CN116295752B - Test strength control method and system for SMT (surface mounting technology) feeding equipment - Google Patents
Test strength control method and system for SMT (surface mounting technology) feeding equipment Download PDFInfo
- Publication number
- CN116295752B CN116295752B CN202310533104.8A CN202310533104A CN116295752B CN 116295752 B CN116295752 B CN 116295752B CN 202310533104 A CN202310533104 A CN 202310533104A CN 116295752 B CN116295752 B CN 116295752B
- Authority
- CN
- China
- Prior art keywords
- feature
- vibration
- frequency domain
- decoding
- value
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01G—WEIGHING
- G01G19/00—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
- B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
- B65G27/00—Jigging conveyors
- B65G27/10—Applications of devices for generating or transmitting jigging movements
- B65G27/32—Applications of devices for generating or transmitting jigging movements with means for controlling direction, frequency or amplitude of vibration or shaking movement
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01G—WEIGHING
- G01G23/00—Auxiliary devices for weighing apparatus
- G01G23/01—Testing or calibrating of weighing apparatus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K13/00—Apparatus or processes specially adapted for manufacturing or adjusting assemblages of electric components
- H05K13/02—Feeding of components
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K13/00—Apparatus or processes specially adapted for manufacturing or adjusting assemblages of electric components
- H05K13/08—Monitoring manufacture of assemblages
- H05K13/086—Supply management, e.g. supply of components or of substrates
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Manufacturing & Machinery (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Mechanical Engineering (AREA)
- Operations Research (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
A test strength control method and system for SMT feeding equipment are disclosed. Firstly, carrying out frequency domain analysis based on Fourier transform on a vibration signal to obtain a plurality of frequency domain feature statistic values, then, passing the plurality of frequency domain feature statistic values and a waveform diagram of the vibration signal through a joint encoder to obtain a multi-mode vibration feature matrix, then, passing the multi-mode vibration feature matrix through a spatial attention module to obtain a decoding feature matrix, then, passing the decoding feature matrix through a decoder to obtain a decoding value used for representing a weight compensation value, finally, comparing a corrected weight value obtained by correcting the weight value of the feeder based on the decoding value with a preset threshold value, and controlling the force provided by the feeder based on a comparison result. Thus, the accuracy control and stability of the SMT feeding equipment force can be improved.
Description
Technical Field
The application relates to the field of intelligent control, in particular to a test force control method and a test force control system for SMT feeding equipment.
Background
In the SMT production process, the control of the test force of the feeding equipment is very important. If the force provided by the feeder is unstable or too large, the component placement is inaccurate, and therefore the product quality and the production efficiency are affected. In addition, excessive force may damage the components or the PCB, increasing production costs and maintenance costs.
Therefore, a test force control scheme for SMT feeding devices is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides a test strength control method and a test strength control system for SMT (surface mounting technology) feeding equipment. Firstly, carrying out frequency domain analysis based on Fourier transform on a vibration signal to obtain a plurality of frequency domain feature statistic values, then, passing the plurality of frequency domain feature statistic values and a waveform diagram of the vibration signal through a joint encoder to obtain a multi-mode vibration feature matrix, then, passing the multi-mode vibration feature matrix through a spatial attention module to obtain a decoding feature matrix, then, passing the decoding feature matrix through a decoder to obtain a decoding value used for representing a weight compensation value, finally, comparing a corrected weight value obtained by correcting the weight value of the feeder based on the decoding value with a preset threshold value, and controlling the force provided by the feeder based on a comparison result. Thus, the accuracy control and stability of the SMT feeding equipment force can be improved.
According to one aspect of the application, there is provided a test force control method for an SMT feeding apparatus, comprising:
acquiring a weight value of a feeder acquired by a weighing sensor and a vibration signal when the weight value of the feeder is measured;
performing frequency domain analysis based on Fourier transform on the vibration signal to obtain a plurality of frequency domain feature statistic values;
passing the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder comprising a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix;
passing the multi-mode vibration feature matrix through a spatial attention module to obtain a decoding feature matrix;
passing the decoding feature matrix through a decoder to obtain a decoded value, the decoded value being used to represent a weight compensation value;
correcting the weight value of the feeder based on the decoding value to obtain a corrected weight value; and controlling the force provided by the feeder based on the comparison between the corrected weight value and a predetermined threshold.
In the test strength control method for an SMT feeding device, the step of passing the plurality of frequency domain feature statistics values and the waveform diagram of the vibration signal through a joint encoder including a sequence encoder and an image encoder to obtain a multi-mode vibration feature matrix includes:
Inputting the plurality of frequency domain feature statistics into the sequence encoder to obtain a frequency domain feature associated feature vector;
passing the waveform diagram of the vibration signal through the image encoder to obtain a vibration waveform characteristic vector; and performing association coding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
In the test strength control method for the SMT feeding equipment, the sequence encoder is a multi-scale neighborhood feature extraction module, the multi-scale neighborhood feature extraction module comprises a first convolution layer and a second convolution layer which are parallel, and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer and the second convolution layer use one-dimensional convolution kernels with different scales.
In the test strength control method for the SMT feeding equipment, the image encoder is a convolutional neural network model serving as a filter.
In the test strength control method for SMT feeding equipment, performing association coding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix, including:
The Helmholtz type free energy factors of the vibration waveform characteristic vector and the frequency domain characteristic association characteristic vector are respectively calculated according to the following factor calculation formula:
wherein, the liquid crystal display device comprises a liquid crystal display device,a classification probability value representing the vibration waveform feature vector obtained by a classifier,/for the vibration waveform feature vector>A classification probability value representing the frequency domain feature associated feature vector obtained by a classifier,/for each of the frequency domain feature associated feature vectors>Characteristic values representing respective positions of the vibration waveform characteristic vector, +.>Characteristic values representing respective positions of the frequency domain characteristic-associated characteristic vector, and +.>Is the length of the feature vector, +.>Representing an exponential operation, ++>Represented by 2Logarithmic function of the bottom>Represents a first Helmholtz-like free energy factor, -/-, a second Helmholtz-like free energy factor, -/-, a third Helmholtz-like free energy factor, -, a fourth Helmholtz-like free energy factor, -, a second Helmh>Representing a second helmholtz-like free energy factor; the frequency domain characteristic association feature vector and the vibration waveform feature vector are weighted by taking the first Helmholtz class free energy factor and the second Helmholtz class free energy factor as weighting weights respectively to obtain an optimized frequency domain characteristic association feature vector and an optimized vibration waveform feature vector; and performing association coding on the optimized frequency domain feature association feature vector and the optimized vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
In the test force control method for the SMT feeding device, the multi-mode vibration feature matrix is passed through a spatial attention module to obtain a decoding feature matrix, including:
performing depth convolution encoding on the multi-mode vibration feature matrix by using a convolution encoding part of the spatial attention module to obtain a multi-mode vibration convolution feature map;
inputting the multi-modal vibration convolution feature map into a spatial attention portion of the spatial attention module to obtain a spatial attention map;
-passing said spatial attention map through a Softmax activation function to obtain a spatial attention profile; and calculating the position-wise dot multiplication of the spatial attention characteristic diagram and the multi-mode vibration convolution characteristic diagram to obtain the decoding characteristic matrix.
In the test strength control method for the SMT feeding device, the decoding feature matrix is passed through a decoder to obtain a decoding value, where the decoding value is used to represent a weight compensation value, and the method includes:
performing decoding regression on the decoding feature matrix by using a plurality of full-connection layers of the decoder according to a decoding formula to obtain the decoding value, wherein the decoding formula is as follows: Wherein->Is the decoding feature matrix,>is the decoded value,/->Is a weight matrix, < >>Representing a matrix multiplication.
According to another aspect of the present application, there is provided a test force control system for an SMT feeding apparatus, comprising:
the data acquisition module is used for acquiring the weight value of the feeder acquired by the weighing sensor and a vibration signal when the weight value of the feeder is measured;
the Fourier transform module is used for carrying out frequency domain analysis based on Fourier transform on the vibration signal so as to obtain a plurality of frequency domain characteristic statistical values;
the joint coding module is used for enabling the plurality of frequency domain feature statistics values and the waveform diagram of the vibration signal to pass through a joint encoder comprising a sequence encoder and an image encoder so as to obtain a multi-mode vibration feature matrix;
the spatial attention coding module is used for enabling the multi-mode vibration feature matrix to pass through the spatial attention module to obtain a decoding feature matrix;
the decoding module is used for enabling the decoding characteristic matrix to pass through a decoder to obtain a decoding value, and the decoding value is used for representing a weight compensation value;
the correction module is used for correcting the weight value of the feeder based on the decoding value to obtain a corrected weight value; and the force control module is used for controlling the force provided by the feeder based on the comparison between the corrected weight value and a preset threshold value.
In the test strength control system for the SMT feeding device, the joint coding module includes:
a sequence encoding unit, configured to input the plurality of frequency domain feature statistics values into the sequence encoder to obtain a frequency domain feature associated feature vector;
the image coding unit is used for passing the waveform diagram of the vibration signal through the image coder to obtain a vibration waveform characteristic vector; and the association coding unit is used for carrying out association coding on the frequency domain feature association characteristic vector and the vibration waveform feature vector so as to obtain the multi-mode vibration characteristic matrix.
In the test strength control system for the SMT feeding equipment, the sequence encoder is a multi-scale neighborhood feature extraction module, the multi-scale neighborhood feature extraction module comprises a first convolution layer and a second convolution layer which are parallel, and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer and the second convolution layer use one-dimensional convolution kernels with different scales.
Compared with the prior art, the test force control method and the test force control system for the SMT feeding equipment are characterized in that firstly, frequency domain analysis based on Fourier transform is conducted on vibration signals to obtain a plurality of frequency domain feature statistics values, then, the frequency domain feature statistics values and waveform diagrams of the vibration signals are combined with an encoder to obtain a multi-mode vibration feature matrix, then, the multi-mode vibration feature matrix is subjected to a spatial attention module to obtain a decoding feature matrix, then, the decoding feature matrix is subjected to a decoder to obtain a decoding value used for representing a weight compensation value, finally, corrected weight values obtained by correcting the weight values of the feeding equipment based on the decoding value are compared with a preset threshold value, and the force provided by the feeding equipment is controlled based on a comparison result. Thus, the accuracy control and stability of the SMT feeding equipment force can be improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort to a person of ordinary skill in the art. The following drawings are not intended to be drawn to scale, emphasis instead being placed upon illustrating the principles of the application.
Fig. 1 is an application scenario diagram of a test force control method for an SMT feeding device according to an embodiment of the application.
Fig. 2 is a flowchart of a test strength control method for an SMT feeding apparatus according to an embodiment of the application.
Fig. 3 is a schematic diagram of a test strength control method for an SMT feeding device according to an embodiment of the application.
Fig. 4 is a flowchart of substep S130 of the test force control method for the SMT feeding apparatus according to an embodiment of the present application.
Fig. 5 is a flowchart of substep S140 of the test force control method for the SMT feeding apparatus according to an embodiment of the present application.
Fig. 6 is a block diagram of a test force control system for an SMT feeding apparatus according to an embodiment of the application.
Fig. 7 is a block diagram of the joint coding module in the test strength control system for the SMT feeding apparatus according to an embodiment of the present application.
Description of the embodiments
The following description of the embodiments of the present application 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 application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are also within the scope of the application.
As used in the specification 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.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
Correspondingly, a test dynamics control for SMT (Surface Mounting Technique, surface mounting technology) feeding equipment includes: the weight of the feeder is measured through a weighing sensor, and the force provided by the feeder is controlled according to a preset weight threshold. However, when the load cell is used to measure the weight of the loader, the measurement result of the sensor may be affected by various factors, such as vibration, which may cause deviation of the measurement result, thereby affecting the providing force and stability of the loader.
Aiming at the technical problems, the technical conception of the application is to correct the measured weight value based on the vibration signal during weight test to obtain a weight value with higher accuracy, and in such a way, the problem of deviation caused by vibration interference of the sensor measurement result during weighing is solved, so that the accurate control and the stability improvement of the SMT feeding equipment force are realized.
Specifically, in the technical scheme of the application, firstly, the weight value of the feeder acquired by the weighing sensor and a vibration signal when the weight value of the feeder is measured are acquired. As described above, since the weight measurement result of the load cell is affected by the vibration factor, it is necessary to acquire the vibration signal at the time of measurement in addition to the weight value of the loader acquired by the load cell at the time of weighing. By analyzing and processing the vibration signals, the influence of vibration on the weighing result can be eliminated or reduced, the weighing precision is improved, and the stability and consistency of the feeder are ensured.
Then, frequency domain analysis based on Fourier transform is carried out on the vibration signal to obtain a plurality of frequency domain characteristic statistical values. Here, the vibration signal acquired by the sensor during the measurement is a time-series signal in which vibration components of a plurality of frequencies are contained. By performing a fourier transform-based frequency domain analysis of the vibration signal, the signal in the time domain may be converted into a signal in the frequency domain, and a plurality of frequency domain feature statistics, such as frequency, power spectral density, etc., may be obtained, which may provide detailed information and characteristics about the vibration signal, such as the main frequency, amplitude, phase, etc., of the vibration signal. By analyzing and processing the characteristics, the source of the vibration signal can be accurately judged so as to eliminate or reduce the interference of the vibration symmetrical weight result, thereby improving the weighing precision and the stability of the feeder.
Further, the plurality of frequency domain feature statistics and the waveform map of the vibration signal are passed through a joint encoder comprising a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix. That is, in the technical scheme of the present application, a joint encoder including a sequence encoder and an image encoder is used to combine a plurality of frequency domain feature statistics obtained from the vibration signal and a waveform diagram of the vibration signal to obtain more comprehensive and accurate vibration information. It will be appreciated by those skilled in the art that it is a technical difficulty that the plurality of frequency domain feature statistics and the waveform map of the vibration signal belong to different types of data (i.e., data of different modalities) and that information of two different modalities is organically integrated together. Based on this, in the solution of the application, the joint encoder is constructed comprising a sequence encoder and an image encoder, which is essentially a variant of the CLIP model, which is able to learn to express the links between different types of data and encode them as a vector representation.
Specifically, in the technical scheme of the application, the sequence encoder is a multi-scale neighborhood feature extraction module, wherein the multi-scale neighborhood feature extraction module comprises a first convolution layer and a second convolution layer which are parallel at a structural level, and a multi-scale feature fusion layer connected with the first convolution layer and the second convolution layer and used for capturing correlation mode features among different frequency domain feature statistics values. Meanwhile, in the technical scheme of the application, the image encoder is a deep convolutional neural network model comprising a plurality of convolutional layers, and one of ordinary skill in the art should know that the convolutional neural network model has excellent performance and characterization capability in terms of image feature extraction.
After the frequency domain feature association feature vector is obtained through the sequence encoder and the vibration waveform feature vector is obtained through the image encoder, carrying out association encoding on the frequency domain feature association feature vector and the vibration waveform feature vector so as to obtain the multi-mode vibration feature matrix.
In particular, in the technical solution of the present application, when the multi-mode vibration feature matrix is processed, the relationships between different types of data are not the same, and some information may be more important or relevant. Therefore, there is a need for an efficient method to capture relationships between different types of data and extract the most useful information. The spatial attention module may help achieve this goal. By means of the spatial attention module, the importance of different regions can be identified and emphasized in the multi-modal vibration feature matrix and useful information can be extracted therefrom. This information can be converted into a decoding feature matrix for subsequent correction and control of the force provided by the feeder.
That is, in the technical scheme of the present application, after the multi-mode vibration feature matrix is obtained, the multi-mode vibration feature matrix is passed through a spatial attention module to obtain a decoding feature matrix. And further passing the decoded feature matrix through a decoder to obtain decoded values, the decoded values being indicative of weight compensation values. And correcting the weight value of the feeder based on the decoded value to obtain a corrected weight value, and controlling the force provided by the feeder based on the comparison between the corrected weight value and a preset threshold value. By means of the method, errors and deviations can be eliminated by correcting the decoding values, so that more accurate and stable weight values are obtained, and the weight compensation values can participate in a subsequent control algorithm to realize accurate control of the force of the SMT feeding equipment and improve stability.
In the technical scheme of the application, when the frequency domain feature correlation feature vector and the vibration waveform feature vector are subjected to correlation coding to obtain the multi-mode vibration feature matrix, as the frequency domain feature correlation feature vector and the vibration waveform feature vector respectively express the channel dimension distribution of the image semantic features of the vibration signals and the sample correlation distribution of the frequency domain statistical features, no matter whether the source data and the feature distribution have significant differences, the frequency domain feature correlation feature vector and the vibration waveform feature vector have similar weak correlation distribution examples relative to a decoder in the integral feature distribution, namely, the compatibility of the integral feature distribution of the multi-mode vibration feature matrix obtained by performing correlation coding on the frequency domain feature correlation feature vector and the vibration waveform feature vector in the decoder is lower, and the accuracy of the decoding value obtained by the decoder of the subsequent decoding feature matrix is influenced.
Based on this, it is preferable to calculate the vibration waveform feature vectors separatelyAssociated feature vector with the frequency domain feature>The helmholtz-like free energy factor of (c) is specifically:
And->Respectively represent the characteristic vector of the vibration waveform +.>Associated feature vector with the frequency domain feature>Classification probability value obtained by classifier, and +.>Is the length of the feature vector.
Here, the vibration waveform feature vector may be calculated based on the helmholtz free energy formulaAssociated feature vector with the frequency domain feature>The respective feature value sets describe the energy value of the predetermined class label as the class free energy of the feature vector as a whole by using the energy value of the predetermined class label as the vibration waveform feature vector +.>Associated feature vector with the frequency domain feature>Weighting is performed to add the vibration waveform feature vector +.>Associated feature vector with the frequency domain feature>Focusing on a class-related prototype instance (prototype instance) distribution of features overlapping with a true instance (groundtruth instance) distribution in a class target domain to facilitate a classification of a waveform in the vibration waveform feature vector>Associated feature vector with the frequency domain feature>And under the condition that a similar weak correlation distribution example exists in the integral feature distribution, incremental learning is realized by carrying out ambiguity labeling on the integral feature distribution, so that the compatibility of the integral feature distribution under a similar label is improved, the accuracy of feature expression of the multi-mode vibration feature matrix is improved, and the accuracy of a decoding value obtained by a decoder by the subsequent decoding feature matrix is improved.
Fig. 1 is an application scenario diagram of a test force control method for an SMT feeding device according to an embodiment of the application. As shown in fig. 1, in this application scenario, first, a weight value (e.g., D1 shown in fig. 1) of a loader (e.g., N shown in fig. 1) acquired by a load cell (e.g., C shown in fig. 1) and a vibration signal (e.g., D2 shown in fig. 1) when the weight value of the loader is measured are acquired, then, the weight value of the loader and the vibration signal are input to a server (e.g., S shown in fig. 1) where a test force control algorithm for an SMT loader is deployed, wherein the server can process the weight value of the loader and the vibration signal using the test force control algorithm for an SMT loader to obtain a decoded value for representing a weight compensation value, then, correct the weight value of the loader based on the decoded value to obtain a corrected weight value, and finally, control the force provided by the loader based on a comparison between the corrected weight value and a predetermined threshold value.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.
Fig. 2 is a flowchart of a test strength control method for an SMT feeding apparatus according to an embodiment of the application. As shown in fig. 2, the test force control method for the SMT feeding device according to the embodiment of the application includes the steps of: s110, acquiring a weight value of a feeder acquired by a weighing sensor and a vibration signal when the weight value of the feeder is measured; s120, carrying out frequency domain analysis based on Fourier transform on the vibration signal to obtain a plurality of frequency domain feature statistic values; s130, passing the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder comprising a sequence encoder and an image encoder to obtain a multi-mode vibration feature matrix; s140, passing the multi-mode vibration feature matrix through a spatial attention module to obtain a decoding feature matrix; s150, the decoding characteristic matrix passes through a decoder to obtain a decoding value, wherein the decoding value is used for representing a weight compensation value; s160, correcting the weight value of the feeder based on the decoding value to obtain a corrected weight value; and S170, controlling the force provided by the feeder based on the comparison between the corrected weight value and a preset threshold value.
Fig. 3 is a schematic diagram of a test strength control method for an SMT feeding device according to an embodiment of the application. As shown in fig. 3, in the network architecture, firstly, a weight value of a loader acquired by a load cell and a vibration signal when measuring the weight value of the loader are acquired; then, carrying out frequency domain analysis based on Fourier transform on the vibration signal to obtain a plurality of frequency domain feature statistic values; then, passing the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder comprising a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix; then, the multi-mode vibration feature matrix passes through a spatial attention module to obtain a decoding feature matrix; then, the decoding characteristic matrix is passed through a decoder to obtain a decoding value, wherein the decoding value is used for representing a weight compensation value; then, correcting the weight value of the feeder based on the decoded value to obtain a corrected weight value; and finally, controlling the force provided by the feeder based on the comparison between the corrected weight value and a preset threshold value.
More specifically, in step S110, a weight value of the loader acquired by the load cell and a vibration signal at the time of measuring the weight value of the loader are acquired. The weight measurement result of the load cell is affected by vibration factors, so that vibration signals during measurement need to be acquired simultaneously in addition to the weight value of the feeder acquired by the load cell when weighing. By analyzing and processing the vibration signals, the influence of vibration on the weighing result can be eliminated or reduced, the weighing precision is improved, and the stability and consistency of the feeder are ensured.
More specifically, in step S120, a fourier transform-based frequency domain analysis is performed on the vibration signal to obtain a plurality of frequency domain feature statistics. The vibration signal collected by the sensor is a time series signal, which contains vibration components of a plurality of frequencies. By performing a fourier transform-based frequency domain analysis of the vibration signal, the signal in the time domain may be converted into a signal in the frequency domain, and a plurality of frequency domain feature statistics, such as frequency, power spectral density, etc., may be obtained, which may provide detailed information and characteristics about the vibration signal, such as the main frequency, amplitude, phase, etc., of the vibration signal. By analyzing and processing the characteristics, the source of the vibration signal can be accurately judged so as to eliminate or reduce the interference of the vibration symmetrical weight result, thereby improving the weighing precision and the stability of the feeder.
More specifically, in step S130, the plurality of frequency domain feature statistics and the waveform map of the vibration signal are passed through a joint encoder including a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix. That is, in the technical scheme of the present application, a joint encoder including a sequence encoder and an image encoder is used to combine a plurality of frequency domain feature statistics obtained from the vibration signal and a waveform diagram of the vibration signal to obtain more comprehensive and accurate vibration information.
Accordingly, in one specific example, as shown in fig. 4, passing the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder including a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix includes: s131, inputting the plurality of frequency domain feature statistics into the sequence encoder to obtain frequency domain feature associated feature vectors; s132, passing the waveform diagram of the vibration signal through the image encoder to obtain a vibration waveform characteristic vector; and S133, performing association coding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
Accordingly, in one specific example, the sequence encoder is a multi-scale neighborhood feature extraction module comprising a first convolution layer and a second convolution layer in parallel, and a multi-scale feature fusion layer connected to the first convolution layer and the second convolution layer, wherein the first convolution layer and the second convolution layer use one-dimensional convolution kernels having different scales. Inputting the plurality of frequency domain feature statistics into the sequence encoder to obtain a frequency domain feature associated feature vector, comprising: arranging the plurality of frequency domain feature statistics into a frequency domain input vector; performing one-dimensional convolution encoding on the frequency domain input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module to obtain a first-scale frequency domain feature vector, wherein the first convolution layer is provided with a first one-dimensional convolution kernel with a first length; performing one-dimensional convolution encoding on the frequency domain input vector by using a second convolution layer of the multi-scale neighborhood feature extraction module to obtain a second-scale frequency domain feature vector, wherein the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; and cascading the first scale frequency domain feature vector and the second scale frequency domain feature vector by using a multi-scale feature fusion layer of the multi-scale neighborhood feature extraction module to obtain the frequency domain feature associated feature vector.
Accordingly, in one specific example, the image encoder is a convolutional neural network model that acts as a filter. The convolutional neural network model has excellent performance and characterization capability in terms of image feature extraction. Passing the waveform map of the vibration signal through the image encoder to obtain a vibration waveform feature vector, comprising: and respectively carrying out two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transmission of layers by using each layer of the convolutional neural network model as a filter to output the vibration waveform feature vector by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is a waveform diagram of the vibration signal.
It should be appreciated that convolutional neural network (Convolutional Neural Network, CNN) is an artificial neural network and has wide application in the fields of image recognition and the like. The convolutional neural network may include an input layer, a hidden layer, and an output layer, where the hidden layer may include a convolutional layer, a pooling layer, an activation layer, a full connection layer, etc., where the previous layer performs a corresponding operation according to input data, outputs an operation result to the next layer, and obtains a final result after the input initial data is subjected to a multi-layer operation. The convolutional neural network model has excellent performance in the aspect of image local feature extraction by taking a convolutional kernel as a feature filtering factor, and has stronger feature extraction generalization capability and fitting capability compared with the traditional image feature extraction algorithm based on statistics or feature engineering.
Accordingly, in a specific example, performing association coding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix, including: the Helmholtz type free energy factors of the vibration waveform characteristic vector and the frequency domain characteristic association characteristic vector are respectively calculated according to the following factor calculation formula:
wherein, the liquid crystal display device comprises a liquid crystal display device,a classification probability value representing the vibration waveform feature vector obtained by a classifier,/for the vibration waveform feature vector>A classification probability value representing the frequency domain feature associated feature vector obtained by a classifier,/for each of the frequency domain feature associated feature vectors>Characteristic values representing respective positions of the vibration waveform characteristic vector, +.>Characteristic values representing respective positions of the frequency domain characteristic-associated characteristic vector, and +.>Is the length of the feature vector, +.>Representing an exponential operation, ++>Represents a logarithmic function with base 2, +.>Represents a first Helmholtz-like free energy factor, -/-, a second Helmholtz-like free energy factor, -/-, a third Helmholtz-like free energy factor, -, a fourth Helmholtz-like free energy factor, -, a second Helmh>Representing a second helmholtz-like free energy factor; the frequency domain characteristic association characteristic vector and the vibration waveform characteristic vector are weighted by taking the first Helmholtz class free energy factor and the second Helmholtz class free energy factor as weighting weights respectively to obtain an optimized frequency domain characteristic association characteristic vector and an optimized vibration Waveform feature vectors; and performing association coding on the optimized frequency domain feature association feature vector and the optimized vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
In the technical scheme of the application, when the frequency domain feature correlation feature vector and the vibration waveform feature vector are subjected to correlation coding to obtain the multi-mode vibration feature matrix, as the frequency domain feature correlation feature vector and the vibration waveform feature vector respectively express the channel dimension distribution of the image semantic features of the vibration signals and the sample correlation distribution of the frequency domain statistical features, no matter whether the source data and the feature distribution have significant differences, the frequency domain feature correlation feature vector and the vibration waveform feature vector have similar weak correlation distribution examples relative to a decoder in the integral feature distribution, namely, the compatibility of the integral feature distribution of the multi-mode vibration feature matrix obtained by performing correlation coding on the frequency domain feature correlation feature vector and the vibration waveform feature vector in the decoder is lower, and the accuracy of the decoding value obtained by the decoder of the subsequent decoding feature matrix is influenced.
Based on a Helmholtz free energy formula, the energy values of the vibration waveform feature vector and the frequency domain feature associated feature vector can be described by the free energy of the feature vector for the whole feature vector of the energy value of a preset class label, and by weighting the vibration waveform feature vector and the frequency domain feature associated feature vector by the free energy, the class associated prototype instance distribution of the feature, which has overlapping property with the true instance distribution in the class target domain, of the vibration waveform feature vector and the frequency domain feature associated feature vector can be focused, so that incremental learning can be realized by carrying out fuzzy labeling on the vibration waveform feature vector and the frequency domain feature associated feature vector under the condition that the class weak associated distribution instance exists in the whole feature distribution of the vibration waveform feature vector and the frequency domain feature associated feature vector, and the compatibility of the whole feature distribution under the class label is improved, so that the accuracy of the feature expression of the multi-mode vibration feature matrix is improved, and the accuracy of the decoding value obtained by a decoder of the follow-up decoding feature matrix is improved.
More specifically, in step S140, the multi-modal vibration feature matrix is passed through a spatial attention module to obtain a decoded feature matrix. In particular, in the technical solution of the present application, when the multi-mode vibration feature matrix is processed, the relationships between different types of data are not the same, and some information may be more important or relevant. Therefore, there is a need for an efficient method to capture relationships between different types of data and extract the most useful information. The spatial attention module may help achieve this goal. By means of the spatial attention module, the importance of different regions can be identified and emphasized in the multi-modal vibration feature matrix and useful information can be extracted therefrom. This information can be converted into a decoding feature matrix for subsequent correction and control of the force provided by the feeder.
Accordingly, in one specific example, as shown in fig. 5, passing the multi-modal vibration feature matrix through a spatial attention module to obtain a decoded feature matrix includes: s141, performing depth convolution coding on the multi-mode vibration feature matrix by using a convolution coding part of the spatial attention module to obtain a multi-mode vibration convolution feature diagram; s142, inputting the multi-mode vibration convolution characteristic diagram into a space attention part of the space attention module to obtain a space attention diagram; s143, the spatial attention is subjected to a Softmax activation function to obtain a spatial attention profile; and S144, calculating the position-wise point multiplication of the spatial attention characteristic diagram and the multi-mode vibration convolution characteristic diagram to obtain the decoding characteristic matrix.
More specifically, in step S150, the decoding feature matrix is passed through a decoder to obtain a decoded value, which is used to represent a weight compensation value. By means of the method, errors and deviations can be eliminated by correcting the decoding values, so that more accurate and stable weight values are obtained, and the weight compensation values can participate in a subsequent control algorithm to realize accurate control of the force of the SMT feeding equipment and improve stability.
Accordingly, in one specific example, one wouldThe decoding feature matrix is passed through a decoder to obtain a decoded value, the decoded value being used to represent a weight compensation value, comprising: performing decoding regression on the decoding feature matrix by using a plurality of full-connection layers of the decoder according to a decoding formula to obtain the decoding value, wherein the decoding formula is as follows:wherein->Is the decoding feature matrix,>is the decoded value,/->Is a weight matrix, < >>Representing a matrix multiplication.
More specifically, in step S160, the weight value of the loader is corrected based on the decoded value to obtain a corrected weight value.
More specifically, in step S170, the force provided by the loader is controlled based on the comparison between the corrected weight value and a predetermined threshold value.
In summary, according to the test force control method for the SMT feeding device according to the embodiment of the present application, firstly, a frequency domain analysis based on fourier transform is performed on a vibration signal to obtain a plurality of frequency domain feature statistics values, then, the plurality of frequency domain feature statistics values and a waveform diagram of the vibration signal are passed through a joint encoder to obtain a multi-mode vibration feature matrix, then, the multi-mode vibration feature matrix is passed through a spatial attention module to obtain a decoding feature matrix, then, the decoding feature matrix is passed through a decoder to obtain a decoding value for representing a weight compensation value, finally, a corrected weight value obtained by correcting the weight value of the feeding device based on the decoding value is compared with a predetermined threshold value, and the force provided by the feeding device is controlled based on the comparison result. Thus, the accuracy control and stability of the SMT feeding equipment force can be improved.
Fig. 6 is a block diagram of a test force control system 100 for an SMT feeding apparatus according to an embodiment of the application. As shown in fig. 6, a test force control system 100 for SMT feeding apparatus according to an embodiment of the present application includes: a data acquisition module 110, configured to acquire a weight value of the loader acquired by the load cell and a vibration signal when the weight value of the loader is measured; a fourier transform module 120, configured to perform fourier transform-based frequency domain analysis on the vibration signal to obtain a plurality of frequency domain feature statistics values; a joint encoding module 130, configured to pass the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder including a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix; a spatial attention encoding module 140, configured to pass the multi-mode vibration feature matrix through a spatial attention module to obtain a decoded feature matrix; a decoding module 150, configured to pass the decoding feature matrix through a decoder to obtain a decoded value, where the decoded value is used to represent a weight compensation value; the correction module 160 is configured to correct the weight value of the loader based on the decoded value to obtain a corrected weight value; and a force control module 170, configured to control a force provided by the feeder based on a comparison between the corrected weight value and a predetermined threshold.
In one example, as shown in fig. 7, in the test strength control system 100 for SMT feeding apparatus, the joint encoding module 130 includes: a sequence encoding unit 131, configured to input the plurality of frequency domain feature statistics into the sequence encoder to obtain a frequency domain feature associated feature vector; an image encoding unit 132 for passing the waveform diagram of the vibration signal through the image encoder to obtain a vibration waveform feature vector; and a correlation encoding unit 133, configured to perform correlation encoding on the frequency domain feature correlation feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
In one example, in the test strength control system 100 for SMT feeding devices, the sequence encoder is a multi-scale neighborhood feature extraction module, and the multi-scale neighborhood feature extraction module includes a first convolution layer and a second convolution layer in parallel, and a multi-scale feature fusion layer connected to the first convolution layer and the second convolution layer, where the first convolution layer and the second convolution layer use one-dimensional convolution kernels with different scales.
In one example, in the test dynamics control system 100 for SMT feeding devices described above, the image encoder is a convolutional neural network model as a filter.
In one example, in the test force control system 100 for SMT feeding apparatus, performing association encoding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix, includes: the Helmholtz type free energy factors of the vibration waveform characteristic vector and the frequency domain characteristic association characteristic vector are respectively calculated according to the following factor calculation formula:
wherein, the liquid crystal display device comprises a liquid crystal display device,a classification probability value representing the vibration waveform feature vector obtained by a classifier,/for the vibration waveform feature vector>A classification probability value representing the frequency domain feature associated feature vector obtained by a classifier,/for each of the frequency domain feature associated feature vectors>Characteristic values representing respective positions of the vibration waveform characteristic vector, +.>Characteristic values representing respective positions of the frequency domain characteristic-associated characteristic vector, and +.>Is the length of the feature vector, +.>Representing an exponential operation, ++>Represents a logarithmic function with base 2, +.>Represents a first Helmholtz-like free energy factor, -/-, a second Helmholtz-like free energy factor, -/-, a third Helmholtz-like free energy factor, -, a fourth Helmholtz-like free energy factor, -, a second Helmh>Representing a second helmholtz-like free energy factor; the frequency domain characteristic association feature vector and the vibration waveform feature vector are weighted by taking the first Helmholtz class free energy factor and the second Helmholtz class free energy factor as weighting weights respectively to obtain an optimized frequency domain characteristic association feature vector and an optimized vibration waveform feature vector; and performing association coding on the optimized frequency domain feature association feature vector and the optimized vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
In one example, in the test strength control system 100 for SMT feeding devices, the spatial attention encoding module is configured to: performing depth convolution encoding on the multi-mode vibration feature matrix by using a convolution encoding part of the spatial attention module to obtain a multi-mode vibration convolution feature map; inputting the multi-modal vibration convolution feature map into a spatial attention portion of the spatial attention module to obtain a spatial attention map; -passing said spatial attention map through a Softmax activation function to obtain a spatial attention profile; and calculating the position-wise point multiplication of the spatial attention characteristic diagram and the multi-mode vibration convolution characteristic diagram to obtain the decoding characteristic matrix.
In one example, in the test force control system 100 for SMT feeding apparatus, the decoding module 150 is configured to: using the decoderPerforming a decoding regression on the decoding feature matrix with a decoding formula to obtain the decoded value, wherein the decoding formula is:wherein->Is the decoding feature matrix,>is the decoded value,/->Is a weight matrix, < >>Representing a matrix multiplication.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective modules in the above-described test force control system 100 for SMT feeding apparatuses have been described in detail in the above description of the test force control method for SMT feeding apparatuses with reference to fig. 1 to 5, and thus, repetitive descriptions thereof will be omitted.
As described above, the test intensity control system 100 for the SMT feeding apparatus according to the embodiment of the application may be implemented in various wireless terminals, for example, a server or the like having a test intensity control algorithm for the SMT feeding apparatus. In one example, the test effort control system 100 for SMT feeding devices according to an embodiment of the present application may be integrated into the wireless terminal as one software module and/or hardware module. For example, the test effort control system 100 for SMT feeding devices 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 test strength control system 100 for SMT feeding devices may also be one of a plurality of hardware modules of the wireless terminal.
Alternatively, in another example, the test effort control system 100 for the SMT feeding apparatus and the wireless terminal may be separate devices, and the test effort control system 100 for the SMT feeding apparatus may be connected to the wireless terminal through a wired and/or wireless network, and transmit the interaction information according to an agreed data format.
According to another aspect of the present application there is also provided a non-volatile computer readable storage medium having stored thereon computer readable instructions which when executed by a computer can perform a method as described above.
Program portions of the technology may be considered to be "products" or "articles of manufacture" in the form of executable code and/or associated data, embodied or carried out by a computer readable medium. A tangible, persistent storage medium may include any memory or storage used by a computer, processor, or similar device or related module. Such as various semiconductor memories, tape drives, disk drives, or the like, capable of providing storage functionality for software.
All or a portion of the software may sometimes communicate over a network, such as the internet or other communication network. Such communication may load software from one computer device or processor to another. For example: a hardware platform loaded from a server or host computer of the video object detection device to a computer environment, or other computer environment implementing the system, or similar functioning system related to providing information needed for object detection. Thus, another medium capable of carrying software elements may also be used as a physical connection between local devices, such as optical, electrical, electromagnetic, etc., propagating through cable, optical cable, air, etc. Physical media used for carrier waves, such as electrical, wireless, or optical, may also be considered to be software-bearing media. Unless limited to a tangible "storage" medium, other terms used herein to refer to a computer or machine "readable medium" mean any medium that participates in the execution of any instructions by a processor.
The application uses specific words to describe embodiments of the application. Reference to "a first/second embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic is associated with at least one embodiment of the application. Thus, it should be emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various positions in this specification are not necessarily referring to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of the application may be combined as suitable.
Furthermore, those skilled in the art will appreciate that the various aspects of the application are illustrated and described in the context of a number of patentable categories or circumstances, including any novel and useful procedures, machines, products, or materials, or any novel and useful modifications thereof. Accordingly, aspects of the application may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.) or by a combination of hardware and software. The above hardware or software may be referred to as a "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of the application may take the form of a computer product, comprising computer-readable program code, embodied in one or more computer-readable media.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The foregoing is illustrative of the present invention and is not to be construed as limiting thereof. Although a few exemplary embodiments of this invention have been described, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the novel teachings and advantages of this invention. Accordingly, all such modifications are intended to be included within the scope of this invention as defined in the following claims. It is to be understood that the foregoing is illustrative of the present invention and is not to be construed as limited to the specific embodiments disclosed, and that modifications to the disclosed embodiments, as well as other embodiments, are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.
Claims (10)
1. The test strength control method for the SMT feeding equipment is characterized by comprising the following steps of:
acquiring a weight value of a feeder acquired by a weighing sensor and a vibration signal when the weight value of the feeder is measured;
performing frequency domain analysis based on Fourier transform on the vibration signal to obtain a plurality of frequency domain feature statistic values;
passing the plurality of frequency domain feature statistics and the waveform diagram of the vibration signal through a joint encoder comprising a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix;
passing the multi-mode vibration feature matrix through a spatial attention module to obtain a decoding feature matrix;
passing the decoding feature matrix through a decoder to obtain a decoded value, the decoded value being used to represent a weight compensation value;
correcting the weight value of the feeder based on the decoding value to obtain a corrected weight value; and controlling the force provided by the feeder based on the comparison between the corrected weight value and a predetermined threshold.
2. The method for controlling test strength of SMT feeding apparatus according to claim 1, wherein passing the plurality of frequency domain feature statistics and the waveform pattern of the vibration signal through a joint encoder including a sequence encoder and an image encoder to obtain a multi-modal vibration feature matrix, comprises:
Inputting the plurality of frequency domain feature statistics into the sequence encoder to obtain a frequency domain feature associated feature vector;
passing the waveform diagram of the vibration signal through the image encoder to obtain a vibration waveform characteristic vector; and performing association coding on the frequency domain feature association feature vector and the vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
3. The method of claim 2, wherein the sequence encoder is a multi-scale neighborhood feature extraction module comprising a first convolution layer and a second convolution layer in parallel, and a multi-scale feature fusion layer connected to the first convolution layer and the second convolution layer, wherein the first convolution layer and the second convolution layer use one-dimensional convolution kernels having different scales.
4. A test strength control method for an SMT feeding apparatus according to claim 3, wherein said image encoder is a convolutional neural network model as a filter.
5. The method for controlling test strength of SMT feeding equipment according to claim 4, wherein performing association encoding on said frequency domain feature association feature vector and said vibration waveform feature vector to obtain said multi-modal vibration feature matrix comprises:
The Helmholtz type free energy factors of the vibration waveform characteristic vector and the frequency domain characteristic association characteristic vector are respectively calculated according to the following factor calculation formula:
wherein (1)>A classification probability value representing the vibration waveform feature vector obtained by a classifier,/for the vibration waveform feature vector>Representing classification probability values obtained by the classifier of the frequency domain feature associated feature vectors,characteristic values representing respective positions of the vibration waveform characteristic vector, +.>Characteristic values representing respective positions of the frequency domain characteristic-associated characteristic vector, and +.>Is the length of the feature vector, +.>Representing an exponential operation, ++>Represents a logarithmic function with base 2, +.>Represents a first Helmholtz-like free energy factor, -/-, a second Helmholtz-like free energy factor, -/-, a third Helmholtz-like free energy factor, -, a fourth Helmholtz-like free energy factor, -, a second Helmh>Representing a second helmholtz-like free energy factor;
the frequency domain characteristic association feature vector and the vibration waveform feature vector are weighted by taking the first Helmholtz class free energy factor and the second Helmholtz class free energy factor as weighting weights respectively to obtain an optimized frequency domain characteristic association feature vector and an optimized vibration waveform feature vector; and performing association coding on the optimized frequency domain feature association feature vector and the optimized vibration waveform feature vector to obtain the multi-mode vibration feature matrix.
6. The method for controlling test strength of an SMT feeding apparatus according to claim 5, wherein passing the multi-modal vibration feature matrix through a spatial attention module to obtain a decoded feature matrix comprises:
performing depth convolution encoding on the multi-mode vibration feature matrix by using a convolution encoding part of the spatial attention module to obtain a multi-mode vibration convolution feature map;
inputting the multi-modal vibration convolution feature map into a spatial attention portion of the spatial attention module to obtain a spatial attention map;
-passing said spatial attention map through a Softmax activation function to obtain a spatial attention profile; and calculating the position-wise dot multiplication of the spatial attention characteristic diagram and the multi-mode vibration convolution characteristic diagram to obtain the decoding characteristic matrix.
7. The method for controlling test strength of an SMT feeding apparatus according to claim 6, wherein the decoding feature matrix is passed through a decoder to obtain a decoded value, the decoded value being used to represent a weight compensation value, comprising:
performing decoding regression on the decoding feature matrix by using a plurality of full-connection layers of the decoder according to a decoding formula to obtain the decoding value, wherein the decoding formula is as follows: Wherein->Is the decoding feature matrix,>is the decoded value,/->Is a weight matrix, < >>Representing a matrix multiplication.
8. A test dynamics control system for SMT material loading equipment, characterized in that includes:
the data acquisition module is used for acquiring the weight value of the feeder acquired by the weighing sensor and a vibration signal when the weight value of the feeder is measured;
the Fourier transform module is used for carrying out frequency domain analysis based on Fourier transform on the vibration signal so as to obtain a plurality of frequency domain characteristic statistical values;
the joint coding module is used for enabling the plurality of frequency domain feature statistics values and the waveform diagram of the vibration signal to pass through a joint encoder comprising a sequence encoder and an image encoder so as to obtain a multi-mode vibration feature matrix;
the spatial attention coding module is used for enabling the multi-mode vibration feature matrix to pass through the spatial attention module to obtain a decoding feature matrix;
the decoding module is used for enabling the decoding characteristic matrix to pass through a decoder to obtain a decoding value, and the decoding value is used for representing a weight compensation value;
the correction module is used for correcting the weight value of the feeder based on the decoding value to obtain a corrected weight value; and the force control module is used for controlling the force provided by the feeder based on the comparison between the corrected weight value and a preset threshold value.
9. The test strength control system for an SMT feeding apparatus according to claim 8, wherein said joint encoding module comprises:
a sequence encoding unit, configured to input the plurality of frequency domain feature statistics values into the sequence encoder to obtain a frequency domain feature associated feature vector;
the image coding unit is used for passing the waveform diagram of the vibration signal through the image coder to obtain a vibration waveform characteristic vector; and the association coding unit is used for carrying out association coding on the frequency domain feature association characteristic vector and the vibration waveform feature vector so as to obtain the multi-mode vibration characteristic matrix.
10. The test strength control system for an SMT feeding device according to claim 9, wherein said sequence encoder is a multi-scale neighborhood feature extraction module comprising a first convolution layer and a second convolution layer in parallel, and a multi-scale feature fusion layer connected to said first convolution layer and said second convolution layer, wherein said first convolution layer and said second convolution layer use one-dimensional convolution kernels having different scales.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310533104.8A CN116295752B (en) | 2023-05-12 | 2023-05-12 | Test strength control method and system for SMT (surface mounting technology) feeding equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310533104.8A CN116295752B (en) | 2023-05-12 | 2023-05-12 | Test strength control method and system for SMT (surface mounting technology) feeding equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116295752A CN116295752A (en) | 2023-06-23 |
| CN116295752B true CN116295752B (en) | 2023-08-15 |
Family
ID=86830851
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310533104.8A Active CN116295752B (en) | 2023-05-12 | 2023-05-12 | Test strength control method and system for SMT (surface mounting technology) feeding equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116295752B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116563291B (en) * | 2023-07-11 | 2023-09-22 | 深圳市蓝眼科技有限公司 | SMT intelligent error-proofing feeding detector |
| CN116734750B (en) * | 2023-08-15 | 2023-10-27 | 山西锦烁生物医药科技有限公司 | Intelligent detection method and system for thickness of ice layer of ice rink based on optical fiber sensor |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08110261A (en) * | 1994-10-11 | 1996-04-30 | Yamato Scale Co Ltd | Apparatus for mass measurement and apparatus for weight measurement |
| JPH09113346A (en) * | 1995-10-16 | 1997-05-02 | Ishida Co Ltd | Method and apparatus for measurement |
| CN103499387A (en) * | 2013-09-26 | 2014-01-08 | 北京空间飞行器总体设计部 | Micro-vibration signal processing method |
| CN103954344A (en) * | 2014-05-12 | 2014-07-30 | 浙江大学 | Acceleration sensor based dynamic weighing signal real-time compensation device and method |
| CN108363374A (en) * | 2018-03-29 | 2018-08-03 | 中山市鸿菊自动化设备制造有限公司 | A kind of SMT materials supply method and system |
| CN208314553U (en) * | 2018-03-29 | 2019-01-01 | 中山市鸿菊自动化设备制造有限公司 | A kind of SMT material delivery system |
| CN109298457A (en) * | 2018-11-20 | 2019-02-01 | 中国科学技术大学 | Vibration noise correction-compensation method suitable for atomic interference gravimeter |
| CN110352000A (en) * | 2019-06-17 | 2019-10-18 | 江苏华志珹智能科技有限公司 | A kind of high stability can bilateral feeding arch (bridge) type SMT chip mounter |
| CN215004043U (en) * | 2021-06-01 | 2021-12-03 | 深圳市蓝眼科技有限公司 | Automatic regulating and controlling mechanism of SMT charging equipment measured value degree of depth |
| CN215500341U (en) * | 2021-07-26 | 2022-01-11 | 苏州帕兰提尼智能科技股份有限公司 | Flexible circuit board chip mounter with automatic calibration function |
-
2023
- 2023-05-12 CN CN202310533104.8A patent/CN116295752B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08110261A (en) * | 1994-10-11 | 1996-04-30 | Yamato Scale Co Ltd | Apparatus for mass measurement and apparatus for weight measurement |
| JPH09113346A (en) * | 1995-10-16 | 1997-05-02 | Ishida Co Ltd | Method and apparatus for measurement |
| CN103499387A (en) * | 2013-09-26 | 2014-01-08 | 北京空间飞行器总体设计部 | Micro-vibration signal processing method |
| CN103954344A (en) * | 2014-05-12 | 2014-07-30 | 浙江大学 | Acceleration sensor based dynamic weighing signal real-time compensation device and method |
| CN108363374A (en) * | 2018-03-29 | 2018-08-03 | 中山市鸿菊自动化设备制造有限公司 | A kind of SMT materials supply method and system |
| CN208314553U (en) * | 2018-03-29 | 2019-01-01 | 中山市鸿菊自动化设备制造有限公司 | A kind of SMT material delivery system |
| CN109298457A (en) * | 2018-11-20 | 2019-02-01 | 中国科学技术大学 | Vibration noise correction-compensation method suitable for atomic interference gravimeter |
| CN110352000A (en) * | 2019-06-17 | 2019-10-18 | 江苏华志珹智能科技有限公司 | A kind of high stability can bilateral feeding arch (bridge) type SMT chip mounter |
| CN215004043U (en) * | 2021-06-01 | 2021-12-03 | 深圳市蓝眼科技有限公司 | Automatic regulating and controlling mechanism of SMT charging equipment measured value degree of depth |
| CN215500341U (en) * | 2021-07-26 | 2022-01-11 | 苏州帕兰提尼智能科技股份有限公司 | Flexible circuit board chip mounter with automatic calibration function |
Non-Patent Citations (1)
| Title |
|---|
| 物联网驱动的SMT车间制造执行系统研究;张浩;马瑞;葛世伦;刘平;;现代制造工程(第02期);第46-54页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116295752A (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116295752B (en) | Test strength control method and system for SMT (surface mounting technology) feeding equipment | |
| Gamboa et al. | Validation of the rapid detection approach for enhancing the electronic nose systems performance, using different deep learning models and support vector machines | |
| US11715190B2 (en) | Inspection system, image discrimination system, discrimination system, discriminator generation system, and learning data generation device | |
| US11416717B2 (en) | Classification model building apparatus and classification model building method thereof | |
| WO2020189189A1 (en) | Inspection device and method | |
| CN109886337B (en) | Depth measurement learning method and system based on self-adaptive sampling | |
| CN114219306B (en) | Method, apparatus, medium for establishing welding quality detection model | |
| CN116805061B (en) | Leakage event judging method based on optical fiber sensing | |
| CN116577677A (en) | Discharging test system and method for retired power battery | |
| US20210240167A1 (en) | Method and apparatus for detecting abnormality of manufacturing facility | |
| CN117171696B (en) | Sensor production monitoring method and system based on Internet of things | |
| CN117434429B (en) | Chip stability testing method and related device | |
| CN116563291B (en) | SMT intelligent error-proofing feeding detector | |
| CN117152152B (en) | Production management system and method for detection kit | |
| CN113554624A (en) | Anomaly detection method and device and computer storage medium | |
| CN117191816A (en) | Method and device for detecting surface defects of electronic component based on multispectral fusion | |
| CN117233682A (en) | Quick calibration system of balance bridge | |
| CN116312860B (en) | Agricultural product soluble solid matter prediction method based on supervised transfer learning | |
| CN115266583B (en) | Ambient light filtering method, system, computer equipment and computer readable storage medium | |
| CN116502899A (en) | Risk rating model generation method, device and storage medium based on artificial intelligence | |
| RU2458397C1 (en) | Method of searching for and recognising objects on digital images | |
| CN114399669A (en) | Target detection method and device | |
| CN116539952B (en) | Millimeter wave therapeutic apparatus power intelligent sampling method and system based on Internet of things | |
| CN113033757A (en) | Method, apparatus and computer-readable storage medium for testing operator precision in neural networks | |
| Chao et al. | Reading pointer meter through one stage end-to-end deep regression |
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 |