CN117102988A - Centerless grinding machine and control method thereof - Google Patents
Centerless grinding machine and control method thereof Download PDFInfo
- Publication number
- CN117102988A CN117102988A CN202311333621.7A CN202311333621A CN117102988A CN 117102988 A CN117102988 A CN 117102988A CN 202311333621 A CN202311333621 A CN 202311333621A CN 117102988 A CN117102988 A CN 117102988A
- Authority
- CN
- China
- Prior art keywords
- temperature
- time
- training
- feeding speed
- sequence
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 230000003247 decreasing effect Effects 0.000 claims abstract description 19
- 238000004458 analytical method Methods 0.000 claims abstract description 13
- 239000013598 vector Substances 0.000 claims description 127
- 238000012549 training Methods 0.000 claims description 86
- 230000004044 response Effects 0.000 claims description 49
- 230000011218 segmentation Effects 0.000 claims description 10
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000012512 characterization method Methods 0.000 claims description 4
- 230000001419 dependent effect Effects 0.000 claims description 4
- 238000002075 inversion recovery Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 abstract description 18
- 238000003754 machining Methods 0.000 abstract description 11
- 238000009825 accumulation Methods 0.000 abstract description 10
- 238000012545 processing Methods 0.000 description 25
- 230000008569 process Effects 0.000 description 21
- 230000006870 function Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 239000010453 quartz Substances 0.000 description 9
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 9
- 230000008859 change Effects 0.000 description 8
- 239000000463 material Substances 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 235000019580 granularity Nutrition 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 238000011084 recovery Methods 0.000 description 3
- 239000004065 semiconductor Substances 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012300 Sequence Analysis Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000003746 surface roughness Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B5/00—Machines or devices designed for grinding surfaces of revolution on work, including those which also grind adjacent plane surfaces; Accessories therefor
- B24B5/18—Machines or devices designed for grinding surfaces of revolution on work, including those which also grind adjacent plane surfaces; Accessories therefor involving centreless means for supporting, guiding, floating or rotating work
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B47/00—Drives or gearings; Equipment therefor
- B24B47/20—Drives or gearings; Equipment therefor relating to feed movement
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B49/00—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation
- B24B49/006—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation taking regard of the speed
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B49/00—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation
- B24B49/14—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation taking regard of the temperature during grinding
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B51/00—Arrangements for automatic control of a series of individual steps in grinding a workpiece
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Constituent Portions Of Griding Lathes, Driving, Sensing And Control (AREA)
Abstract
A centerless grinder and a control method thereof are disclosed. Firstly, acquiring feeding speed values of a plurality of preset time points and grinding temperature values of the preset time points in a preset time period, then, carrying out time sequence cooperative analysis on the feeding speed values of the preset time points and the grinding temperature values of the preset time points to obtain feeding speed-temperature time sequence correlation characteristics, and then, determining that the feeding speed value of the current time point should be increased or decreased based on the feeding speed-temperature time sequence correlation characteristics. In this way, the feeding speed can be adaptively adjusted based on the time sequence distribution characteristics of the temperature data, so that the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed is avoided, the machining efficiency is improved, and the machining quality is ensured.
Description
Technical Field
The present disclosure relates to the field of centerless grinding machines, and more particularly, to a centerless grinding machine and a control method thereof.
Background
The centerless grinder is a machine tool for machining shaft parts, is widely applied to industries such as machine manufacturing, automobile manufacturing, aerospace and the like, and is used for machining various precise shaft parts. The centerless grinder is characterized in that in the machining process, the workpiece does not need to be supported in the center, but is kept stable through the inertia of the workpiece. The machining mode can avoid workpiece deformation and errors caused by the center support, and improves machining precision and efficiency.
In the semiconductor industry, centerless grinding machines are commonly used to process quartz materials. Quartz has the characteristics of high hardness, high thermal stability, low thermal expansion coefficient and the like, so that the quartz is widely applied to the fields of optical devices, crystal oscillators, sensors and the like in semiconductor manufacturing. The centerless grinder can carry out high-precision grinding processing on the quartz material so as to meet the requirements of the semiconductor industry on quartz parts.
However, in the conventional method, a fixed feeding speed is generally adopted for processing the quartz material in the process of using the centerless grinder, and dynamic adjustment cannot be performed according to actual processing conditions. In the process of processing quartz materials, a great amount of heat is generated in the grinding process due to the high hardness and thermal stability. The fixed feed rate in the conventional control scheme cannot effectively control the grinding heat accumulation phenomenon, and heat accumulation in the grinding process can be caused, so that the processing quality and the surface roughness of the workpiece are affected. Also, the fixed feed rate in conventional control schemes may not maximize the performance of the grinding apparatus, resulting in insufficient grinding efficiency. Particularly in quartz material processing under high precision requirements, the fixed feed rate may not meet the processing requirements, and higher processing efficiency and quality may not be achieved.
Accordingly, an optimized control scheme for centerless grinding machines is desired.
Disclosure of Invention
In view of the above, the present application provides a centerless grinding machine and a control method thereof, which can adaptively adjust a feeding speed based on a time sequence distribution characteristic of temperature data, and avoid a problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed, thereby improving processing efficiency and ensuring processing quality.
According to an aspect of the present disclosure, there is provided a control method of a centerless grinder, including:
acquiring feeding speed values of a plurality of preset time points in a preset time period and grinding temperature values of the preset time points;
performing time sequence collaborative analysis on the feeding speed values of the plurality of preset time points and the grinding temperature values of the plurality of preset time points to obtain feeding speed-temperature time sequence correlation characteristics; and
based on the feed speed-temperature timing correlation characteristic, it is determined that the feed speed value at the current point in time should be increased or decreased.
According to another aspect of the present disclosure, there is provided a centerless grinder wherein the centerless grinder operates in a control method of the centerless grinder as previously described.
According to an embodiment of the present disclosure, a feeding speed value at a plurality of predetermined time points and a grinding temperature value at the plurality of predetermined time points within a predetermined period of time are first acquired, then, a time-series cooperative analysis is performed on the feeding speed value at the plurality of predetermined time points and the grinding temperature value at the plurality of predetermined time points to obtain a feeding speed-temperature time-series correlation feature, and then, it is determined that the feeding speed value at the current time point should be increased or decreased based on the feeding speed-temperature time-series correlation feature. In this way, the feeding speed can be adaptively adjusted based on the time sequence distribution characteristics of the temperature data, so that the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed is avoided, the machining efficiency is improved, and the machining quality is ensured.
Compared with the prior art, the application provides the centerless grinding machine and the control method thereof, which can adaptively adjust the feeding speed based on the time sequence distribution characteristics of the temperature data, avoid the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed, thereby improving the processing efficiency and ensuring the processing quality.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 shows a flowchart of a control method of a centerless grinder according to an embodiment of the present disclosure.
Fig. 2 shows a schematic architecture diagram of a control method of a centerless grinder according to an embodiment of the present disclosure.
Fig. 3 shows a flowchart of substep S120 of the control method of the centerless grinder according to an embodiment of the present disclosure.
Fig. 4 shows a flowchart of training steps further included in the control method of the centerless grinder according to an embodiment of the present disclosure.
Fig. 5 shows a block diagram of a control system of a centerless grinder according to an embodiment of the present disclosure.
Fig. 6 shows an application scenario diagram of a control method of a centerless grinder according to an embodiment of the present disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this disclosure and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following detailed description in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order not to obscure the present disclosure.
Example 1:
aiming at the technical problems, the technical concept of the disclosure is to collect temperature data of a grinding process based on a temperature sensor in the grinding process, and introduce a data processing and analyzing algorithm at the rear end to perform time sequence analysis of the temperature data of the grinding process, so as to adaptively adjust the feeding speed based on time sequence distribution characteristics of the temperature data, avoid the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed, thereby improving the processing efficiency and ensuring the processing quality.
Fig. 1 shows a flowchart of a control method of a centerless grinder according to an embodiment of the present disclosure. Fig. 2 shows a schematic architecture diagram of a control method of a centerless grinder according to an embodiment of the present disclosure. As shown in fig. 1 and 2, a control method of a centerless grinder according to an embodiment of the present disclosure includes the steps of: s110, acquiring feeding speed values of a plurality of preset time points in a preset time period and grinding temperature values of the preset time points; s120, carrying out time sequence collaborative analysis on the feeding speed values of the plurality of preset time points and the grinding temperature values of the plurality of preset time points to obtain feeding speed-temperature time sequence correlation characteristics; and S130, determining that the feeding speed value of the current time point should be increased or decreased based on the feeding speed-temperature time sequence correlation characteristic.
Specifically, in the technical scheme of the present disclosure, first, feed speed values at a plurality of predetermined time points within a predetermined period of time and grinding temperature values at the plurality of predetermined time points are acquired. Then, considering that the feeding speed value and the grinding temperature value have a dynamic change rule in the time dimension and have mutual influence, the combined action of the feeding speed value and the grinding temperature value has influence on the processing efficiency and quality of the quartz material. Accordingly, in the technical solution of the present disclosure, it is necessary to perform real-time control of the feed speed based on the time-series distribution correlation characteristic between the feed speed value and the grinding temperature value. Specifically, first, the feeding speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points are respectively arranged into a feeding speed time sequence input vector and a grinding temperature time sequence input vector according to a time dimension, so that time sequence distribution information of the feeding speed values and the grinding temperature values in the time dimension is respectively integrated.
Then, the time-by-time element response between the feeding speed time sequence input vector and the grinding temperature time sequence input vector is calculated again to obtain a feeding speed-temperature response time sequence input vector, so that the time-by-time response association between the feeding speed value and the grinding temperature value is established, and the time sequence characteristic distribution information of the feeding speed value and the grinding temperature value is conveniently and effectively captured and described.
Then, when extracting the time sequence correlation characteristic between the feeding speed value and the grinding temperature value, in order to better capture the time sequence correlation relation between the feeding speed value and the grinding temperature value in the time dimension, in the technical scheme of the disclosure, vector segmentation is further performed on the feeding speed-temperature response time sequence input vector so as to obtain a sequence of feeding speed-temperature response local time sequence input vector. By vector slicing the feed rate-temperature response time series input vector, it can be decomposed into a plurality of shorter local time series input vectors, each corresponding to a time period in the grinding process, and containing time series correlation distribution information between the feed rate and the temperature in that time period. Therefore, the model can pay more attention to local time sequence correlation change and detail between the feeding speed and the grinding temperature in the grinding process, and is helpful for better understanding of the correlation dynamic change between the feeding speed and the grinding temperature in the grinding process and capturing time sequence correlation characteristic information between the feeding speed and the grinding temperature.
Further, the sequence of the feeding speed-temperature response local time sequence input vector is subjected to feature mining in a time sequence feature extractor based on a one-dimensional convolution layer, so that local time sequence related feature distribution information related to the feeding speed value and the grinding temperature value in each local time period in the preset time period is extracted, namely, the local time sequence cooperative change feature of the feeding speed value and the grinding temperature value in the time dimension is obtained, and the sequence of the feeding speed-temperature response local time sequence feature vector is obtained.
Further, it is also considered that since the feeding speed value and the grinding temperature value have not only a local time-series correlation in each local time period but also overall correlation characteristic distribution information in the entire predetermined time period. That is, the local time series cooperative correlation characteristic information between the feed speed value and the grinding temperature value has a time series overall correlation relationship throughout a predetermined period. Therefore, in order to capture global time sequence related information between the feeding speed and the grinding temperature so as to better understand dynamic change and mutual influence between the feeding speed and the grinding temperature, in the technical scheme of the disclosure, the sequence of the feeding speed-temperature response local time sequence characteristic vector is further encoded in a context time sequence encoder based on a converter so as to extract local time sequence cooperative change characteristics between the feeding speed value and the grinding temperature value in each local time period based on the time sequence global context related characteristic information, thereby obtaining the feeding speed-temperature global time sequence context related characteristic vector. In particular, the converter-based context sequential encoder is a sequence modeling method that uses a self-attention mechanism to encode and associative model elements in a sequence. In this process, each local timing feature vector interacts with other local feature vectors to obtain global context-related information.
Accordingly, as shown in fig. 3, performing a time-series collaborative analysis on the feeding speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points to obtain a feeding speed-temperature time-series correlation feature, includes: s121, arranging the feeding speed values of the plurality of preset time points and the grinding temperature values of the plurality of preset time points into a feeding speed time sequence input vector and a grinding temperature time sequence input vector according to a time dimension respectively; s122, calculating a time-element-by-time response between the feeding speed time sequence input vector and the grinding temperature time sequence input vector to obtain a feeding speed-temperature response time sequence input vector; s123, vector segmentation is carried out on the feeding speed-temperature response time sequence input vector so as to obtain a sequence of feeding speed-temperature response local time sequence input vector; s124, extracting features of the sequence of the feeding speed-temperature response local time sequence input vectors through a time sequence feature extractor based on a deep neural network model to obtain the sequence of the feeding speed-temperature response local time sequence feature vectors; and S125, carrying out full-time sequence association coding on the sequence of the feeding speed-temperature response local time sequence characteristic vectors to obtain a feeding speed-temperature global time sequence context association characteristic vector as the feeding speed-temperature time sequence association characteristic. It should be understood that the purpose of step S121 is to arrange the time-series data in order of time for subsequent analysis and processing; the purpose of step S122 is to analyze the correlation between the feed rate and the grinding temperature, determining their variation over time; the purpose of step S123 is to segment the time series data into a plurality of local time series input vectors for feature extraction and analysis at each local time series; the purpose of step S124 is to extract key features in each local timing input vector to capture timing correlations between feed speed and grinding temperature; the purpose of step S125 is to integrate the local timing feature vector into a global timing context correlation feature vector to provide more comprehensive timing correlation information between feed speed and grinding temperature.
More specifically, in step S124, the timing feature extractor based on the deep neural network model is a timing feature extractor based on a one-dimensional convolution layer. It should be noted that a one-dimensional convolutional layer is a common layer type in deep neural networks, and is used for processing one-dimensional sequence data, and has an important role in feature extraction and pattern recognition of the sequence data. The one-dimensional convolution layer performs local perception on the input sequence by sliding a convolution kernel (one-dimensional filter), thereby extracting local features at different positions. The parameters of the convolution kernel are learned through back propagation in the training process, so that the network can automatically learn important features in the input sequence, and the one-dimensional convolution layer can extract local features with different sizes from the input sequence through convolution operation. These features can capture patterns, trends, and important structures in the sequence, facilitating the processing and analysis of subsequent tasks. One-dimensional convolution layers typically use pooling operations (e.g., maximum pooling or average pooling) to reduce the size of feature maps, thereby reducing the dimensionality of the data, which helps reduce the amount of parameters and computational complexity of the model, while retaining critical feature information. The one-dimensional convolution layer typically applies an activation function, such as ReLU (Rectified Linear Unit), after the convolution operation to introduce a non-linear mapping, which can enhance the expressive power of the model, enabling it to learn more complex sequence patterns and associations. A one-dimensional convolutional layer based timing feature extractor can automatically learn timing patterns and associated features in an input sequence to extract high-level feature representations useful for subsequent tasks. It shows good effects in many time series data analysis tasks, such as speech recognition, natural language processing, time series prediction, etc.
More specifically, in step S125, performing full-time-series correlation encoding on the sequence of the feed speed-temperature response local-time-series feature vectors to obtain a feed speed-temperature global-time-series context-correlation feature vector as the feed speed-temperature time-series correlation feature, including: passing the sequence of feed rate-temperature responsive local timing feature vectors through a converter-based context timing encoder to obtain the feed rate-temperature global timing context-dependent feature vector. It is worth mentioning that the converter (transducer) is a deep learning model based on self-attention mechanism, and the core idea of the converter is to capture global dependency in the sequence by using self-attention mechanism without relying on the traditional Recurrent Neural Network (RNN) or Convolutional Neural Network (CNN). The self-attention mechanism allows the model to establish a global context association in the input sequence between the encoder and decoder, thereby better capturing long-range dependencies in the sequence. In step S125, the context-based sequential encoder of the converter is used to perform full-time correlated encoding of the sequence of feed rate-temperature responsive local sequential feature vectors. This means that the converter model takes as input the entire sequence of local time series feature vectors and learns the correlations between the individual time steps in the sequence by a self-attention mechanism. Through the coding process of the converter, the model can perform global context correlation coding on the local time sequence characteristic vector of each time step, so that global time sequence correlation characteristics between the feeding speed and the grinding temperature are captured. These global timing context-dependent feature vectors can provide more information that helps understand and analyze the timing relationship between feed speed and grinding temperature, resulting in timing-dependent features of feed speed-temperature. The advantage of the converter is that it is able to handle long sequences and capture global dependencies, making it a powerful tool for handling time series data.
And then, the feeding speed-temperature global time sequence context associated characteristic vector is further passed through a classifier to obtain a classification result, wherein the classification result is used for indicating that the feeding speed value of the current time point is increased or decreased. That is, the full-time cooperative correlation change characteristic information between the feed speed value and the grinding temperature value is subjected to classification processing, whereby the feed speed value control at the current time point is performed. In this way, the feeding speed can be adaptively adjusted based on the time sequence distribution characteristics of the temperature data, so that the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed is avoided, the machining efficiency is improved, and the machining quality is ensured.
Accordingly, determining, based on the feed speed-temperature timing correlation characteristic, whether the feed speed value at the current point in time should be increased or decreased includes: the feed speed-temperature global time sequence context associated feature vector is passed through a classifier to obtain a classification result, which is used to indicate whether the feed speed value at the current point in time should be increased or decreased.
More specifically, passing the feed speed-temperature global time series context associated feature vector through a classifier to obtain a classification result, wherein the classification result is used for representing that the feed speed value of the current time point should be increased or decreased, and the classification result comprises the following steps: performing full-connection coding on the feeding speed-temperature global time sequence context associated feature vector by using a full-connection layer of the classifier to obtain a coded classification feature vector; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
That is, in the technical solution of the present disclosure, the tag of the classifier includes that the feeding speed value of the current time point should be increased (first tag) and that the feeding speed value of the current time point should be decreased (second tag), wherein the classifier determines to which classification tag the feeding speed-temperature global time-series context-associated feature vector belongs through a soft maximum function. It is worth noting that the first tag p1 and the second tag p2 here do not contain the concept of human setting, and in fact, during the training process, the computer model does not have the concept of "the feeding speed value of the current time point should be increased or should be decreased", which is only two kinds of classification tags and the probability that the output feature is under these two kinds of classification tags, i.e. the sum of p1 and p2 is one. Therefore, the classification result that the feeding speed value of the current time point should be increased or decreased is actually a classification probability distribution converted from classifying the tag into a classification conforming to the natural law, and the physical meaning of the natural probability distribution of the tag is essentially used instead of the language text meaning that the feeding speed value of the current time point should be increased or decreased.
It should be appreciated that the role of the classifier is to learn the classification rules and classifier using a given class, known training data, and then classify (or predict) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
Further, in the technical solution of the present disclosure, the control method of the centerless grinder further includes a training step: for training the one-dimensional convolutional layer-based timing feature extractor, the converter-based context timing encoder, and the classifier. It should be understood that in the technical solution of the present disclosure, the training step is used to train the one-dimensional convolution layer-based time sequence feature extractor, the converter-based context time sequence encoder and the classifier, and this training process is aimed at enabling the model to automatically learn, through a large amount of sample data, the feature representation and the association relationship in the input sequence, so as to realize the control of the centerless grinder. The training steps include: 1. and (3) feature learning: by training the time sequence feature extractor, the model can learn key features in the input sequence, the one-dimensional convolution layer converts the input sequence into advanced feature representation, modes and structures in the sequence are captured, and in the training process, the model adjusts parameters of the convolution layer through a back propagation algorithm, so that the model can better extract useful features. 2. Context coding: by training the context sequential encoder, the model can learn global context correlation features in the sequence, the converter model encodes each time step in the sequence using a self-attention mechanism, capturing the correlation between time steps, and during the training process, the model learns the ability to better model the sequence context by optimizing the objective function. 3. Classification decision: the ability to correlate the extracted features with the control targets of the centerless grinder can be learned by the model by training the classifier, which maps the input feature representations to corresponding control decisions, such as adjusting feed speed or monitoring grinding temperature, where the model optimizes the parameters of the classifier by minimizing classification errors to enable it to accurately make classification decisions. Through the training step, the model can gradually optimize the parameters of the model, the accuracy and performance of the control of the centerless grinder are improved, and a large amount of sample data used in the training process can help the model to fully learn the characteristics and association relation of the input sequence, so that the effect and the robustness of the control method are improved.
More specifically, as shown in fig. 4, the training step includes: s210, acquiring training data, wherein the training data comprises training feeding speed values of a plurality of preset time points in a preset time period and training grinding temperature values of the preset time points, and a real value that the feeding speed value of the current time point should be increased or decreased; s220, training feeding speed values at the plurality of preset time points and training grinding temperature values at the plurality of preset time points are respectively arranged into training feeding speed time sequence input vectors and training grinding temperature time sequence input vectors according to a time dimension; s230, calculating a time element-by-time element response between the training feeding speed time sequence input vector and the training grinding temperature time sequence input vector to obtain a training feeding speed-temperature response time sequence input vector; s240, vector segmentation is carried out on the training feeding speed-temperature response time sequence input vector so as to obtain a sequence of training feeding speed-temperature response local time sequence input vector; s250, passing the sequence of training feed speed-temperature response local time sequence input vectors through the time sequence feature extractor based on the one-dimensional convolution layer to obtain a sequence of training feed speed-temperature response local time sequence feature vectors; s260, passing the sequence of training feed speed-temperature response local time sequence feature vectors through the context time sequence encoder based on the converter to obtain training feed speed-temperature global time sequence context correlation feature vectors; s270, passing the training feeding speed-temperature global time sequence context correlation feature vector through the classifier to obtain a classification loss function value; and S280, training the one-dimensional convolutional layer-based time sequence feature extractor, the converter-based context time sequence encoder and the classifier based on the classification loss function value and through gradient descent direction propagation, wherein in each round of iteration of the training, feature precision alignment based on dimension representation and inversion type recovery is carried out on the training feeding speed-temperature global time sequence context association feature vector.
In particular, in the technical scheme of the application, when the sequence of the training feeding speed-temperature response local time sequence input vectors is obtained through the time sequence feature extractor based on the one-dimensional convolution layer, each training feeding speed-temperature response local time sequence feature vector can express the local time sequence associated feature of the feeding speed-temperature response in the time point granularity based on the scale of the one-dimensional convolution kernel in the segmentation time domain, and after the sequence of the training feeding speed-temperature response local time sequence feature vector passes through the context time sequence encoder based on the converter, the context associated code based on the segmentation time sequence granularity in the global time domain can be further obtained, but in consideration of the imbalance of feature distribution in each segmentation time domain, the obtained overall feature distribution of the training feeding speed-temperature global time sequence context associated feature vector also has time sequence associated feature representation deviation in different time domain granularities in the global time domain, so that when the classification regression is carried out through the classifier, the multi-time sequence associated feature representation of the training feeding speed-temperature context associated feature vector has a problem of associating the time sequence context associated feature vector, and the training feeding speed-global time sequence context associated feature vector has a problem of associating accuracy.
Thus, in the training process, the applicant of the present application performs feature precision alignment based on dimension characterization and inversion recovery on the training feeding speed-temperature global time sequence context associated feature vector, for example, denoted as V, specifically expressed as: in each iteration of the training, performing feature precision alignment based on dimension characterization and inversion recovery on the training feeding speed-temperature global time sequence context associated feature vector by using the following optimization formula;
wherein, the optimization formula is:
vi'=V0L1+(V0L)2×vi+ α×(V0×vivi∈Vvi)
where vi is the feature value of the ith position of the training feed speed-temperature global time series context associated feature vector V, V0 represents the zero norm of the training feed speed-temperature global time series context associated feature vector V, L is the length of the training feed speed-temperature global time series context associated feature vector V, and α is a weight override parameter, vi' is the feature value of the ith position of the optimized training feed speed-temperature global time series context associated feature vector.
Here, for the contradiction of precision between high-dimensional feature space coding of parameter time sequence features and multi-granularity context feature association editing based on time sequence association dimension, the feature precision alignment based on dimension representation and inversion type recovery is generated by regarding multi-granularity context feature association editing as inversion type embedded generation of high-dimensional feature space coding distributed by parameter time sequence features, sparse distribution balance of scale representation is equipped for feature values serving as coding representation, inversion type recovery of association details is performed based on vector counting, so that self-adaptive alignment of precision difference in a training process is realized, and training effect of the training feeding speed-temperature global time sequence context association feature vector in classification regression training is improved through a classifier. Therefore, the feeding speed can be adaptively adjusted based on the time sequence change condition of the temperature data, and the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed is avoided, so that the processing efficiency is improved, and the processing quality is ensured.
In summary, according to the control method of the centerless grinding machine according to the embodiment of the disclosure, the feeding speed can be adaptively adjusted based on the time sequence distribution characteristics of the temperature data, so that the problem of grinding heat accumulation or insufficient grinding efficiency caused by too high or too low feeding speed is avoided, and the processing efficiency is improved and the processing quality is ensured.
Example 2:
further, in an embodiment of the present disclosure, there is also provided a centerless grinder, wherein the centerless grinder operates in a control method of the centerless grinder as described in any one of the foregoing. Further, the centerless grinder further includes: a grinding head; the workpiece clamping device is used for fixing and clamping a workpiece; a feed system for controlling the feed speed and direction of the workpiece; and, a control system.
Fig. 5 shows a block diagram of a control system 100 of a centerless grinder according to an embodiment of the present disclosure. As shown in fig. 5, a control system 100 of a centerless grinder according to an embodiment of the present disclosure includes: a data acquisition module 110, configured to acquire feeding speed values at a plurality of predetermined time points and grinding temperature values at the plurality of predetermined time points within a predetermined time period; a timing cooperative analysis module 120 for performing timing cooperative analysis on the feeding speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points to obtain feeding speed-temperature timing correlation characteristics; and, a feed rate control module 130 for determining, based on the feed rate-temperature timing correlation characteristics, whether the feed rate value at the current point in time should be increased or decreased.
In one possible implementation, the timing collaborative analysis module 120 includes: an input vector arrangement unit configured to arrange the feed speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points into a feed speed timing input vector and a grinding temperature timing input vector, respectively, according to a time dimension; a time-by-time element response unit for calculating a time-by-time element response between the feed speed timing input vector and the grinding temperature timing input vector to obtain a feed speed-temperature response timing input vector; the vector segmentation unit is used for carrying out vector segmentation on the feeding speed-temperature response time sequence input vector so as to obtain a sequence of feeding speed-temperature response local time sequence input vector; a feature extraction unit, configured to perform feature extraction on the sequence of feeding speed-temperature response local time sequence input vectors by using a time sequence feature extractor based on a deep neural network model, so as to obtain a sequence of feeding speed-temperature response local time sequence feature vectors; and the full-time sequence associated coding unit is used for carrying out full-time sequence associated coding on the sequence of the feeding speed-temperature response local time sequence characteristic vectors so as to obtain a feeding speed-temperature global time sequence context associated characteristic vector as the feeding speed-temperature time sequence associated characteristic.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the control system 100 of the centerless grinder described above have been described in detail in the above description of the control method of the centerless grinder with reference to fig. 1 to 4, and thus, repetitive descriptions thereof will be omitted.
As described above, the control system 100 of the centerless grinder according to the embodiment of the present disclosure may be implemented in various wireless terminals, such as a server or the like having a control algorithm of the centerless grinder. In one possible implementation, the control system 100 of the centerless grinder according to an embodiment of the present disclosure may be integrated into the wireless terminal as one software module and/or hardware module. For example, the control system 100 of the centerless grinder may be a software module in the operating system of the wireless terminal or may be an application developed for the wireless terminal; of course, the control system 100 of the centerless grinder may also be one of a number of hardware modules of the wireless terminal.
Alternatively, in another example, the control system 100 of the centerless grinder and the wireless terminal may be separate devices, and the control system 100 of the centerless grinder may be connected to the wireless terminal through a wired and/or wireless network and transmit interactive information in a agreed data format.
Fig. 6 shows an application scenario diagram of a control method of a centerless grinder according to an embodiment of the present disclosure. As shown in fig. 6, in this application scenario, first, feed speed values (e.g., D1 illustrated in fig. 6) at a plurality of predetermined time points and grinding temperature values (e.g., D2 illustrated in fig. 6) at the plurality of predetermined time points within a predetermined period of time are acquired, and then the feed speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points are input to a server (e.g., S illustrated in fig. 6) where a control algorithm of a centerless grinder is deployed, wherein the server is capable of processing the feed speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points using the control algorithm of the centerless grinder to obtain a classification result indicating that the feed speed value at the current time point should be increased or decreased.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (8)
1. A control method of a centerless grinder, comprising:
acquiring feeding speed values of a plurality of preset time points in a preset time period and grinding temperature values of the preset time points;
performing time sequence collaborative analysis on the feeding speed values of the plurality of preset time points and the grinding temperature values of the plurality of preset time points to obtain feeding speed-temperature time sequence correlation characteristics; and
determining, based on the feed speed-temperature timing correlation characteristic, whether a feed speed value at a current point in time should be increased or decreased;
wherein the performing a time-series collaborative analysis on the feed speed values at the plurality of predetermined time points and the grinding temperature values at the plurality of predetermined time points to obtain a feed speed-temperature time-series correlation feature includes:
arranging the feeding speed values of the plurality of preset time points and the grinding temperature values of the plurality of preset time points into a feeding speed time sequence input vector and a grinding temperature time sequence input vector according to a time dimension respectively;
calculating a time-by-time element response between the feed rate timing input vector and the grinding temperature timing input vector to obtain a feed rate-temperature response timing input vector;
vector segmentation is carried out on the feeding speed-temperature response time sequence input vector so as to obtain a sequence of feeding speed-temperature response local time sequence input vector;
extracting features of the sequence of the feeding speed-temperature response local time sequence input vectors through a time sequence feature extractor based on a deep neural network model to obtain the sequence of the feeding speed-temperature response local time sequence feature vectors; and
and performing full-time sequence correlation coding on the sequence of the feeding speed-temperature response local time sequence characteristic vectors to obtain a feeding speed-temperature global time sequence context correlation characteristic vector as the feeding speed-temperature time sequence correlation characteristic.
2. The control method of a centerless grinder according to claim 1, wherein the deep neural network model based timing feature extractor is a one-dimensional convolutional layer based timing feature extractor.
3. The control method of a centerless grinder according to claim 2, wherein full-time-series associative encoding the sequence of feed rate-temperature responsive local-time-series feature vectors to obtain a feed rate-temperature global-time-series context-associated feature vector as the feed rate-temperature time-series associated feature includes:
passing the sequence of feed rate-temperature responsive local timing feature vectors through a converter-based context timing encoder to obtain the feed rate-temperature global timing context-dependent feature vector.
4. A control method of a centerless grinder according to claim 3 wherein determining whether the feed rate value at the current point in time should be increased or decreased based on the feed rate-temperature timing correlation feature includes:
the feed speed-temperature global time sequence context associated feature vector is passed through a classifier to obtain a classification result, which is used to indicate whether the feed speed value at the current point in time should be increased or decreased.
5. The method of controlling a centerless grinder according to claim 4, further comprising the step of training: for training the one-dimensional convolutional layer-based timing feature extractor, the converter-based context timing encoder, and the classifier.
6. The method of claim 5, wherein the training step comprises:
acquiring training data, wherein the training data comprises training feeding speed values of a plurality of preset time points in a preset time period and training grinding temperature values of the preset time points, and a real value that the feeding speed value of the current time point should be increased or decreased;
the training feeding speed values of the plurality of preset time points and the training grinding temperature values of the plurality of preset time points are respectively arranged into a training feeding speed time sequence input vector and a training grinding temperature time sequence input vector according to the time dimension;
calculating a time-by-time element response between the training feed rate timing input vector and the training grinding temperature timing input vector to obtain a training feed rate-temperature response timing input vector;
vector segmentation is carried out on the training feeding speed-temperature response time sequence input vector so as to obtain a sequence of training feeding speed-temperature response local time sequence input vector;
passing the sequence of training feed rate-temperature response local time sequence input vectors through the one-dimensional convolution layer based time sequence feature extractor to obtain a sequence of training feed rate-temperature response local time sequence feature vectors;
passing the sequence of training feed rate-temperature responsive local timing feature vectors through the converter-based context timing encoder to obtain a training feed rate-temperature global timing context associated feature vector;
passing the training feed speed-temperature global time sequence context associated feature vector through the classifier to obtain a classification loss function value; and
training the one-dimensional convolutional layer-based timing feature extractor, the converter-based context timing encoder, and the classifier based on the classification loss function values and by gradient descent direction propagation, wherein, in each iteration of the training, feature accuracy alignment based on dimension characterization and inversion recovery is performed on the training feed speed-temperature global timing context correlation feature vector.
7. The method of claim 6, wherein in each iteration of the training, feature accuracy alignment based on dimension characterization and inversion recovery is performed on the training feed speed-temperature global timing context correlation feature vector with the following optimization formula;
wherein, the optimization formula is:
vi'=V0L1+(V0L)2×vi+ α×(V0×vivi∈Vvi)
where vi is the feature value of the ith position of the training feed speed-temperature global time series context associated feature vector V, V0 represents the zero norm of the training feed speed-temperature global time series context associated feature vector V, L is the length of the training feed speed-temperature global time series context associated feature vector V, and α is a weight override parameter, vi' is the feature value of the ith position of the optimized training feed speed-temperature global time series context associated feature vector.
8. A centerless grinder operating in accordance with the control method of the centerless grinder of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311333621.7A CN117102988B (en) | 2023-10-16 | 2023-10-16 | Centerless grinding machine and control method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311333621.7A CN117102988B (en) | 2023-10-16 | 2023-10-16 | Centerless grinding machine and control method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117102988A true CN117102988A (en) | 2023-11-24 |
| CN117102988B CN117102988B (en) | 2024-01-05 |
Family
ID=88800400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311333621.7A Active CN117102988B (en) | 2023-10-16 | 2023-10-16 | Centerless grinding machine and control method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117102988B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4043767A (en) * | 1975-09-20 | 1977-08-23 | Minoru Suda | Centerless grinding machine using tangential-feed method |
| US6669532B1 (en) * | 1997-12-24 | 2003-12-30 | Toyota Jidosha Kabushiki Kaisha | Device and method for cool air cooling type machining |
| CN113618633A (en) * | 2020-05-06 | 2021-11-09 | 湖南大学 | Grinding speed optimization method for dry grinding centerless grinding machine |
| CN115374700A (en) * | 2022-08-17 | 2022-11-22 | 南昌大学 | Polishing process parameter self-adaption method based on normal contact force change |
| CN116714204A (en) * | 2023-06-05 | 2023-09-08 | 江西佳音王文化科技有限公司 | Injection molding mechanism and molding control method for motor turntable |
| CN116772944A (en) * | 2023-08-25 | 2023-09-19 | 克拉玛依市燃气有限责任公司 | Intelligent monitoring system and method for gas distribution station |
| CN116765925A (en) * | 2023-06-12 | 2023-09-19 | 深圳市捷辉创科技有限公司 | CNC-based die cutting machining control system and method thereof |
| CN116832588A (en) * | 2023-08-25 | 2023-10-03 | 湖北鼎信成套设备有限公司 | Acid regeneration flue gas purifying device and method thereof |
-
2023
- 2023-10-16 CN CN202311333621.7A patent/CN117102988B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4043767A (en) * | 1975-09-20 | 1977-08-23 | Minoru Suda | Centerless grinding machine using tangential-feed method |
| US6669532B1 (en) * | 1997-12-24 | 2003-12-30 | Toyota Jidosha Kabushiki Kaisha | Device and method for cool air cooling type machining |
| CN113618633A (en) * | 2020-05-06 | 2021-11-09 | 湖南大学 | Grinding speed optimization method for dry grinding centerless grinding machine |
| CN115374700A (en) * | 2022-08-17 | 2022-11-22 | 南昌大学 | Polishing process parameter self-adaption method based on normal contact force change |
| CN116714204A (en) * | 2023-06-05 | 2023-09-08 | 江西佳音王文化科技有限公司 | Injection molding mechanism and molding control method for motor turntable |
| CN116765925A (en) * | 2023-06-12 | 2023-09-19 | 深圳市捷辉创科技有限公司 | CNC-based die cutting machining control system and method thereof |
| CN116772944A (en) * | 2023-08-25 | 2023-09-19 | 克拉玛依市燃气有限责任公司 | Intelligent monitoring system and method for gas distribution station |
| CN116832588A (en) * | 2023-08-25 | 2023-10-03 | 湖北鼎信成套设备有限公司 | Acid regeneration flue gas purifying device and method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 邓朝晖;陶能如;唐浩;万林林;刘伟;: "磨削加工仿真预报系统的研究现状及发展趋势", 金刚石与磨料磨具工程, no. 03, pages 64 - 68 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117102988B (en) | 2024-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104657744B (en) | A kind of multi-categorizer training method and sorting technique based on non-determined Active Learning | |
| CN110580496A (en) | Deep migration learning system and method based on entropy minimization | |
| CN116832588B (en) | Acid regeneration flue gas purifying device and method thereof | |
| CN111633467A (en) | Cutter wear state monitoring method based on one-dimensional depth convolution automatic encoder | |
| CN112987664A (en) | Flow shop scheduling method based on deep reinforcement learning | |
| CN116760772B (en) | Control system and method for converging flow divider | |
| CN113723438A (en) | Classification model calibration | |
| KR20220133998A (en) | Systems, methods and media for manufacturing processes | |
| CN117689278B (en) | Construction quality intelligent management system and method | |
| CN116204774A (en) | Cutter abrasion stability prediction method based on hierarchical element learning | |
| CN112580798A (en) | Intelligent early warning method for equipment based on multi-input multi-output ResNet | |
| CN117072740A (en) | Automatic control system and method for pneumatic regulating valve | |
| CN117308077A (en) | Water supply control system for evaporator of nuclear power unit of reactor | |
| CN117575685A (en) | Data analysis early warning system and method | |
| CN117748496B (en) | Distributed photovoltaic energy storage regulation and control system and method based on intelligent prediction | |
| CN118014041A (en) | Training method and device for power equipment energy consumption prediction model | |
| CN117102988B (en) | Centerless grinding machine and control method thereof | |
| CN118192447A (en) | Control optimization method and system for intelligent manufacturing process | |
| CN117951646A (en) | Data fusion method and system based on edge cloud | |
| Liu et al. | Focusformer: Focusing on what we need via architecture sampler | |
| CN116992954A (en) | UMAP data dimension reduction-based similarity measurement transfer learning method | |
| CN117070741A (en) | Control method and system of pickling line | |
| CN114372181A (en) | Intelligent planning method for equipment production based on multi-mode data | |
| CN117067042B (en) | Grinder and control method thereof | |
| CN113743593A (en) | Neural network quantization method, system, storage medium and terminal |
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 |