CN109732450B - Blade polishing and grinding method based on neural network - Google Patents
Blade polishing and grinding method based on neural network Download PDFInfo
- Publication number
- CN109732450B CN109732450B CN201910144036.XA CN201910144036A CN109732450B CN 109732450 B CN109732450 B CN 109732450B CN 201910144036 A CN201910144036 A CN 201910144036A CN 109732450 B CN109732450 B CN 109732450B
- Authority
- CN
- China
- Prior art keywords
- blade
- neural network
- model
- machining
- theoretical
- 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
Images
Landscapes
- Numerical Control (AREA)
- Constituent Portions Of Griding Lathes, Driving, Sensing And Control (AREA)
Abstract
The invention relates to a blade polishing and grinding processing method based on a neural network, which comprises the steps of scanning and measuring a rough part of a blade through a measuring device, transmitting information to computer software to obtain a reconstructed three-dimensional model, comparing the reconstructed three-dimensional model with characteristic points on a theoretical model, adjusting and enabling coordinate systems of the two models to be superposed, and calculating the theoretical processing allowance of the blade; training a neural network model by using historical data, so that the neural network model can identify the processing process parameter configuration under different theoretical processing allowances; inputting the theoretical machining allowance of the blade into the neural network model, and calculating machining technological parameters of blade machining by the neural network model for polishing and grinding; scanning and measuring the processed blade, transmitting information to computer software to obtain an updated model, comparing the updated model with the three-dimensional model before processing, and calculating the material removal amount of the blade; and if the size of the machined blade meets the tolerance requirement, the machining is finished.
Description
Technical Field
The invention belongs to the field of machining, and particularly relates to a blade polishing and grinding machining method based on a neural network.
Background
The level of mechanical manufacture is the manifestation of comprehensive national force of a country. Although the total value of the manufacturing industry in China is the first world, the key core technology and the top products are far behind developed countries. The research and development and the manufacture of the aeroengine are used as important components of national defense industry, and the technical level of the aeroengine is related to national defense safety and comprehensive national force of a country. Blades are key components of aircraft engines. At present, the conventional process route of die forging blank-rough milling-finish milling-polishing is mainly adopted for the processing method of the blade. The purpose of milling is to improve the accuracy of the blade profile and the purpose of grinding and polishing is to improve the surface quality. At present, the mode of combining manual coping and numerical control polishing is mainly adopted for grinding and polishing the blade profile. When the mode of manual grinding is adopted, the influence of human factors on the processing quality of the blade is large, and the consistency of the processing quality is difficult to ensure. When a numerical control polishing mode is adopted, the processing parameters are usually selected by technicians according to experience or an empirical formula, and the processing efficiency is difficult to ensure.
A neural network is an operational model, which is formed by connecting a large number of nodes (or neurons). The essence of neural networks is to closely model the functional relationship of inputs to outputs through a number of parameters and activation functions. The neural network algorithm discards an original mathematical and physical modeling method, does not need to study the relation between each process parameter and a processing process, and can output related parameters to guide the realization of processing by acquiring enough data.
CN105160059A discloses a blade machining cutting amount optimization selection method based on BP and GA, and an aluminum alloy blade mirror cutting model is established by taking a ball nose cutter spiral mirror cutting machining process as a research object. And (3) performing optimized cutting processing on the simulated aeroengine blade by using ABAQUS nonlinear finite element analysis software. And summarizing the change rule between the cutting parameters and the deformation in the mirror cutting simulation and the relation between the cutting parameters and the deformation by using an artificial neural network algorithm, and establishing a blade mirror cutting machining deformation prediction model. Determining design variables, objective functions and constraint conditions of cutting parameter optimization, constructing a cutting parameter optimization model of the aluminum alloy blade, and optimizing cutting parameters by using a genetic algorithm. It does not involve secondary modeling analysis of the machined blade.
CN107807610A discloses a complex curved surface part mechanical arm processing system and method based on a feature library, wherein the system comprises a feature database, a process planning and feature matching module, a code conversion interface, an information processing module, a processing simulation module and an actual processing module. The method comprises the steps of analyzing and extracting the characteristics of the complex curved surface part, matching similar information from a characteristic database, formulating a processing technological process and setting corresponding processing technological parameters. And carrying out digital simulation analysis on the machining process of the part with the complex curved surface to be machined. And (4) carrying out mechanical arm mirror cutting on the part to be machined with the complex curved surface through the actual machining module. However, in this method, the quality of the blade machining depends on the accuracy of the data in the feature library, and the level of the technician affects the accuracy of the data, so that it is difficult to ensure the machining quality.
Disclosure of Invention
The invention aims to provide a blade polishing and grinding processing method based on a neural network, which takes processing technology parameters and material removal amount as a database, and realizes self-learning and self-evolution of equipment in the processing process through continuously optimizing related parameters by a neural network model so as to achieve the purpose of improving the quality and the efficiency of the processing technology.
The invention relates to a blade polishing and grinding processing method based on a neural network, which comprises the following steps:
the method comprises the following steps: reconstructing the blade through a reverse technology to obtain a three-dimensional model, comparing the three-dimensional model with characteristic points on a theoretical model of the blade, adjusting and enabling coordinate systems of the two models to be overlapped, and calculating the theoretical machining allowance of the blade;
step two: training a neural network model by using historical data in an existing database, so that the neural network model can identify the processing process parameter configuration under different theoretical processing allowances;
step three: inputting the theoretical machining allowance of the blade into the neural network model, calculating machining technological parameters of blade machining by the neural network model, and outputting the machining technological parameters to machining equipment for polishing and grinding;
step four: reconstructing the processed blade again through a reverse technology to obtain an updated three-dimensional model, comparing the updated three-dimensional model with the three-dimensional model before processing, and calculating the material removal amount of the blade;
step five: uploading new data continuously generated in the processing process to a database so as to update the data; retraining the neural network model by using new data, and improving the recognition capability and accuracy of the neural network model;
step six: judging whether the size of the processed blade meets the tolerance requirement, and if so, judging that the processing is finished; and otherwise, returning to the first step and the second step, calculating a corrected machining allowance and transmitting the corrected machining allowance to the neural network model to recalculate the machining process parameters of the blade for polishing machining until the size of the blade meets the tolerance requirement.
Preferably, the reverse technology in the step one is realized by scanning and measuring the rough part of the blade through a measuring device and transmitting the information to computer three-dimensional simulation software to reconstruct and obtain a three-dimensional model; the measuring device is a three-coordinate detector.
Preferably, the specific method for reconstruction is to obtain a closed data point set of the blade at a plurality of different cross-section positions by a three-dimensional detector, and fit the closed data point set by computer three-dimensional simulation software.
Preferably, the specific method for calculating the theoretical machining allowance of the blade in the first step is to obtain a plurality of tool contact points, corresponding normal vectors and straight lines along the direction of the normal vectors on the theoretical model, each straight line intersects with the reconstructed three-dimensional model at a point, and the distance between the point and the tool contact points is the theoretical machining allowance.
Preferably, the neural network model in the second step, the third step, the fifth step and the sixth step is a hybrid network model composed of a convolutional neural network and a fully connected neural network, the convolutional neural network is used for identifying and processing the theoretical machining allowance and the corrected machining allowance of the blade, and the fully connected neural network is used for connecting the convolutional neural network and the output layer.
Preferably, the neural network model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a convolutional layer, a pooling layer and a full-link layer.
Preferably, the input layers are 22 x 1 in size, the output layers are 3 in size, the first hidden layer filter is 3 x 3 in size and 6 in depth, the second hidden layer filter is 3 x 3 in size and 16 in depth, the pooled layer filters are 2 x 2 in size and the length and width steps are 2; the convolution process is de-linearized using the ReLU activation function.
Preferably, the new data in step five includes updated machining process parameters and the actual material removal of the blade.
The invention has the beneficial effects that:
the method takes the material removal amount of the blade in the machining process as the input amount of the neural network model for training, so that the characteristic of surface contact between the abrasive belt and the blade in the machining process is truly reflected.
The measuring device is used for carrying out secondary scanning measurement on the processed blade and transmitting information to computer software to obtain an updated model, the updated model is compared with the three-dimensional model before processing, and the material removal amount of the blade is calculated through comparison with a real object, so that the data is more objective and accurate.
Uploading new data (such as material removal) generated continuously in the machining process to a database so as to update the data; and the new data is used for retraining the neural network model, so that the recognition capability and accuracy of the neural network model are improved, and the accuracy of subsequent processing is improved.
Drawings
FIG. 1 is a flow chart of the present method;
FIG. 2 is a schematic representation of a three-dimensional model of an aircraft engine blade manufactured by the present method;
FIG. 3 is a diagram of a neural network model involved in the method;
1-an input layer; 2-a hidden layer; 3-output layer.
Detailed Description
The technical solution is further explained below with reference to the detailed description and the accompanying drawings.
A blade polishing and grinding processing method based on a neural network comprises the following steps: acquiring blade machining allowance information; training a neural network model by using historical data; determining processing technological parameters through a neural network model according to the processing allowance, and performing online processing; and uploading the parameters in the processing process to a database, and updating the neural network model.
Referring to fig. 1, the method specifically includes the following steps:
the method comprises the following steps: the blade is reconstructed through a reverse technology to obtain a three-dimensional model (as shown in fig. 2), and the three-dimensional model is compared with characteristic points on a theoretical model of the blade, so that coordinate systems of the two models are adjusted and overlapped, and the theoretical machining allowance of the blade is calculated.
The reverse technique is realized by scanning and measuring the rough part of the blade by a measuring device and transmitting the information to computer three-dimensional simulation software (such as UG or ABAQUS) to obtain a three-dimensional model.
The blank piece of the blade is a precision casting blade for an aeroengine and is made of high-temperature alloy. The blade is approximately 150mm in length and 40mm in width.
The measuring device is a three-coordinate detector. The coordinate measuring machine has a displacement measuring system (such as a grating ruler) capable of moving in three directions, and coordinates (x, y, z) of each point of the workpiece are calculated by a data processor or a computer as raw data for reconstructing a three-dimensional model.
The characteristic points refer to points which can play a role of position identification or other roles of positioning and identification.
The specific method of reconstruction is to obtain a plurality of (for example, 22) closed data point sets at different cross-section positions of the blade through a three-dimensional detector, and to fit the closed data point sets through three-dimensional software such as UG (unigraphics).
The specific method for calculating the theoretical machining allowance of the blade is to obtain a plurality of (for example, 22) blade contacts and corresponding normal vectors on a theoretical model, and straight lines along the direction of the normal vectors, wherein each straight line intersects with the reconstructed three-dimensional model at a point, and the distance between the point and the blade contacts is the theoretical machining allowance.
Step two: and training the neural network model by using historical data in the existing database, so that the neural network model can identify the processing process parameter configuration under different theoretical processing allowances.
The historical data in the database comprises machining process parameters used in the blade polishing and grinding machining process and the corresponding material removal amount.
The information storage and transmission form of the material removal amount is a two-dimensional data matrix. The information storage and transmission form of the processing technological parameters is a one-dimensional data matrix. The processing parameters comprise abrasive belt rotating speed, feeding speed, normal pressure and the like. The polishing abrasive belt used was a silica carbide abrasive belt of type XK870F with an abrasive grain size of 120.
The neural network model is a hybrid network model composed of a convolutional neural network and a fully-connected neural network, and comprises an input layer 1, a hidden layer 2 and an output layer 3, as shown in fig. 3. The hidden layer 2 comprises a convolutional layer, a pooling layer and a fully-connected layer, and is generally connected in the following order: input-convolutional layer-pooling layer-full-link layer-output. Pooling is an operation used in convolutional neural networks. The convolution neural network is used for identifying and processing the theoretical machining allowance and the correction machining allowance of the blade, and the full-connection neural network is used for connecting the convolution neural network and the output layer.
The input layers are 22 x 1 in size, the output layers are 3 in size, the first hidden layer filter is 3 x 3 in size, the depth is 6, the second hidden layer filter is 3 x 3 in size, the depth is 16, the pooled layer filters are 2 x 2 in size, and the length and width steps are 2. The convolution process is de-linearized using the ReLU activation function.
The input quantity of the neural network model is the material removal quantity of the blade in the machining process, and the output quantity is a machining process parameter.
The training of the neural network model refers to the construction of a neural network model framework through python, inputting historical data in a database into the model, and learning and training through continuous iteration to finally obtain the model.
Step three: inputting the theoretical machining allowance of the blade into the neural network model, calculating machining technological parameters of blade machining by the neural network model, and outputting the machining technological parameters to machining equipment for polishing and grinding;
step four: and reconstructing the processed blade again through a reverse technology to obtain an updated three-dimensional model, comparing the updated three-dimensional model with the three-dimensional model before processing, and calculating the material removal amount of the blade.
The reverse technology is to scan and measure the processed blade again through the measuring device and transmit the information to the computer three-dimensional simulation software to obtain an updated three-dimensional model.
The measuring device is a three-coordinate detector.
Step five: uploading new data (such as material removal) generated continuously in the machining process to a database so as to update the data; and the new data is used for retraining the neural network model, so that the recognition capability and the accuracy of the neural network model are improved.
The new data includes updated machining process parameters and actual material removal of the blade.
When the data in the database is updated to a certain threshold value, the neural network model automatically calls a training function to learn and train on the existing basis to obtain an updated model.
Step six: judging whether the size of the processed blade meets the tolerance requirement, and if so, judging that the processing is finished; and otherwise, returning to the first step and the second step, calculating a corrected machining allowance and transmitting the corrected machining allowance to the neural network model to recalculate the machining process parameters of the blade for polishing machining until the size of the blade meets the tolerance requirement.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and substitutions can be made without departing from the principle of the present invention, and these modifications and substitutions should also be regarded as the protection scope of the present invention.
Claims (5)
1. A blade polishing and grinding processing method based on a neural network is characterized by comprising the following steps:
the method comprises the following steps: reconstructing the blade through a reverse technology to obtain a three-dimensional model, comparing the three-dimensional model with characteristic points on a theoretical model of the blade, adjusting and enabling coordinate systems of the two models to be overlapped, and calculating the theoretical machining allowance of the blade;
step two: training a neural network model by using historical data in an existing database, so that the neural network model can identify the processing process parameter configuration under different theoretical processing allowances;
step three: inputting the theoretical machining allowance of the blade into the neural network model, calculating machining technological parameters of blade machining by the neural network model, and outputting the machining technological parameters to machining equipment for polishing and grinding;
step four: reconstructing the processed blade again through a reverse technology to obtain an updated three-dimensional model, comparing the updated three-dimensional model with the three-dimensional model before processing, and calculating the material removal amount of the blade;
step five: uploading new data continuously generated in the processing process to a database so as to update the data; retraining the neural network model by using new data, and improving the recognition capability and accuracy of the neural network model;
step six: judging whether the size of the processed blade meets the tolerance requirement, and if so, judging that the processing is finished; otherwise, returning to the first step and the second step, calculating a corrected machining allowance and transmitting the corrected machining allowance to the neural network model to recalculate the machining technological parameters of the blade for polishing and grinding until the size of the blade meets the tolerance requirement;
the neural network model in the second step, the third step, the fifth step and the sixth step is a mixed network model consisting of a convolutional neural network and a fully-connected neural network, the convolutional neural network is used for identifying and processing the theoretical machining allowance and the corrected machining allowance of the blade, and the fully-connected neural network is used for connecting the convolutional neural network and an output layer;
the neural network model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a convolutional layer, a pooling layer and a full-connection layer;
the size of the input layers is 22 x 1, the size of the output layers is 3, the size of the first hidden layer filter is 3 x 3, the depth is 6, the size of the second hidden layer filter is 3 x 3, the depth is 16, the size of the pooling layer filter is 2 x 2, and the length and width step size are both 2; the convolution process is de-linearized using the ReLU activation function.
2. The method for polishing and grinding the blades based on the neural network as claimed in claim 1, wherein the reverse technology in the step one is realized by scanning and measuring the rough parts of the blades through a measuring device and transmitting information to computer three-dimensional simulation software to reconstruct and obtain a three-dimensional model; the measuring device is a three-coordinate detector.
3. The blade polishing and grinding processing method based on the neural network as claimed in claim 2, wherein the specific method of reconstruction is to obtain a closed data point set of a plurality of different cross-section positions of the blade through a three-dimensional detector and to fit the closed data point set through computer three-dimensional simulation software.
4. The method for polishing and grinding the blades based on the neural network as claimed in claim 1, wherein the specific method for calculating the theoretical machining allowance of the blade in the first step is to obtain a plurality of blade contacts and corresponding normal vectors on the theoretical model and straight lines along the direction of the normal vectors, each straight line intersects with the reconstructed three-dimensional model at a point, and the distance between the point and the blade contacts is the theoretical machining allowance.
5. The neural network-based blade polishing machining method according to claim 1, wherein the new data in the fifth step includes updated machining process parameters and an actual material removal amount of the blade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910144036.XA CN109732450B (en) | 2019-02-27 | 2019-02-27 | Blade polishing and grinding method based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910144036.XA CN109732450B (en) | 2019-02-27 | 2019-02-27 | Blade polishing and grinding method based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109732450A CN109732450A (en) | 2019-05-10 |
| CN109732450B true CN109732450B (en) | 2021-02-26 |
Family
ID=66368458
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910144036.XA Active CN109732450B (en) | 2019-02-27 | 2019-02-27 | Blade polishing and grinding method based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109732450B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110153851B (en) * | 2019-05-28 | 2020-08-18 | 华中科技大学 | High-order contact enveloping type grinding and polishing processing control system and method |
| CN110744406A (en) * | 2019-10-29 | 2020-02-04 | 华中科技大学 | Blade profile grinding and polishing self-adaptive machining system and method through robot |
| CN113910085B (en) * | 2021-09-26 | 2022-11-18 | 江苏集萃华科智能装备科技有限公司 | Control method, control device and control system of metal mirror polishing production line |
| CN113681469B (en) * | 2021-10-22 | 2022-03-11 | 南通豪派金属制品有限公司 | Intelligent control method and system for polishing equipment |
| CN114905515B (en) * | 2022-05-26 | 2023-05-12 | 无锡斯帝尔科技有限公司 | Robot control method and system based on flexible perception neural network |
| CN115625595B (en) * | 2022-12-21 | 2023-03-17 | 太原理工大学 | Quantitative control blade polishing and grinding method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000343380A (en) * | 1999-05-31 | 2000-12-12 | Toshiba Mach Co Ltd | Method for determining condition of operating in machine tool |
| CN101901283A (en) * | 2010-06-22 | 2010-12-01 | 北京理工大学 | Prediction method of numerical control bending forming quality of conduit and device |
| CN102880908A (en) * | 2012-09-11 | 2013-01-16 | 天津大学 | Method for calculating remanufactured part environmental loss based on back propagation (BP) neural network |
| CN102930337A (en) * | 2012-10-31 | 2013-02-13 | 贵州大学 | Prediction method for cutting amount in cutting machining |
| CN103659602A (en) * | 2012-09-25 | 2014-03-26 | 宁波工程学院 | Method and device for detecting cylindrical grinding surface roughness through data fusion technology |
| CN108356606A (en) * | 2018-03-19 | 2018-08-03 | 西北工业大学 | Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural |
| CN108573310A (en) * | 2017-03-08 | 2018-09-25 | 发那科株式会社 | Finishing amount prediction meanss and machine learning device |
| CN109048492A (en) * | 2018-07-30 | 2018-12-21 | 北京航空航天大学 | Cutting-tool wear state detection method, device and equipment based on convolutional neural networks |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000233369A (en) * | 1999-02-15 | 2000-08-29 | Noritake Co Ltd | Grinding condition monitoring device and dressing condition monitoring device |
| US7001243B1 (en) * | 2003-06-27 | 2006-02-21 | Lam Research Corporation | Neural network control of chemical mechanical planarization |
| CN102880771B (en) * | 2012-10-31 | 2014-12-03 | 贵州大学 | Method for predicting surface roughness of workpiece during high-speed cutting machining |
| CN105160059A (en) * | 2015-07-11 | 2015-12-16 | 西安工业大学 | BP and GA based blade machining cutting quantity optimization selection method |
| CN106709149B (en) * | 2016-11-25 | 2019-11-19 | 中南大学 | A kind of aluminium cell three-dimensional burner hearth shape real-time predicting method neural network based and system |
-
2019
- 2019-02-27 CN CN201910144036.XA patent/CN109732450B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000343380A (en) * | 1999-05-31 | 2000-12-12 | Toshiba Mach Co Ltd | Method for determining condition of operating in machine tool |
| CN101901283A (en) * | 2010-06-22 | 2010-12-01 | 北京理工大学 | Prediction method of numerical control bending forming quality of conduit and device |
| CN102880908A (en) * | 2012-09-11 | 2013-01-16 | 天津大学 | Method for calculating remanufactured part environmental loss based on back propagation (BP) neural network |
| CN103659602A (en) * | 2012-09-25 | 2014-03-26 | 宁波工程学院 | Method and device for detecting cylindrical grinding surface roughness through data fusion technology |
| CN102930337A (en) * | 2012-10-31 | 2013-02-13 | 贵州大学 | Prediction method for cutting amount in cutting machining |
| CN108573310A (en) * | 2017-03-08 | 2018-09-25 | 发那科株式会社 | Finishing amount prediction meanss and machine learning device |
| CN108356606A (en) * | 2018-03-19 | 2018-08-03 | 西北工业大学 | Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural |
| CN109048492A (en) * | 2018-07-30 | 2018-12-21 | 北京航空航天大学 | Cutting-tool wear state detection method, device and equipment based on convolutional neural networks |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109732450A (en) | 2019-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109732450B (en) | Blade polishing and grinding method based on neural network | |
| US11656614B2 (en) | In-process digital twinning | |
| CN105739440A (en) | Adaptive machining method of wide-chord hollow fan blade | |
| Lin et al. | A generic uniform scallop tool path generation method for five-axis machining of freeform surface | |
| CN101653921A (en) | Error compensation method of numerical control grinding contour of camshaft | |
| CN109683552B (en) | Numerical control machining path generation method on complex point cloud model guided by base curve | |
| Gohari et al. | A digital twin for integrated inspection system in digital manufacturing | |
| CN106202822A (en) | B-spline surface model reconstruction method towards blade adaptive machining | |
| CN112508895B (en) | Propeller blade quality assessment method based on curved surface registration | |
| CN113408195B (en) | Method for predicting wear state of cutter in process | |
| CN109408857A (en) | The intelligent parameter drive module design method of spiral bevel gear shape Collaborative Manufacturing | |
| CN110703686A (en) | On-line measuring path planning method for blade section of blisk | |
| CN116373304A (en) | In-situ repair path planning method for composite damaged member | |
| CN116276328A (en) | Robot polishing track optimization method based on digital twin and visual transmission technology | |
| CN116068959A (en) | Processing method based on tool path compensation, electronic equipment and storage medium | |
| CN115562159A (en) | Tool path planning method, device and computer readable storage medium | |
| Brecher et al. | Evaluation of toolpath quality: User-assisted CAM for complex milling processes | |
| CN102495593A (en) | Optimized cutter selecting method for numerical control milling for free-form surface | |
| CN117010666B (en) | Intelligent management method and system for production workstation of automobile repair tool | |
| WO2021221916A1 (en) | Morphic manufacturing | |
| Hsieh et al. | Multi-pass progressive tool path planning in five-axis flank milling by particle swarm optimisation | |
| Tapoglou et al. | Optimal machining parameter selection based on real-time machine monitoring using IEC 61499 function blocks for use in a cloud manufacturing environment: A case study for face milling | |
| CN105867300A (en) | Reverse remachining method for large forming welded part with complex contour | |
| CN111241752A (en) | Centrifugal impeller comprehensive optimization method based on digital twinning and reinforcement learning | |
| CN114170429A (en) | Method and device for identifying tool path characteristic points |
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 |