed, feed rate, cutting depth, vice clamping torque, etc.).
Step 104: and (3) aggregating (fusing) and preprocessing the dynamic processing characteristic data to be trained obtained in the step (103) and the static processing characteristic data to be trained provided by the mechanical milling data set to obtain preprocessed aggregate processing characteristic data to be trained. The data fusion adopts a transverse splicing mode, and the data preprocessing mainly comprises data cleaning, data normalization, gray correlation analysis and the like.
Step 105: and (3) training a multi-core support vector regression machining roughness prediction model (multi-core support vector regression prediction model) through the aggregated machining characteristic data to be trained obtained in the step (104), and optimizing model parameters of the prediction model by utilizing an improved particle swarm optimization algorithm to obtain a trained roughness prediction model (trained machining quality prediction model).
The multi-core support vector regression machining roughness prediction model makes a 'interval band' on both sides of the mixed kernel function, and does not calculate loss for all samples falling into the interval band; samples outside the interval band are taken into account for the loss function. Finally, the model is optimized by minimizing the width of the spacer bands and the total loss through an improved particle swarm optimization algorithm (PSO is not used for the existing SVR surface roughness prediction). The multi-core support vector regression machining roughness prediction model introduces a focus mechanism in the kernel function part of support vector regression.
And when the multi-core support vector regression machining roughness prediction model is trained, dynamic machining feature data to be trained and aggregate machining feature data to be trained which are spliced by static machining feature data to be trained are taken as input, and the real roughness training model provided by the mechanical milling data set, namely the multi-core support vector regression machining roughness prediction model is trained by taking the dynamic machining feature data to be trained and the aggregate machining feature data to be trained which are spliced by the static machining feature data to be trained as input and the surface roughness of the workpiece corresponding to the aggregate machining feature data to be trained as output. The workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained is provided by a mechanical milling data set.
The multi-core support vector regression machining roughness prediction model adopts a combination of a linear kernel function, a polynomial kernel function, a Gaussian kernel function and a Sigmoid kernel function.
Step 106: and (3) performing feature extraction on vibration time sequence data (to-be-predicted data set) transmitted back in real time in the manufacturing process (in the machining process) by using the feature extraction model (the trained feature extraction model) trained in the step 102, so as to obtain real-time dynamic machining feature data.
Step 107: and (3) aggregating and preprocessing the real-time dynamic processing characteristic data obtained in the step (106) and the static processing characteristic data provided by the data set to be predicted to obtain preprocessed real-time aggregate processing characteristic data (aggregating the real-time dynamic processing characteristic data and the static processing characteristic data corresponding to the current processing state into the real-time aggregate processing characteristic data).
Step 108: and (3) predicting the real-time aggregate processing characteristic data by using the multi-core support vector regression roughness prediction model (a trained processing quality prediction model) trained in the step (105) to obtain real-time mechanical processing roughness (predicting real-time processing quality).
The multi-core support vector regression roughness prediction model is a multi-core support vector regression machining roughness prediction model. Machining roughness is the surface roughness of a workpiece.
Step 109: and (3) performing incremental fine adjustment (online optimization of the multi-core support vector regression roughness prediction model) on the multi-core support vector regression machining roughness prediction model by utilizing the real-time aggregate machining characteristic data (the real-time aggregate machining characteristic data obtained in step 107) and the real roughness at the last moment, enhancing the adaptability of the prediction model to dynamic factors such as environment and the like, and improving the machining roughness (machining quality) prediction precision.
The incremental learning method utilizes real-time aggregate processing characteristic data and real roughness, and combines an attention mechanism to carry out incremental fine adjustment on the multi-core support vector regression machining roughness prediction model so as to enhance the adaptability of the prediction model to dynamic factors such as environment and the like and improve the machining roughness prediction precision.
Step 110: and (5) carrying out on-line prediction on the machining roughness (machining quality) by cycling all the steps until the machining process is finished.
According to the method, dynamic machining characteristic data in multidimensional vibration time series data are automatically extracted based on a gating circulation unit vibration time series data characteristic extraction model, dynamic information related to surface roughness is fully utilized, and real-time machining roughness conditions are predicted on line through an incremental multi-core support vector regression roughness on-line prediction model (an incremental multi-core support vector regression prediction model). According to the invention, complex influence factors are integrated into the machining roughness prediction model through the incremental multi-core support vector regression model so as to improve the prediction performance.
The invention integrates the surface roughness prediction algorithm based on the traditional data mining and the surface roughness prediction algorithm based on the deep learning, wherein the deep learning is used for extracting the vibration signal characteristics, and the data mining is used for roughness prediction. The invention introduces vibration, force and other dynamic signals capable of reflecting environmental factors and static factors to predict the surface roughness. The invention uses GRU to extract end-to-end characteristics, so as to improve the quality of characteristic extraction. The GRU feature extraction belongs to a deep learning model, so that the GRU feature extraction is not suitable for manufacturing scenes of the aerospace industry, but the model is suitable for processing time series data such as vibration signals, and although the model cannot be fully trained by small data quantity, the feature which is superior to the artificial feature and combined with the coarse granularity can be extracted, and the SVR model can be trained by the feature to more efficiently and accurately predict the surface roughness, which is equivalent to longitudinal integrated learning. Compared with the prior art, the invention avoids manual extraction in feature extraction, thereby being capable of extracting a large amount of useful information.
Example two
In order to perform a corresponding method of the above embodiments to achieve the corresponding functions and technical effects, a system for predicting surface roughness of a workpiece during machining is provided below, the system comprising:
the data set to be predicted acquisition module is used for acquiring the data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted.
The feature extraction module is used for carrying out feature extraction on the data set to be predicted by utilizing the feature extraction model of the vibration time sequence data of the gating circulating unit to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set includes vibration signals on the spindle and on-axis vise, process parameter statics, tool statics, workpiece statics, and workpiece surface roughness corresponding to the vibration signals on the spindle and on-axis vise, process parameter statics, tool statics, and workpiece statics.
The feature fusion module is used for aggregating the dynamic processing feature data and the static processing feature data to obtain aggregated processing feature data; the static tooling signature data is provided by the data set to be predicted.
The prediction module is used for inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using a mechanical milling data set and a gating circulation unit vibration time sequence data characteristic extraction model.
The to-be-trained aggregation processing characteristic data is obtained by aggregating to-be-trained dynamic processing characteristic data and to-be-trained static processing characteristic data. The dynamic processing characteristic data to be trained is obtained by extracting the characteristics of the data set to be trained by using the trained gating circulating unit vibration time sequence data characteristic extraction model. The static tooling feature data is provided by the data set to be trained.
The vibration time sequence data feature extraction model of the gate control circulation unit takes the aggregate features spliced by the dynamic features and the static features as input and takes the workpiece surface roughness corresponding to the aggregate features provided by the mechanical milling data set as output training. The dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set. The static features are provided by the mechanical milling dataset.
The multi-core support vector regression machining roughness prediction model utilizes an improved particle swarm optimization algorithm to optimize model parameters.
The invention provides a machining roughness online prediction system integrating feature extraction and quality prediction, which is a machining roughness online prediction system based on incremental multi-core support vector regression, and specifically comprises the following modules:
the pretreatment module is used for carrying out pretreatment on the mechanical processing multisource vibration time sequence data to obtain pretreated data.
The first training module is used for training the characteristic extraction model of the vibration time sequence data of the gating circulating unit through the preprocessing data and establishing a trained characteristic extraction model.
The first feature extraction module is used for carrying out feature extraction on the vibration time sequence data by utilizing the trained feature extraction model to obtain dynamic processing feature data; the feature extraction data includes dynamic machining feature extraction data of the vibration timing data of a plurality of sets of different experimental parameter settings (spindle rotation speed, feed rate, cutting depth, vise clamping torque, etc.).
And the first feature fusion module is used for aggregating and preprocessing the dynamic processing feature data and the static processing feature data to obtain preprocessed aggregate processing feature data.
And the second training module is used for training the multi-core support vector regression machining roughness prediction model through the aggregate machining feature data, and optimizing model parameters of the prediction model by utilizing an improved particle swarm optimization algorithm to obtain a trained roughness prediction model.
And the second feature extraction module is used for carrying out feature extraction on vibration time sequence data transmitted back in real time in the manufacturing process by utilizing the trained feature extraction model to obtain real-time dynamic processing feature data.
And the second feature fusion module is used for aggregating and preprocessing the real-time dynamic processing feature data and the static processing feature data to obtain preprocessed real-time aggregate processing feature data.
And the prediction module is used for predicting the real-time aggregate processing characteristic data by using the trained roughness prediction model to obtain the real-time machining roughness.
And the increment module is used for carrying out increment fine adjustment on the mechanical processing roughness prediction model by utilizing the real-time aggregate processing characteristic data and the real roughness, enhancing the adaptability of the prediction model to dynamic factors such as environment and the like, and improving the mechanical processing roughness prediction precision.
Example III
An electronic device according to a third embodiment of the present invention includes a memory for storing a computer program and a processor for executing the computer program to cause the electronic device to execute the method for predicting surface roughness of a workpiece during machining according to the first embodiment.
The electronic device may be a server.
Example IV
A fourth embodiment of the present invention provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements the method for predicting surface roughness of a workpiece during machining of the first embodiment.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (10)
1. A method for predicting surface roughness of a workpiece during machining, the method comprising:
acquiring a data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted;
performing feature extraction on the data set to be predicted by using a vibration time sequence data feature extraction model of the gate control circulation unit to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set comprises vibration signals on a main shaft and an on-shaft vice, technological parameter static factors, tool static factors, workpiece static factors and workpiece surface roughness corresponding to the vibration signals on the main shaft and the on-shaft vice, the technological parameter static factors, the tool static factors and the workpiece static factors;
aggregating the dynamic processing characteristic data and the static processing characteristic data to obtain aggregated processing characteristic data; the static tooling characteristic data is provided by the data set to be predicted;
inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using the mechanical milling data set and the gating circulation unit vibration time sequence data characteristic extraction model.
2. The method for predicting the surface roughness of a workpiece in a machining process according to claim 1, wherein the to-be-trained aggregated machining feature data is obtained by aggregating to-be-trained dynamic machining feature data and to-be-trained static machining feature data;
the dynamic processing characteristic data to be trained are obtained by utilizing the trained gating circulation unit vibration time sequence data characteristic extraction model to extract the characteristics of the data set to be trained; the static tooling feature data is provided by the data set to be trained.
3. The method for predicting the surface roughness of a workpiece in a machining process according to claim 1, wherein the gating cycle unit vibration time sequence data feature extraction model takes an aggregate feature of dynamic feature and static feature splicing as input and takes the surface roughness of the workpiece corresponding to the aggregate feature provided by the mechanical milling data set as output training;
the dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set; the static feature is provided by the mechanical milling dataset.
4. The method of claim 1, wherein the multi-core support vector regression machining roughness prediction model utilizes an improved particle swarm optimization algorithm for model parameter optimization.
5. A system for predicting surface roughness of a workpiece during a machining process, the system comprising:
the data set to be predicted acquisition module is used for acquiring the data set to be predicted in the machining process in real time; the data set to be predicted comprises vibration signals on a main shaft to be predicted and a vice on the shaft, static factors of technological parameters to be predicted, static factors of a cutter to be predicted and static factors of a workpiece to be predicted;
the feature extraction module is used for carrying out feature extraction on the data set to be predicted by utilizing a gating circulation unit vibration time sequence data feature extraction model to obtain dynamic processing feature data; the gating circulation unit vibration time sequence data characteristic extraction model is trained by using a mechanical milling data set; the mechanical milling data set comprises vibration signals on a main shaft and an on-shaft vice, technological parameter static factors, tool static factors, workpiece static factors and workpiece surface roughness corresponding to the vibration signals on the main shaft and the on-shaft vice, the technological parameter static factors, the tool static factors and the workpiece static factors;
the feature fusion module is used for aggregating the dynamic processing feature data and the static processing feature data to obtain aggregated processing feature data; the static tooling characteristic data is provided by the data set to be predicted;
the prediction module is used for inputting the aggregate processing characteristic data into a multi-core support vector regression machining roughness prediction model, and predicting the surface roughness of the workpiece by using the multi-core support vector regression machining roughness prediction model; the multi-core support vector regression machining roughness prediction model is trained by utilizing a data set to be trained; the data set to be trained comprises aggregate processing characteristic data to be trained and workpiece surface roughness corresponding to the aggregate processing characteristic data to be trained; the aggregate processing characteristic data to be trained is obtained by using the mechanical milling data set and the gating circulation unit vibration time sequence data characteristic extraction model.
6. The system for predicting the surface roughness of a workpiece during machining of claim 5, wherein the aggregate machining feature data to be trained is obtained by aggregating dynamic machining feature data to be trained with static machining feature data to be trained;
the dynamic processing characteristic data to be trained are obtained by utilizing the trained gating circulation unit vibration time sequence data characteristic extraction model to extract the characteristics of the data set to be trained; the static tooling feature data is provided by the data set to be trained.
7. The system according to claim 5, wherein the gating cycle unit vibration time sequence data feature extraction model takes an aggregate feature of dynamic feature and static feature concatenation as input, and takes a workpiece surface roughness corresponding to the aggregate feature provided by the mechanical milling data set as output training;
the dynamic characteristics are obtained by carrying out wavelet denoising on the mechanical milling data set; the static feature is provided by the mechanical milling dataset.
8. The in-machine work piece surface roughness prediction system of claim 5, wherein the multi-core support vector regression machining roughness prediction model utilizes a modified particle swarm optimization algorithm for model parameter optimization.
9. An electronic device comprising a memory for storing a computer program and a processor that runs the computer program to cause the electronic device to perform the method of predicting surface roughness of a workpiece during machining as claimed in any one of claims 1 to 4.
10. A computer-readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements the method for predicting the surface roughness of a workpiece during machining as claimed in any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310030459.5A CN116089818B (en) | 2023-01-10 | 2023-01-10 | Workpiece surface roughness prediction method, system and product in machining process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310030459.5A CN116089818B (en) | 2023-01-10 | 2023-01-10 | Workpiece surface roughness prediction method, system and product in machining process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116089818A CN116089818A (en) | 2023-05-09 |
| CN116089818B true CN116089818B (en) | 2023-10-27 |
Family
ID=86211646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310030459.5A Active CN116089818B (en) | 2023-01-10 | 2023-01-10 | Workpiece surface roughness prediction method, system and product in machining process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116089818B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102592035A (en) * | 2012-03-20 | 2012-07-18 | 北京航空航天大学 | Method for predicating surface roughness and surface topography simulation of car milling compound machining |
| CN108596158A (en) * | 2018-05-15 | 2018-09-28 | 同济大学 | A kind of Surface Roughness in Turning prediction technique based on energy consumption |
| CN111859566A (en) * | 2020-07-17 | 2020-10-30 | 重庆大学 | Surface roughness stabilizing method based on digital twinning |
| CN113704922A (en) * | 2021-09-01 | 2021-11-26 | 江苏师范大学 | Method for predicting surface roughness based on sound vibration and texture features |
| CN113703395A (en) * | 2021-07-07 | 2021-11-26 | 西北工业大学 | Variable working condition milling machining clamping force prediction method for machining deformation control |
| CN115422978A (en) * | 2022-09-20 | 2022-12-02 | 温州大学 | Workpiece surface roughness prediction method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021174525A1 (en) * | 2020-03-06 | 2021-09-10 | 大连理工大学 | Parts surface roughness and cutting tool wear prediction method based on multi-task learning |
| WO2022076451A1 (en) * | 2020-10-06 | 2022-04-14 | Sentient Science Corporation | Systems and methods for modeling performance in a part manufactured using an additive manufacturing process |
-
2023
- 2023-01-10 CN CN202310030459.5A patent/CN116089818B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102592035A (en) * | 2012-03-20 | 2012-07-18 | 北京航空航天大学 | Method for predicating surface roughness and surface topography simulation of car milling compound machining |
| CN108596158A (en) * | 2018-05-15 | 2018-09-28 | 同济大学 | A kind of Surface Roughness in Turning prediction technique based on energy consumption |
| CN111859566A (en) * | 2020-07-17 | 2020-10-30 | 重庆大学 | Surface roughness stabilizing method based on digital twinning |
| CN113703395A (en) * | 2021-07-07 | 2021-11-26 | 西北工业大学 | Variable working condition milling machining clamping force prediction method for machining deformation control |
| CN113704922A (en) * | 2021-09-01 | 2021-11-26 | 江苏师范大学 | Method for predicting surface roughness based on sound vibration and texture features |
| CN115422978A (en) * | 2022-09-20 | 2022-12-02 | 温州大学 | Workpiece surface roughness prediction method |
Non-Patent Citations (3)
| Title |
|---|
| 基于Elman网络的切削表面粗糙度预测方法;粟潘;孙华;张永顺;张江平;;机械管理开发(第05期);第203页至第206页 * |
| 基于深度学习的雾天夜间机场能见度预测方法;徐悦等;《电子制作》;第60页至第62页 * |
| 涂层刀具高速铣削模具钢SKD11的表面粗糙度模型预测;谢英星;《工具技术》;第122页至126页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116089818A (en) | 2023-05-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Milling force prediction model based on transfer learning and neural network | |
| Palanikumar et al. | Analysis on drilling of glass fiber–reinforced polymer (GFRP) composites using grey relational analysis | |
| CN108803486B (en) | Numerical control machine tool thermal error prediction and compensation method based on parallel deep learning network | |
| CN113414638B (en) | Variable working condition milling cutter wear state prediction method based on milling force time sequence diagram deep learning | |
| CN114235330B (en) | Multi-source pneumatic load model construction method for correlation wind tunnel test and calculation data | |
| CN113554621B (en) | Cutter wear state identification system and method based on wavelet scale map and deep migration learning | |
| Deng et al. | Reliability analysis of chatter stability for milling process system with uncertainties based on neural network and fourth moment method | |
| Feng et al. | A novel energy evaluation approach of machining processes based on data analysis | |
| Zhang et al. | Tool wear online monitoring method based on DT and SSAE-PHMM | |
| Wu et al. | Quality estimation method for gear hobbing based on attention and adversarial transfer learning | |
| Liu et al. | An accurate prediction method of multiple deterioration forms of tool based on multitask learning with low rank tensor constraint | |
| Gupta et al. | Predictive soft modeling of turning parameters using artificial neural network | |
| CN115099135A (en) | Improved artificial neural network multi-type operation power consumption prediction method | |
| CN116089818B (en) | Workpiece surface roughness prediction method, system and product in machining process | |
| Yang et al. | Milling cutter wear prediction method under variable working conditions based on LRCN | |
| Chen et al. | Milling chatter monitoring under variable cutting conditions based on time series features | |
| Yang et al. | Tool wear prediction based on parallel dual-channel adaptive feature fusion | |
| Adesta et al. | Prediction of cutting temperatures by using back propagation neural network modeling when cutting hardened H-13 steel in CNC end milling | |
| Zhao et al. | CNC thermal compensation based on mind evolutionary algorithm optimized BP neural network | |
| Deng et al. | On-line surface roughness classification for multiple CNC milling conditions based on transfer learning and neural network | |
| Sun et al. | A New Semi-supervised Tool-wear Monitoring Method using Unreliable Pseudo-Labels | |
| CN114357851A (en) | Numerical control milling surface roughness prediction method based on DAE-RNN | |
| Liu et al. | Prediction of cutting force via machine learning: state of the art, challenges and potentials | |
| Jin et al. | Milling process stability detection for curved workpiece based on MVMD and LSTM | |
| Cheng | The application of computer vision technology in the field of industrial automation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |