CN114355234A - Intelligent quality detection method and system for power module - Google Patents
Intelligent quality detection method and system for power module Download PDFInfo
- Publication number
- CN114355234A CN114355234A CN202111664611.2A CN202111664611A CN114355234A CN 114355234 A CN114355234 A CN 114355234A CN 202111664611 A CN202111664611 A CN 202111664611A CN 114355234 A CN114355234 A CN 114355234A
- Authority
- CN
- China
- Prior art keywords
- power supply
- appearance
- quality detection
- data set
- sound
- 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.)
- Withdrawn
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 166
- 238000000034 method Methods 0.000 claims abstract description 44
- 230000009467 reduction Effects 0.000 claims abstract description 32
- 238000012795 verification Methods 0.000 claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims abstract description 12
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims description 50
- 238000011156 evaluation Methods 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 9
- 238000000513 principal component analysis Methods 0.000 claims description 7
- 238000007689 inspection Methods 0.000 claims 4
- 238000003860 storage Methods 0.000 description 17
- 230000000694 effects Effects 0.000 description 11
- 230000008569 process Effects 0.000 description 11
- 238000003466 welding Methods 0.000 description 8
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 239000011324 bead Substances 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000007789 sealing Methods 0.000 description 5
- 230000003287 optical effect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 230000032683 aging Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000002035 prolonged effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000003925 brain function Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013506 data mapping Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009713 electroplating Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 229910000679 solder Inorganic materials 0.000 description 1
- 238000005476 soldering Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Landscapes
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The invention provides an intelligent quality detection method and system for a power module, wherein the method comprises the following steps: carrying out sound collection on the first power supply module through a sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions; carrying out noise reduction and filtering on the first sound data set to obtain a second sound data set; obtaining a feature set of a second sound dataset; inputting the characteristic set into a power quality detection model to obtain a first detection result; acquiring first image information through an image acquisition device, wherein the first image information is appearance image information of a first power supply module; performing feature extraction analysis on the first image information to obtain a first appearance feature set; obtaining a first check factor according to the first appearance characteristic set; and verifying the first detection result according to the first verification factor to obtain a second detection result.
Description
Technical Field
The invention relates to the technical field related to intelligent manufacturing devices, in particular to an intelligent quality detection method and system for a power module.
Background
Power supplies are devices that provide power to electronic devices, all of which are provided with power supplies. The quality of the power supply is very important for normal operation of the electronic equipment, and the service life of the power supply is prolonged to a certain time, so that the aging can occur, and the operation performance of the electronic equipment is influenced.
The detection of the power supply in the existing equipment is mainly realized by detecting whether the power supply is conducted or not and detecting the electricity storage performance, so that whether the power supply is aged or not is judged, and whether the power supply needs to be maintained or replaced or not is judged, so that the normal operation of the electronic equipment is maintained.
However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the inventors of the present application find that the above-mentioned technology has at least the following technical problems:
the detection method for the power supply in the prior art is mainly embodied in whether the power supply in the electronic equipment is conducted or not and power, but other modules in the electronic equipment also influence the performance of the power supply, and the existing detection method cannot carry out quantifiable quality detection on the power supply and has the technical problems of inaccurate and not intelligent detection results.
Disclosure of Invention
The embodiment of the application provides an intelligent quality detection method and system for a power module, and aims to solve the technical problems that in the prior art, a detection method for a power supply is mainly embodied in whether the power supply in an electronic device is conducted or not and power, but other modules in the electronic device can also influence the performance of the power supply, the existing detection method cannot carry out quantifiable quality detection on the power supply, and the detection result is inaccurate and not intelligent enough.
In view of the foregoing problems, the embodiments of the present application provide an intelligent quality detection method and system for a power module.
In a first aspect of the embodiments of the present application, an intelligent quality detection method for a power module is provided, where the method is applied to an intelligent quality detection system for a power module, the system includes a sound sensor and an image capture device, and the method includes: carrying out sound collection on a first power supply module through a sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions; carrying out noise reduction and filtering on the first sound data set to obtain a second sound data set; obtaining a feature set of the second sound dataset; inputting the characteristic set into a power quality detection model to obtain a first detection result; acquiring first image information through an image acquisition device, wherein the first image information is appearance image information of the first power supply module; performing feature extraction analysis on the first image information to obtain a first appearance feature set; obtaining a first check factor according to the first appearance characteristic set; and verifying the first detection result according to the first verification factor to obtain a second detection result.
In a second aspect of the embodiments of the present application, an intelligent quality detection system for a power module is provided, where the system includes: the first obtaining unit is used for carrying out sound collection on a first power supply module through a sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions; a first processing unit, configured to perform noise reduction filtering on the first sound data set to obtain a second sound data set; a second obtaining unit configured to obtain a feature set of the second sound dataset; the second processing unit is used for inputting the feature set into a power quality detection model to obtain a first detection result; a third obtaining unit, configured to obtain first image information through an image acquisition device, where the first image information is appearance image information of the first power module; a third processing unit, configured to perform feature extraction analysis on the first image information to obtain a first appearance feature set; a fourth processing unit, configured to obtain a first check factor according to the first appearance feature set; and the fifth processing unit is used for verifying the first detection result according to the first verification factor to obtain a second detection result.
In a third aspect of the embodiments of the present application, an intelligent quality detection system for a power module is provided, including: a processor coupled to a memory for storing a program that, when executed by the processor, causes a system to perform the steps of the method according to the first aspect.
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
according to the embodiment of the application, sound data sets of a power supply module under different vibration conditions are collected through a sound sensor, the sound data sets are filtered, then a feature set of the filtered sound data sets is obtained, and the feature set is input into a power supply quality detection model, so that a first detection result can be obtained; and then, an image acquisition device is adopted to obtain image information of the power supply module, feature extraction analysis is carried out to obtain a first check factor, and the first detection result is checked to obtain a second checked detection result. According to the embodiment of the application, the power supply is detected and noise is reduced through the vibration of the power supply modules with different frequencies and amplitudes, the welding condition of the tin beads welded inside the power supply module and the stability condition of other structures can be obtained, characteristic analysis is carried out by combining images of the power supply modules, the sealing and appearance or other power supply model information is verified, the quality detection result after verification is obtained, a method for quantitatively detecting the quality of the power supply modules is established, the power supply quality can be detected in a multi-dimensional and comprehensive mode, the reason or the position of the power supply module which generates faults can be known when the fact that the power supply module can be conducted to provide the power supply is known, and the technical effect of comprehensively, accurately and intelligently detecting the quality of the power supply module is achieved.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
Fig. 1 is a schematic flowchart of an intelligent quality detection method for a power module according to an embodiment of the present disclosure;
fig. 2 is a schematic flow chart illustrating an error evaluation method adopted in an intelligent quality detection method for a power module according to an embodiment of the present disclosure;
fig. 3 is a schematic diagram of an error evaluation method of skew classification in an intelligent quality detection method of a power module according to an embodiment of the present disclosure;
fig. 4 is a schematic flowchart illustrating a dimension reduction process performed on a first appearance feature set in an intelligent quality detection method for a power module according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an intelligent quality detection system of a power module according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of an exemplary electronic device according to an embodiment of the present application.
Description of reference numerals: a first obtaining unit 11, a first processing unit 12, a second obtaining unit 13, a second processing unit 14, a third obtaining unit 15, a third processing unit 16, a fourth processing unit 17, a fifth processing unit 18, an electronic device 300, a memory 301, a processor 302, a communication interface 303, and a bus architecture 304.
Detailed Description
The embodiment of the application provides an intelligent quality detection method and system for a power module, and aims to solve the technical problems that in the prior art, a detection method for a power supply is mainly embodied in whether the power supply in an electronic device is conducted or not and power, but other modules in the electronic device can also influence the performance of the power supply, the existing detection method cannot carry out quantifiable quality detection on the power supply, and the detection result is inaccurate and not intelligent enough.
According to the embodiment of the application, sound data sets of a power supply module under different vibration conditions are collected through a sound sensor, the sound data sets are filtered, then a feature set of the filtered sound data sets is obtained, and the feature set is input into a power supply quality detection model, so that a first detection result can be obtained; and then, an image acquisition device is adopted to obtain image information of the power supply module, feature extraction analysis is carried out to obtain a first check factor, and the first detection result is checked to obtain a second checked detection result. According to the embodiment of the application, the power supply is detected and noise is reduced through the vibration of the power supply modules with different frequencies and amplitudes, the welding condition of the tin beads welded inside the power supply module and the stability condition of other structures can be obtained, characteristic analysis is carried out by combining images of the power supply modules, the sealing and appearance or other power supply model information is verified, the quality detection result after verification is obtained, a method for quantitatively detecting the quality of the power supply modules is established, the power supply quality can be detected in a multi-dimensional and comprehensive mode, the reason or the position of the power supply module which generates faults can be known when the fact that the power supply module can be conducted to provide the power supply is known, and the technical effect of comprehensively, accurately and intelligently detecting the quality of the power supply module is achieved.
Summary of the application
Power supplies are devices that provide power to electronic devices, all of which are provided with power supplies. The quality of the power supply is very important for normal operation of the electronic equipment, and the service life of the power supply is prolonged to a certain time, so that the aging can occur, and the operation performance of the electronic equipment is influenced. The detection of the power supply in the existing equipment is mainly realized by detecting whether the power supply is conducted or not and detecting the electricity storage performance, so that whether the power supply is aged or not is judged, and whether the power supply needs to be maintained or replaced or not is judged, so that the normal operation of the electronic equipment is maintained. The detection method for the power supply in the prior art is mainly embodied in whether the power supply in the electronic equipment is conducted or not and power, but other modules in the electronic equipment also influence the performance of the power supply, and the existing detection method cannot carry out quantifiable quality detection on the power supply and has the technical problems of inaccurate and not intelligent detection results.
In view of the above technical problems, the technical solution provided by the present application has the following general idea:
carrying out sound collection on a first power supply module through a sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions; carrying out noise reduction and filtering on the first sound data set to obtain a second sound data set; obtaining a feature set of the second sound dataset; inputting the characteristic set into a power quality detection model to obtain a first detection result; acquiring first image information through an image acquisition device, wherein the first image information is appearance image information of the first power supply module; performing feature extraction analysis on the first image information to obtain a first appearance feature set; obtaining a first check factor according to the first appearance characteristic set; and verifying the first detection result according to the first verification factor to obtain a second detection result.
Having described the basic principles of the present application, the following embodiments will be described in detail and fully with reference to the accompanying drawings, it being understood that the embodiments described are only some embodiments of the present application, and not all embodiments of the present application, and that the present application is not limited to the exemplary embodiments described herein. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application. It should be further noted that, for the convenience of description, only some but not all of the elements relevant to the present application are shown in the drawings.
Example one
As shown in fig. 1, an embodiment of the present application provides an intelligent quality detection method for a power module, where the method is applied to an intelligent quality detection system for a power module, where the system includes an acoustic sensor and an image capture device, and the method includes:
s100: carrying out sound collection on a first power supply module through the sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions;
specifically, the first power module is a module for providing power provided on any electronic device or mobile electronic device, and the first power module may be any power supply in the prior art, for example: a PC power supply, a rectification power supply, a customized power supply, a heating power supply, a welding power supply/arc power supply, an electroplating power supply, a switching power supply, an inverter power supply, an alternating current stabilized power supply, a direct current stabilized power supply, a DC/DC power supply, a communication power supply, a module power supply, a variable frequency power supply and the like.
The sound sensor is any device for detecting and knowing sound in the surrounding environment in the prior art, and the sound sensor in the embodiment of the present application is not limited to one device, and may include a sensor for receiving sound waves in the surrounding environment, displaying an image of sound vibration, a sensor for detecting sound intensity, and the like.
The first sound data set is a sound data set obtained by the first power module through detection of the sound sensor, and illustratively, the first sound data set includes a waveform, a wavelength, a frequency, an intensity, and the like of sound obtained through detection. The first sound data set is obtained by detecting under different vibration conditions, and specifically, the first power module can be fixed on the detection platform through fixing the first power module, vibrate the first power module according to different frequencies and amplitudes, and then detect and receive the sound generated by the first power module in the vibration process, so that the first sound data set is obtained.
S200: carrying out noise reduction and filtering on the first sound data set to obtain a second sound data set;
specifically, the first power module detects and collects a first sound data set through a sound sensor in the vibration process, and the main sound data used for judging the quality of the first power module in the first sound data set is the welding condition of tin balls in the first power module. However, the first power module generates noise due to irregular vibration and friction of the housing and other structures during the vibration process, which affects the judgment of the first sound data set. And filtering out uneven or unstable sound data in the first sound data set by noise reduction filtering to obtain a second sound data set. The sound data in the second sound data set are stable, the noise is low, and the credibility for detecting and judging the quality of the power supply module is higher.
S300: obtaining a feature set of the second sound dataset;
specifically, the feature set of the second sound data set includes sound features of the first power module vibration sound in the second sound data set after noise reduction and filtering. Illustratively, the second sound data set includes characteristics of intensity, frequency, wavelength, tone and the like of sound generated by vibration of the internal main board and the tin ball in the vibration process of the power module, and the characteristic set can be obtained by extracting the second sound data set through computing software.
S400: inputting the characteristic set into a power quality detection model to obtain a first detection result;
specifically, the feature set is input into the power quality detection model, and the power quality detection model outputs a result according to the feature set, wherein the result includes the first detection result. The power supply quality detection model is a neural network model in machine learning, reflects many basic characteristics of human brain functions, and is a highly complex nonlinear power learning system. After the power quality detection model is trained by a plurality of groups of training data, a first detection result with quality detection result parameters can be output according to the characteristic set after the training is finished. The multiple groups of training data comprise a characteristic information set of a sound data set obtained based on big data and a historical power module vibration experiment, and the set comprises characteristic information of sound data sets generated by power modules of different types, different models and different masses under different vibration conditions. And inputting the feature set of the second sound data set into the power quality detection model, and performing matching calculation on the model to obtain a corresponding first detection result.
The first detection result includes a quality detection result of the first power module obtained according to the feature set, and may illustratively include a plurality of grades, with different grades indicating the quality detection result of the battery module. The first detection result may include only two detection results, i.e., pass and fail, where pass is 1 and fail is 0. The first detection result is used to show the skilled person, and the skilled person can make a judgment according to the first detection result.
S500: acquiring first image information through the image acquisition device, wherein the first image information is appearance image information of the first power supply module;
specifically, the image acquisition device is any device capable of obtaining image information by photographing or imaging, and is preferably a video camera. The first image information is specifically appearance image information acquired by the image acquisition device at any illumination angle and all angles. Illustratively, the first image information specifically includes a degree of sealing with respect to the first power supply module, a soldering condition of the circuit board, external image information of the battery, and the like, and includes different image information for different kinds of power supplies.
S600: performing feature extraction analysis on the first image information to obtain a first appearance feature set;
s700: obtaining a first check factor according to the first appearance characteristic set;
specifically, the first image information is subjected to feature extraction and analysis, for example, without limitation, internal standard image information of the power module, such as standard image information including tin beads, is obtained based on the big data, traversal convolution kernel feature extraction is performed on the first image information according to the standard image information, and then a first appearance feature set is extracted, wherein the first appearance feature set includes features of an appearance image meeting quality standards and image features of an appearance image not meeting the quality standards. The first appearance feature set can also be processed by other image processing methods to obtain a first feature set.
And obtaining a first check factor based on the condition that the appearance features in the first appearance feature set accord with the quality standard, wherein the first check factor is used for checking and calculating the first detection result. The proportion of the appearance features meeting the quality standard in the first appearance feature set is inversely related to the influence of the first check factor on the first detection result. The larger the proportion of the appearance features meeting the quality standard in the first appearance feature set is, the smaller the first verification factor is, and the smaller the weight value of the first detection result for verification is.
S800: and verifying the first detection result according to the first verification factor to obtain a second detection result.
Specifically, the first detection result is verified by using a first verification factor, and the first detection result obtained by the second sound data set is verified and adjusted based on the first image information, so that a second detection result is obtained.
According to the embodiment of the application, the power supply is detected and noise is reduced through the vibration of the power supply modules with different frequencies and amplitudes, the welding condition of the tin beads welded inside the power supply module and the stability condition of other structures can be obtained, characteristic analysis is carried out by combining images of the power supply modules, the sealing and appearance or other power supply model information is verified, the quality detection result after verification is obtained, a method for quantitatively detecting the quality of the power supply modules is established, the power supply quality can be detected in a multi-dimensional and comprehensive mode, the reason or the position of the power supply module which generates faults can be known when the fact that the power supply module can be conducted to provide the power supply is known, and the technical effect of comprehensively, accurately and intelligently detecting the quality of the power supply module is achieved.
As shown in fig. 2, after step S400, the method provided in the embodiment of the present application further includes:
s410: obtaining precision ratio and recall ratio of the characteristic set input power quality detection model by an error evaluation method of skew classification;
s420: obtaining a first evaluation metric value according to the precision ratio and the recall ratio;
s430: and if the first evaluation metric value does not meet the preset standard, performing regularization processing on the characteristic set input power quality detection model.
Specifically, in the process of detecting the battery module, the use condition of most of the battery modules is normal, that is, the quality of most of the battery modules is qualified, while the use of a small number of the battery modules exceeds the age limit, the quality is unqualified or is damaged beyond the age limit, and the quality of the battery modules is unqualified, and the occupation ratio of the small number of the battery modules in the set of all the battery modules is small, so the set of the power module set and the corresponding second sound data set is a skew data sample, wherein the result of the quality unqualified battery module is 1, the result of the quality is 0, and the occupation ratio of the result of the quality unqualified battery module in the sample is very small.
Therefore, the output result of the power quality detection model trained based on the skew-like data samples is affected by the skew-like data samples. The probability of the quality of the power module failing is assumed to be 0.5%, and the probability of the quality of the power module failing is assumed to be 99.5%. Meanwhile, the battery module is subjected to quality detection based on the second sound data set input power quality detection model, the probability that the quality of the power module is qualified is 99% output by the power quality detection model, the probability that the quality of the power module is qualified is high, but actually, the probability that the quality of the power module corresponding to the probability is unqualified is 1%, and the proportion difference between the probability and the actual 0.5% is very large. Therefore, the output result of the power quality detection model is not accurate, and misleading output tends to occur. Therefore, it is necessary to obtain precision (precision) and recall (recall) of the feature set input power quality detection model by using an error evaluation method of skew classification.
Fig. 3 is a schematic diagram of an error evaluation method for skew classification in an embodiment of the present application, where Predicted class is a Predicted class. When the prediction category and the real category are both 1, the prediction error is True Positive (TP), when the real category is 0 and the prediction is 1, the prediction error is False Positive (FP), when the True category is 0 and the prediction is 1, the prediction error is False Negative (FN), and when the prediction and the real are both 0, the prediction error is True Negative (TN). The results predicted by the algorithm are divided into four types:
1. true Positive (True Positive): the prediction is true and the result is true
2. True Negative (True Negative): prediction is false and result is false
3. False Positive (False Positive): prediction is true and result is false
4. False Negative (False Negative): the prediction is false and the result is true
The precision ratio is:
the recall ratio is:
the precision ratio is the probability of success prediction under the condition that all prediction results are that the quality of the power supply module is not qualified; the recall rate is the probability of success of prediction in the case that all the actual results are that the quality of the power supply module is not qualified. In practical application, the recall rate and the accuracy rate both reach higher levels, and the performance of the model is better.
In the embodiment of the present application, to obtain a power quality detection model with better performance, a first evaluation metric of the power quality detection model is calculated, which is specifically as follows:
wherein P is precision, R is recall, F1Scote is the first evaluation metric value. The predetermined standard of the first evaluation metric value can be set based on the previous model in the power module quality detection experiment, and if the first evaluation metric value is larger than the predetermined standard, the corresponding model is considered to have better performance, and the model is adopted for use. And if the first evaluation metric value does not meet the preset standard, performing regularization processing on the characteristic set input power quality detection model.
Step S430 in the method provided in the embodiment of the present application includes:
s431: obtaining a first regularization parameter;
s432: obtaining a first penalty factor;
s433: and performing regularization processing on the cost function of the characteristic set input power supply quality detection model according to the first regularization parameter and the first penalty factor.
Specifically, before inputting the feature set into the power quality detection model, the feature set needs to be fitted, and for example, the sound waves in the feature set in the two sound data sets may be fitted to a wave curve. In the fitting process, an overfitting condition may occur, the fitted curve formed by fitting cannot well express the feature data in the feature set, and at this time, the cost function of inputting the feature set into the power quality detection model needs to be regularized, so that the fitted curve can well express each feature data in the feature set.
The cost function of the characteristic set input power quality detection model is as follows:
the regularization minimizes the cost function while adding a large penalty factor to the parameters in order to guarantee all features. Thus, the cost function is minimized by:
wherein, theta0、θ1Is a parameter;
m is the total amount of feature data in the feature set;
yithe first detection result is obtained;
θ2is the regularization parameter;
λ is the penalty factor.
According to the embodiment of the application, when the first evaluation metric value is not in accordance with the preset standard, the regularization processing is carried out on the characteristic set input power quality detection model, so that the fitting effect of the power quality detection model on the characteristic data in the characteristic set is better, the performance of the model is further improved, the model is more accurate when the data are processed and the result is output, the precision and the recall rate of the model are prevented from being influenced, the technical effects of accurately fitting the characteristic set and improving the accuracy of the data processing and the result output are achieved.
Step S700 in the method provided in the embodiment of the present application includes:
s710: performing principal component analysis on the first appearance characteristic set to obtain a first dimension reduction data set;
s720: and obtaining the first checking factor according to the first dimension reduction data set.
Specifically, the first appearance feature set is an appearance feature set corresponding to the image information of the first power module in a full angle. Because the image characteristic information of the multi-angle and multi-azimuth first power module exists in the first appearance characteristic, and the structure of the power module is more complex. Therefore, the first appearance feature set has a large amount of feature data, and therefore, the dimension reduction of the first appearance feature set is required.
The dimensionality reduction method adopted in the embodiment of the present application is Principal Component Analysis (PCA), which is the most commonly used linear dimensionality reduction method, and its objective is to map high-dimensional data into a low-dimensional space for representation through some linear projection, and expect that the variance of the data is maximum in the projected dimension, so as to use fewer data dimensions, while retaining the characteristics of more raw data points.
By carrying out principal component analysis on the first appearance characteristic set, a first dimension reduction data set is obtained, so that the characteristic data in the processed first appearance characteristic set can show lower data dimension, the processing is convenient, all characteristics of a plurality of image characteristic data in the original first appearance characteristic set are kept, and the technical effect of optimizing the data processing speed is achieved.
As shown in fig. 4, step S710 in the method provided in the embodiment of the present application includes:
s711: performing decentralized processing on the first appearance feature set to obtain a second appearance feature set;
s712: obtaining a first covariance matrix of the second appearance feature set;
s713: calculating the first covariance matrix to obtain a first eigenvector of the first covariance matrix;
s714: and projecting the first appearance feature set to the first feature vector to obtain a first dimension reduction data set, wherein the first dimension reduction data set is the appearance feature set after dimension reduction of the first appearance feature set.
Specifically, the average value of all feature data is subtracted from each feature data in the first appearance feature set to complete decentralized processing, the capability of each processed feature data representing the quality of the first power module is more average, and then a processed second appearance feature set is obtained.
For example, in the second appearance feature set after dimensionality reduction in the method provided by the embodiment of the present application, an n-dimensional vector w is set as one coordinate axis direction of a target subspace (referred to as a mapping vector), and a variance after data mapping is maximized, where:
wherein m is the number of the feature data in the second appearance feature set, XiIs a vector representation of the feature data i,
the method is characterized in that the method is an average vector of all characteristic data, W is defined as a matrix containing all mapping vectors as column vectors, and the following optimization objective function can be obtained through linear algebraic transformation:
where tr represents the trace of the matrix, and:
and A is a first covariance matrix, the A is operated to obtain a first eigenvector of the first covariance matrix, a plurality of first eigenvectors form a group of orthogonal bases, the first appearance characteristic set is projected to the first eigenvector to obtain a first dimension reduction data set, and data characteristic information in the first appearance characteristic set is preferably reserved while dimension reduction is carried out on the data.
According to the method and the device, the first appearance feature set is subjected to decentralization, the first covariance look-up matrix is obtained and calculated, the first dimensionality reduction data set is obtained, the feature data in the processed first appearance feature set can show lower data dimensionality, processing is facilitated, all characteristics of a plurality of image feature data in the original first appearance feature set are kept, and the technical effect of optimizing the data processing speed is achieved.
Step S400 in the method provided in the embodiment of the present application includes:
s410: inputting the characteristic information in the characteristic set as input information into the power supply quality detection model;
s420: the power quality detection model is obtained through training of multiple groups of training data, wherein each group of data in the multiple groups of training data comprises the characteristic information and identification information used for identifying the first detection result;
s430: and obtaining output information of the power quality detection model, wherein the output information comprises the first detection result.
Specifically, the power quality detection model is a neural network model, which can perform continuous self-training learning according to training data, and each set of training data in the multiple sets of training data comprises: the power quality detection model is continuously self-corrected, and when the output information of the power quality detection model reaches a preset accuracy rate/convergence state, the supervised learning process is ended. By carrying out data training on the power quality detection model, the power quality detection model can process input data more accurately, and then output first detection result information is more accurate, so that the technical effects of accurately obtaining data information and improving intellectualization of evaluation results are achieved
In summary, in the embodiments of the present application, the power supply is detected and denoised by the sound sensor through the vibration of the power supply modules with different frequencies and amplitudes, so as to obtain the welding condition of the solder balls welded inside the power supply module and the stability condition of other structures, and the characteristic analysis is performed in combination with the power supply module image, so as to verify the seal, appearance or other power supply model information, obtain the quality detection result after verification, construct a method capable of quantitatively detecting the quality of the power supply module itself, improve the performance of the model through the error evaluation method of skew classification and regularization, improve the accuracy of the model for outputting the first detection result data, optimize the calculation speed of the model, detect the quality of the power supply in a multi-dimensional and comprehensive manner, and know whether the power supply module can be turned on to provide the power supply, and know the cause or position of the power supply module generating a fault, the technical effect of comprehensively, accurately and intelligently detecting the quality of the power module is achieved.
Example two
Based on the same inventive concept as the intelligent quality detection method of the power module in the foregoing embodiment, as shown in fig. 5, an embodiment of the present application provides an intelligent quality detection system of a power module, wherein the system includes:
the first obtaining unit 11 is configured to perform sound collection on a first power module through a sound sensor to obtain a first sound data set, where the first sound data set is a sound data set of the first power module under different vibration conditions;
a first processing unit 12, where the first processing unit 12 is configured to perform noise reduction filtering on the first sound data set to obtain a second sound data set;
a second obtaining unit 13, configured to obtain a feature set of the second sound data set;
the second processing unit 14, where the second processing unit 14 is configured to input the feature set into a power quality detection model, and obtain a first detection result;
a third obtaining unit 15, where the third obtaining unit 15 is configured to obtain first image information through an image acquisition device, where the first image information is appearance image information of the first power module;
a third processing unit 16, where the third processing unit 16 is configured to perform feature extraction analysis on the first image information to obtain a first appearance feature set;
a fourth processing unit 17, where the fourth processing unit 17 is configured to obtain a first verification factor according to the first appearance feature set;
a fifth processing unit 18, where the fifth processing unit 18 is configured to verify the first detection result according to the first verification factor to obtain a second detection result.
Further, the system further comprises:
a sixth processing unit, configured to obtain precision and recall of the feature set input power quality detection model by an error evaluation method of skew classification;
a fourth obtaining unit, configured to obtain a first evaluation metric value according to the precision ratio and the recall ratio;
and the first judgment unit is used for carrying out regularization processing on the characteristic set input power quality detection model if the first evaluation metric value does not meet a preset standard.
Further, the system further comprises:
a fifth obtaining unit, configured to obtain a first regularization parameter;
a sixth obtaining unit, configured to obtain a first penalty factor;
and the seventh processing unit is used for performing regularization processing on the cost function of the characteristic set input power quality detection model according to the first regularization parameter and the first penalty factor.
Further, the system further comprises:
an eighth processing unit, configured to perform principal component analysis on the first appearance feature set to obtain a first dimension reduction data set;
a ninth processing unit configured to obtain the first verification factor according to the first dimension-reduced data set.
Further, the system further comprises:
a seventh obtaining unit, configured to obtain a first covariance matrix of the second appearance feature set;
a tenth processing unit, configured to perform operation on the first covariance matrix to obtain a first eigenvector of the first covariance matrix;
an eleventh processing unit, configured to project the first appearance feature set to the first feature vector, to obtain a first dimension reduction data set, where the first dimension reduction data set is an appearance feature set after dimension reduction of the first appearance feature set.
Further, the system further comprises:
a twelfth processing unit, configured to input feature information in the feature set as input information into the power quality detection model;
the first construction unit is used for obtaining the power quality detection model through training of multiple groups of training data, wherein each group of data in the multiple groups of training data comprises the characteristic information and identification information used for identifying the first detection result;
a thirteenth processing unit, configured to obtain output information of the power quality detection model, where the output information includes the first detection result.
Exemplary electronic device
The electronic device of the embodiment of the present application is described below with reference to figure 6,
based on the same inventive concept as the intelligent quality detection method of the power module in the foregoing embodiment, the embodiment of the present application further provides an intelligent quality detection system of the power module, including: a processor coupled to a memory, the memory for storing a program that, when executed by the processor, causes the system to perform the steps of the method of embodiment one.
The electronic device 300 includes: processor 302, communication interface 303, memory 301. Optionally, the electronic device 300 may also include a bus architecture 304. Wherein, the communication interface 303, the processor 302 and the memory 301 may be connected to each other through a bus architecture 304; the bus architecture 304 may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus architecture 304 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in FIG. 5, but this is not intended to represent only one bus or type of bus.
The communication interface 303 is a system using any transceiver or the like, and is used for communicating with other devices or communication networks, such as ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), wired access network, and the like.
The memory 301 may be a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, an electrically erasable Programmable read-only memory (EEPROM), a compact disc read-only memory (compact disc)
read-only memory, CD-ROM) or other optical disk storage, optical disk storage (including compact disk, laser disk, optical disk, digital versatile disk, blu-ray disk, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited to such. The memory may be self-contained and coupled to the processor through a bus architecture 304. The memory may also be integral to the processor.
The memory 301 is used for storing computer-executable instructions for executing the present application, and is controlled by the processor 302 to execute. The processor 302 is configured to execute the computer-executable instructions stored in the memory 301, so as to implement the intelligent quality detection method for the power module provided by the above-mentioned embodiment of the present application.
Optionally, the computer-executable instructions in the embodiments of the present application may also be referred to as application program codes, which are not specifically limited in the embodiments of the present application.
The embodiment of the application detects and reduces noise of the power supply through the vibration of the power supply modules with different frequencies and amplitudes, can obtain the welding condition of the welding tin beads inside the power supply modules and the stability condition of other structures, performs characteristic analysis by combining with the images of the power supply modules, verifies the information of sealing, appearance or other power supply models, obtains the quality detection result after verification, constructs a method for quantitatively detecting the quality of the power supply modules, improves the performance of the model through an error evaluation method of deviation classification and regularization treatment, improves the accuracy of the model for outputting first detection result data, optimizes the calculation speed of the model, can detect the quality of the power supply in multiple dimensions and comprehensively, and can know the reasons or positions of faults of the power supply modules when knowing whether the power supply modules can be conducted to provide the power supply, the technical effect of comprehensively, accurately and intelligently detecting the quality of the power module is achieved.
Those of ordinary skill in the art will understand that: the various numbers of the first, second, etc. mentioned in this application are only used for the convenience of description and are not used to limit the scope of the embodiments of this application, nor to indicate the order of precedence. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "at least one" means one or more. At least two means two or more. "at least one," "any," or similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one (one ) of a, b, or c, may represent: a, b, c, a-b, a-c, b-c, or a-b-c, wherein a, b, c may be single or multiple.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the application to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device including one or more available media integrated servers, data centers, and the like. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The various illustrative logical units and circuits described in this application may be implemented or operated upon by design of a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.
The steps of a method or algorithm described in the embodiments herein may be embodied directly in hardware, in a software element executed by a processor, or in a combination of the two. The software cells may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. For example, a storage medium may be coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may be disposed in a terminal. In the alternative, the processor and the storage medium may reside in different components within the terminal. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Although the present application has been described in conjunction with specific features and embodiments thereof, it will be evident that various modifications and combinations can be made thereto without departing from the spirit and scope of the application. Accordingly, the specification and figures are merely exemplary of the present application as defined in the appended claims and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations.
Claims (9)
1. An intelligent quality detection method for a power module, wherein the method is applied to an intelligent quality detection system for the power module, the system comprises an acoustic sensor and an image acquisition device, and the method comprises the following steps:
carrying out sound collection on a first power supply module through the sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions;
carrying out noise reduction and filtering on the first sound data set to obtain a second sound data set;
obtaining a feature set of the second sound dataset;
inputting the characteristic set into a power quality detection model to obtain a first detection result;
acquiring first image information through the image acquisition device, wherein the first image information is appearance image information of the first power supply module;
performing feature extraction analysis on the first image information to obtain a first appearance feature set;
obtaining a first check factor according to the first appearance characteristic set;
and verifying the first detection result according to the first verification factor to obtain a second detection result.
2. The method of claim 1, wherein said inputting the feature set into a power quality inspection model, after obtaining a first inspection result, comprises:
obtaining precision ratio and recall ratio of the characteristic set input power quality detection model by an error evaluation method of skew classification;
obtaining a first evaluation metric value according to the precision ratio and the recall ratio;
and if the first evaluation metric value does not meet the preset standard, performing regularization processing on the characteristic set input power quality detection model.
3. The method of claim 2, wherein the regularizing the feature set input power quality detection model if the first evaluation metric value does not meet a predetermined criterion comprises:
obtaining a first regularization parameter;
obtaining a first penalty factor;
and performing regularization processing on the cost function of the characteristic set input power supply quality detection model according to the first regularization parameter and the first penalty factor.
4. The method of claim 3, wherein the cost function of the feature set input power quality detection model is:
the cost function is minimized, and, then,
wherein, theta0、θ1Is a parameter;
m is the total amount of feature data in the feature set;
yithe first detection result is obtained;
θ2is the regularization parameter;
λ is the penalty factor.
5. The method of claim 1, wherein said obtaining a first check factor from said first set of appearance characteristics comprises:
performing principal component analysis on the first appearance characteristic set to obtain a first dimension reduction data set;
and obtaining the first checking factor according to the first dimension reduction data set.
6. The method of claim 5, wherein the performing a principal component analysis on the first set of appearance features to obtain a first set of reduced dimensional data comprises:
performing decentralized processing on the first appearance feature set to obtain a second appearance feature set;
obtaining a first covariance matrix of the second appearance feature set;
calculating the first covariance matrix to obtain a first eigenvector of the first covariance matrix;
and projecting the first appearance feature set to the first feature vector to obtain a first dimension reduction data set, wherein the first dimension reduction data set is the appearance feature set after dimension reduction of the first appearance feature set.
7. The method of claim 1, wherein said inputting the feature set into a power quality inspection model to obtain a first inspection result comprises:
inputting the characteristic information in the characteristic set as input information into the power supply quality detection model;
the power quality detection model is obtained through training of multiple groups of training data, wherein each group of data in the multiple groups of training data comprises the characteristic information and identification information used for identifying the first detection result;
and obtaining output information of the power quality detection model, wherein the output information comprises the first detection result.
8. An intelligent quality detection system for a power module, wherein the system comprises:
the first obtaining unit is used for carrying out sound collection on a first power supply module through a sound sensor to obtain a first sound data set, wherein the first sound data set is a sound data set of the first power supply module under different vibration conditions;
a first processing unit, configured to perform noise reduction filtering on the first sound data set to obtain a second sound data set;
a second obtaining unit configured to obtain a feature set of the second sound dataset;
the second processing unit is used for inputting the feature set into a power quality detection model to obtain a first detection result;
a third obtaining unit, configured to obtain first image information through an image acquisition device, where the first image information is appearance image information of the first power module;
a third processing unit, configured to perform feature extraction analysis on the first image information to obtain a first appearance feature set;
a fourth processing unit, configured to obtain a first check factor according to the first appearance feature set;
and the fifth processing unit is used for verifying the first detection result according to the first verification factor to obtain a second detection result.
9. An intelligent quality detection system for a power module, comprising: a processor coupled to a memory, the memory for storing a program that, when executed by the processor, causes a system to perform the steps of the method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111664611.2A CN114355234A (en) | 2021-12-31 | 2021-12-31 | Intelligent quality detection method and system for power module |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111664611.2A CN114355234A (en) | 2021-12-31 | 2021-12-31 | Intelligent quality detection method and system for power module |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114355234A true CN114355234A (en) | 2022-04-15 |
Family
ID=81105168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111664611.2A Withdrawn CN114355234A (en) | 2021-12-31 | 2021-12-31 | Intelligent quality detection method and system for power module |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114355234A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114627891A (en) * | 2022-05-16 | 2022-06-14 | 山东捷瑞信息技术产业研究院有限公司 | Moving coil loudspeaker quality detection method and device |
-
2021
- 2021-12-31 CN CN202111664611.2A patent/CN114355234A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114627891A (en) * | 2022-05-16 | 2022-06-14 | 山东捷瑞信息技术产业研究院有限公司 | Moving coil loudspeaker quality detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116150676B (en) | Equipment fault diagnosis and identification method and device based on artificial intelligence | |
| WO2019051941A1 (en) | Method, apparatus and device for identifying vehicle type, and computer-readable storage medium | |
| CN112766342A (en) | Abnormity detection method for electrical equipment | |
| CN111932510B (en) | Method and device for determining image definition | |
| CN116453438B (en) | Display screen parameter detection method, device, equipment and storage medium | |
| CN109815988B (en) | Model generation method, classification method, device and computer-readable storage medium | |
| CN108919059A (en) | A kind of electric network failure diagnosis method, apparatus, equipment and readable storage medium storing program for executing | |
| CN109633448B (en) | Method and device for identifying battery health state and terminal equipment | |
| CN113111591B (en) | Automatic diagnosis method, device and equipment based on internal fault of modular power distribution terminal | |
| WO2019176988A1 (en) | Inspection system, identification system, and device for evaluating identification apparatus | |
| CN113820333B (en) | Battery pole piece abnormality detection method, device, upper computer and detection system | |
| WO2019095587A1 (en) | Face recognition method, application server, and computer-readable storage medium | |
| CN113542241A (en) | Intrusion detection method and device based on CNN-BiGRU mixed model | |
| CN115994327A (en) | Equipment fault diagnosis method and device based on edge calculation | |
| CN114997342B (en) | SCR fault diagnosis method, device, equipment and storage medium | |
| CN114355234A (en) | Intelligent quality detection method and system for power module | |
| CN113487223B (en) | Risk assessment method and system based on information fusion | |
| CN117112336B (en) | Intelligent communication equipment abnormality detection method, equipment, storage medium and device | |
| CN117975999A (en) | Valve cooling equipment fault detection method and device, computer equipment and storage medium | |
| CN115932661A (en) | Method for analyzing and judging noise compliance of transformer | |
| US20230204549A1 (en) | Apparatus and automated method for evaluating sensor measured values, and use of the apparatus | |
| CN113139673A (en) | Method, device, terminal and storage medium for predicting air quality | |
| CN112363859A (en) | Method and device for determining abnormality determination threshold | |
| CN111597934A (en) | System and method for processing training data for statistical applications | |
| CN118196567B (en) | Data evaluation method, device, equipment and storage medium based on large language model |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220415 |