CN118226236B - Intelligent recognition method and device for PCB defects - Google Patents
Intelligent recognition method and device for PCB defects Download PDFInfo
- Publication number
- CN118226236B CN118226236B CN202410646815.0A CN202410646815A CN118226236B CN 118226236 B CN118226236 B CN 118226236B CN 202410646815 A CN202410646815 A CN 202410646815A CN 118226236 B CN118226236 B CN 118226236B
- Authority
- CN
- China
- Prior art keywords
- pcb
- abnormal
- electrical
- defect
- test
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 137
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000002159 abnormal effect Effects 0.000 claims abstract description 125
- 238000012360 testing method Methods 0.000 claims abstract description 88
- 238000001514 detection method Methods 0.000 claims abstract description 77
- 230000000007 visual effect Effects 0.000 claims abstract description 38
- 238000004458 analytical method Methods 0.000 claims abstract description 20
- 238000003384 imaging method Methods 0.000 claims abstract description 14
- 238000012795 verification Methods 0.000 claims abstract description 7
- 238000012549 training Methods 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 16
- 230000005856 abnormality Effects 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 12
- 230000008439 repair process Effects 0.000 claims description 12
- 238000007405 data analysis Methods 0.000 claims description 10
- 238000004519 manufacturing process Methods 0.000 claims description 8
- 238000011056 performance test Methods 0.000 claims description 8
- 239000000523 sample Substances 0.000 claims description 7
- 230000003321 amplification Effects 0.000 claims description 6
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 abstract description 3
- 150000003071 polychlorinated biphenyls Chemical class 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 7
- 230000006872 improvement Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 3
- 230000004807 localization Effects 0.000 description 3
- 238000002790 cross-validation Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011179 visual inspection Methods 0.000 description 2
- 238000003466 welding Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
- G01R31/281—Specific types of tests or tests for a specific type of fault, e.g. thermal mapping, shorts testing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/302—Contactless testing
- G01R31/304—Contactless testing of printed or hybrid circuits
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention belongs to the technical field of PCBs and discloses an intelligent recognition method and device for PCB defects, comprising the steps of performing electrical characteristic test on a PCB to be detected, obtaining and analyzing an electrical parameter set, recognizing an abnormal parameter with a difference from an expected value, inputting the electrical parameter set and the abnormal parameter into a trained CNN model for analysis, and preliminarily positioning an abnormal region; imaging the abnormal region by using a high-definition camera to obtain a high-definition image of the abnormal region, amplifying and presenting the high-definition image in a preset proportion, and identifying a specific defect type to obtain a visual detection result; the electrical parameter set and the visual detection result of each abnormal area are synthesized, and the performance index of the PCB to be tested is evaluated; the electric characteristic test is adopted to test the initial screening and rapidly locate the area with the defects, so that the efficiency of the detection flow is effectively improved, and the electric test and visual detection are combined, so that the double verification method not only enhances the detection accuracy, but also obviously reduces the omission rate, and ensures the comprehensive identification of the defects.
Description
Technical Field
The invention relates to the technical field of PCB detection, in particular to an intelligent identification method and device for PCB defects.
Background
In the electronics manufacturing industry, printed Circuit Boards (PCBs) are fundamental components of all electronic devices, responsible for providing connections between electronic components. In order to ensure the quality and functional integrity of PCBs is critical to the overall electronics manufacturing industry, it is necessary to identify and repair any defects or errors on PCBs by accurate and reliable inspection methods to avoid possible performance failures, ensuring the final quality of the electronic product.
In the prior art, the defect detection of the PCB mainly depends on a visual detection technology, wherein the technology captures an image of the PCB by using a high-resolution camera and identifies a defect area by using an image processing technology; although this approach can achieve automated detection to some extent, there are several significant drawbacks; firstly, the whole PCB needs to be scanned in visual detection, and effective positioning measures are lacked, so that the time consumption in the process is long, and particularly for large-size PCBs, the efficiency is low; secondly, because of the dependence on image recognition technology, the accuracy of the method is limited when small or complex defects are recognized, and the phenomenon of missed detection or false detection is easy to occur, especially under the condition of dense PCB or complex design.
In view of this, there is a need for improvements in the prior art PCB defect detection techniques to address the lack of efficient positioning and inefficiency.
Disclosure of Invention
The invention aims to provide an intelligent recognition method and device for PCB defects, and solves the technical problems.
To achieve the purpose, the invention adopts the following technical scheme:
An intelligent recognition method of PCB defects, comprising:
Connecting a PCB to be tested to electrical testing equipment, and performing electrical characteristic test on the PCB to be tested to obtain an electrical parameter set;
analyzing the electrical parameter set, identifying abnormal parameters which are different from expected values, inputting the electrical parameter set and the abnormal parameters into a trained CNN model for analysis, and primarily positioning an abnormal region with defects;
Imaging the abnormal region by using a high-definition camera to obtain a high-definition image of the abnormal region, magnifying and presenting the high-definition image according to a preset proportion by a display unit, and identifying a specific defect type to obtain a visual detection result;
and (3) integrating the electrical parameter set and the visual detection result of each abnormal region, and evaluating the performance index of the PCB to be tested.
Optionally, connecting the PCB to be tested to electrical testing equipment, and performing electrical characteristic test on the PCB to be tested to obtain an electrical parameter set; the method specifically comprises the following steps:
Fixing a PCB to be tested on a test platform, and connecting a test probe of electrical test equipment with a designated test point of the PCB to be tested according to test requirements;
Setting test parameters on the electrical test equipment, and adjusting according to the specification and test requirements of the PCB to be tested;
starting the electrical test equipment, executing a preset test program, automatically performing comprehensive electrical performance test on the PCB to be tested according to set test parameters by the test program, and monitoring test data in real time in the test process;
after the electrical performance test is completed, all relevant electrical parameter sets are collected, including resistance, capacitance, inductance and leakage current.
Optionally, the analyzing the electrical parameter set identifies an abnormal parameter that is different from an expected value; the method specifically comprises the following steps:
The resistance, the capacitance, the inductance and the leakage current of the electrical parameter set are respectively compared with corresponding expected values, the electrical parameter set with abnormality is identified, and the electrical parameter set with abnormality is arranged to form an obtained abnormal parameter;
based on the type and the abnormal amplitude of the abnormal electrical parameter set, the defect type is primarily identified, and a primary defect area is presumed according to the primarily identified defect type.
Optionally, the electrical parameter set and the abnormal parameters are input into a trained CNN model for analysis, and an abnormal area with defects is initially positioned; the method specifically comprises the following steps:
Calling a CNN model, and performing model training of an abnormal region identification function on the CNN model;
Carrying out data formatting processing on the abnormal parameters, and inputting the formatted abnormal parameters into the CNN model after training as input features;
the trained CNN model carries out parameter analysis on the abnormal parameters and outputs one or more predicted abnormal areas;
And integrating the predicted abnormal region and the preliminary defect region to preliminarily locate the abnormal region with the defect.
Optionally, the training process of the CNN model is:
Collecting a preset number of data sets for testing the electrical characteristics of the PCB, wherein the data sets comprise normal electrical parameters, abnormal electrical parameters and corresponding PCB defect area information;
Marking each data point of the abnormal electrical parameter, and indicating the corresponding component or circuit defect type and defect position;
designing a framework of a CNN model, wherein the framework comprises an input layer, a feature extraction layer, a function prediction layer and an output layer;
inputting the collected data set and the labeling information of the abnormal electrical parameters into the CNN model as a training set, and identifying the relationship between the abnormal parameters and expected values by the CNN model through a supervised learning strategy so as to obtain the CNN model with an abnormal region identification function;
Through repeated iterative training of the CNN model, optimizing model parameters, performing cross verification by adopting a data set which does not participate in training, and adjusting the model parameters of the CNN model according to a cross verification result; the model parameters include learning rate, number of layers, and filter size.
Optionally, the high-definition image is amplified and presented through a display unit according to a preset proportion, a specific defect type is identified, and a visual detection result is obtained; specifically comprises
Uploading the obtained high-definition image of the abnormal region to an image processing unit of a detection system;
setting the amplification ratio of the high-definition image of each abnormal region according to the size of the PCB to be tested and the defect detection precision requirement, and amplifying the uploaded high-definition image according to the corresponding amplification ratio through a display unit;
And preprocessing the high-definition image through the image processing unit, visually detecting the defect type of the preprocessed high-definition image, and marking the identified defect position and defect type on the high-definition image to obtain a visual detection result.
Optionally, the electrical parameter set and the visual detection result of each abnormal region are integrated, and the performance index of the PCB to be tested is evaluated; the method specifically comprises the following steps:
analyzing the relevance between the electrical parameter abnormality and the visual detection result through a data analysis model, and correlating the electrical parameter abnormality with the defect type detected visually;
And judging the influence degree of the electrical parameter abnormality and the visually detected defect type on the performance of the PCB, and evaluating the overall performance index of the PCB to obtain the performance index of the PCB to be tested.
Optionally, the evaluating the performance index of the PCB to be tested further includes:
And based on the performance indexes, formulating a repair suggestion aiming at each defect type of the PCB to be tested, and summarizing to form a production guidance report.
The invention provides an intelligent recognition device of PCB defects, which is used for realizing the intelligent recognition method of the PCB defects, and comprises the following steps:
The electrical testing equipment comprises a testing platform and a testing probe, and is used for testing the electrical characteristics of the PCB to be tested to obtain an electrical parameter set;
The image capturing module comprises a high-definition camera, an imaging light source and a positioning device, wherein the high-definition camera is used for imaging the abnormal region to obtain a high-definition image of the abnormal region;
The detection system comprises an image processing unit and a display unit, wherein the display unit is used for amplifying and presenting the high-definition image according to a preset proportion;
The data analysis module comprises a data analysis model and is used for integrating the electrical parameter set and the visual detection result of each abnormal region and evaluating the performance index of the PCB to be tested;
and the recognition software module comprises a CNN model, and is used for carrying out model training on the CNN model and preliminarily positioning an abnormal region with a defect through the trained CNN model.
Compared with the prior art, the invention has the following beneficial effects: during detection, connecting a PCB to be detected to electrical test equipment to perform comprehensive electrical characteristic test on the PCB, quickly obtaining electrical parameters, analyzing the electrical parameters to identify abnormal parameters with obvious differences from expected values, preliminarily positioning areas with defects, imaging the areas by using a high-definition camera after determining the abnormal areas, acquiring high-resolution abnormal area images, amplifying and presenting the images through a display unit according to a preset proportion, and inputting a trained CNN model for analysis to identify specific defect types; the overall performance index of the PCB to be tested is evaluated through the comprehensive electrical parameters and visual detection results, so that the defect detection process is completed; the method can rapidly and accurately locate the area with possible defects through the electric characteristic test as a preliminary screening means, obviously reduces the area range of the subsequent visual detection, effectively improves the efficiency of the detection flow, and not only enhances the detection accuracy, but also obviously reduces the omission ratio, ensures the comprehensive identification and repair of the defects through the combination of the electric test and the visual detection, and provides a more efficient and reliable PCB defect detection scheme.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained from these drawings without inventive faculty for a person skilled in the art.
The structures, proportions, sizes, etc. shown in the drawings are shown only in connection with the present disclosure, and are not intended to limit the scope of the invention, since any modification, variation in proportions, or adjustment of the size, etc. of the structures, proportions, etc. should be considered as falling within the spirit and scope of the invention, without affecting the effect or achievement of the objective.
Fig. 1 is a flow chart of a method for intelligently identifying defects of a PCB according to the first embodiment;
FIG. 2 is a second flowchart of a method for intelligent recognition of PCB defects according to the first embodiment;
FIG. 3 is a third flowchart of a method for intelligent recognition of PCB defects according to the first embodiment;
Fig. 4 is a system layout diagram of an intelligent recognition device for PCB defects in the second embodiment.
Detailed Description
In order to make the objects, features and advantages of the present invention more comprehensible, the technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings, and it is apparent that the embodiments described below are only some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be understood that the directions or positional relationships indicated by the terms "upper", "lower", "top", "bottom", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. It is noted that when one component is referred to as being "connected" to another component, it can be directly connected to the other component or intervening components may also be present.
The technical scheme of the invention is further described below by the specific embodiments with reference to the accompanying drawings.
Embodiment one:
Referring to fig. 1 to 3, an embodiment of the present invention provides an intelligent recognition method for a PCB defect, including:
S1, connecting a PCB to be tested to electrical testing equipment, and performing electrical characteristic testing on the PCB to be tested to obtain an electrical parameter set;
The electrical parameter set of the PCB to be tested is obtained by connecting the PCB to the electrical test equipment, and at this stage, the basic electrical performance of the PCB can be comprehensively evaluated by measuring key parameters such as resistance, capacitance, inductance and leakage current; these electrical parameters are the basis for evaluating whether the PCB board meets design specifications, and are also key clues for identifying potential electrical faults such as shorts, opens, etc.
S2, analyzing the electrical parameter set, identifying abnormal parameters which are different from the expected values, inputting the electrical parameter set and the abnormal parameters into the trained CNN model for analysis, and primarily positioning the abnormal region with the defect;
Analyzing the obtained set of electrical parameters to identify those abnormal parameters that differ significantly from the expected values, which is critical to identifying potential problem areas, because abnormal parameters are typically directed to specific electrical or physical faults; the anomaly parameters are combined with a trained CNN model for preliminary spatial localization of these anomaly regions. By the method, the possible problem areas can be rapidly positioned, the range and the workload of subsequent visual detection can be effectively reduced, and the detection efficiency is improved.
S3, imaging the abnormal region by using a high-definition camera to obtain a high-definition image of the abnormal region, magnifying and presenting the high-definition image according to a preset proportion through a display unit, and identifying a specific defect type to obtain a visual detection result;
Imaging the located abnormal region by using a high-definition camera, and amplifying the obtained high-definition image according to a preset proportion by a display unit so as to more carefully check and identify specific defect types such as welding spot problems, cracks, element dislocation and the like; this step allows the operator to verify and analyze the specifics of the abnormal region from a visual point of view, which is a key element in identifying and detailing defects.
S4, the electrical parameter set and the visual detection result of each abnormal area are synthesized, and the performance index of the PCB to be tested is estimated.
Comprehensively analyzing the electrical parameter set and the visual detection result, and evaluating the performance index of the whole PCB to be tested, wherein the comprehensive analysis comprises the steps of analyzing the association between the abnormal parameters and the visual identification defects and evaluating the possible influence of the defects on the overall performance of the PCB; the comprehensive defect and performance evaluation report is provided, basis is provided for repair decision and quality improvement, and meanwhile, the comprehensive defect and performance evaluation report is also a summary and closed loop of the whole detection flow, so that potential problems are ensured to be identified and corresponding treatment schemes are obtained.
The working principle of the invention is as follows: during detection, connecting a PCB to be detected to electrical test equipment to perform comprehensive electrical characteristic test on the PCB, quickly obtaining electrical parameters, analyzing the electrical parameters to identify abnormal parameters with obvious differences from expected values, preliminarily positioning areas with defects, imaging the areas by using a high-definition camera after determining the abnormal areas, acquiring high-resolution abnormal area images, amplifying and presenting the images through a display unit according to a preset proportion, and inputting a trained CNN model for analysis to identify specific defect types; the overall performance index of the PCB to be tested is evaluated through the comprehensive electrical parameters and visual detection results, so that the defect detection process is completed; compared with the existing PCB defect detection technology, the method can rapidly and accurately locate the area with the possible defects through the electrical characteristic test as a primary screening means, the area range of the subsequent visual detection is obviously reduced, the efficiency of the detection flow is effectively improved, the detection accuracy is enhanced through the combination of the electrical test and the visual detection, the omission ratio is obviously reduced, the comprehensive identification and repair of the defects are ensured, and a more efficient and reliable PCB defect detection scheme is provided.
In this embodiment, it is specifically described that step S1 specifically includes:
S11, fixing the PCB to be tested on a test platform, and connecting a test probe of electrical test equipment with a designated test point of the PCB to be tested according to test requirements;
In the testing process, stable and reliable connection between the PCB and the testing equipment is ensured, and especially correct connection of the testing probe is a basis for obtaining accurate electric parameters, so that inaccurate data caused by incorrect connection is avoided, and subsequent defect analysis and positioning are influenced.
S12, setting test parameters on electrical test equipment, and adjusting according to the specification and test requirements of the PCB to be tested;
And setting proper test parameters such as current and voltage levels on the electrical test equipment, wherein the test parameters need to be adjusted according to the specification of the PCB to be tested and specific test requirements so as to ensure the accuracy of the test. Proper parameter setting can reduce test errors to the greatest extent and improve the repeatability and reliability of test results.
S13, starting electrical test equipment, executing a preset test program, automatically performing comprehensive electrical performance test on the PCB to be tested according to set test parameters by the test program, and monitoring test data in real time in the test process;
starting an electrical testing device, automatically performing comprehensive electrical performance test on the PCB to be tested according to preset parameters, automatically recording the electrical parameters by the device, and simultaneously monitoring test data in real time to ensure that no abnormality occurs in the test process; the real-time monitoring is helpful for finding out and solving the problems possibly occurring in the test process in real time, and ensuring the reliability of the data.
S14, after the electrical performance test is completed, all relevant electrical parameter sets are collected, wherein the electrical parameter sets comprise resistance, capacitance, inductance and leakage current.
After the electrical performance test is completed, all relevant electrical parameters including, but not limited to, resistance, capacitance, inductance and leakage current are collected, and the electrical parameter set is used as a basis for subsequent analysis and determination of whether defects exist in the PCB board, so as to determine abnormal parameters in the following steps for defect localization and analysis.
In this embodiment, it is specifically described that step S2 specifically includes:
S21, respectively comparing the resistance, the capacitance, the inductance and the leakage current of the electrical parameter set with corresponding expected values, identifying the electrical parameter set with abnormality, and finishing to form an obtained abnormal parameter;
Comparing the collected electrical parameter sets (resistance, capacitance, inductance and leakage current) with respective expected values, and identifying abnormal parameters through the comparison; abnormal parameters typically indicate possible faults, such as shorts, component damage, etc., and this step is therefore critical for fault detection and localization.
S22, preliminarily identifying the defect type based on the type and the abnormal amplitude of the abnormal electrical parameter set, and presuming a preliminary defect area according to the preliminarily identified defect type;
Based on the identified abnormal electrical parameters, primarily judging possible defect types (such as welding spot problems, broken wires and the like); then, based on these preliminary defect types, the PCB area that may be affected is presumed. The detection range is further narrowed, and a specific direction is provided for visual detection and detailed analysis, so that the subsequent detection work is more targeted and efficient. And meanwhile, the output result of the subsequent CNN model can be complemented.
S23, calling a CNN model, and performing model training of an abnormal region identification function on the CNN model;
By using machine learning, the automation and the accuracy of analysis are improved, the model training ensures the accuracy of analysis, and the reliability of abnormal region identification is improved; meanwhile, the processing process can be quickened, and the working efficiency is improved.
S24, carrying out data formatting processing on the abnormal parameters, and inputting the formatted abnormal parameters into a trained CNN model as input features;
Carrying out necessary data formatting processing on the abnormal parameters to enable the abnormal parameters to meet the input requirements of the CNN model, and then inputting the data serving as characteristics into the trained CNN model; data formatting is a key step in ensuring that machine learning models can properly interpret and process input data.
S25, carrying out parameter analysis on the abnormal parameters by the trained CNN model, and outputting one or more predicted abnormal areas;
The trained CNN model analyzes the input abnormal parameters to predict the possible abnormal areas, and the step utilizes the machine learning technology to deeply analyze the electrical data. The accuracy and efficiency of defect detection can be greatly improved by utilizing the CNN model for prediction, and the result of model analysis helps to more accurately position defects, so that an accurate basis is provided for repair.
S26, comprehensively predicting the abnormal region and the preliminary defect region to preliminarily locate the abnormal region with the defect.
The defect area estimated by the abnormal electrical parameters and the abnormal area predicted by the CNN model are combined to accurately position the defect-containing abnormal area, so that the defect-containing abnormal area can be more comprehensively estimated.
As a preferable scheme of this embodiment, the training process of the CNN model is:
t101, collecting a preset number of data sets for PCB electrical characteristic tests, wherein the data sets comprise normal electrical parameters and abnormal electrical parameters, and corresponding PCB defect area information;
Collecting a data set for training a CNN model, wherein the data set comprises normal and abnormal electrical parameters and corresponding PCB defect area information; these data sets should cover many different types of PCB electrical characteristics to ensure that the model can learn a wide range of conditions; collecting a wide and diverse array of data is critical to training any machine learning model, which ensures that the model has better predictive and adaptive capabilities when subjected to a variety of conditions in practical applications.
T102, marking each data point of the abnormal electrical parameter, and indicating the corresponding component or circuit defect type and defect position;
Each anomaly electrical parameter in the dataset is annotated in detail, including specifying the type and location of the defect for its corresponding component or circuit, which provides the model with the necessary landmark information for training the model to identify not only the presence or absence of anomalies, but also the specific type and location of defects. With these labels, the model learns how to predict specific defect types and locations from the input electrical parameters, which is critical to improving inspection accuracy.
T103, designing a framework of a CNN model, wherein the framework comprises an input layer, a feature extraction layer, a function prediction layer and an output layer;
the architecture of the CNN model is designed, and the architecture comprises an input layer, a feature extraction layer, a functional prediction layer and an output layer, and defines how the model processes input data, extracts and utilizes information and finally outputs a prediction result. Optimization of the model architecture directly affects training efficiency and prediction accuracy.
T104, inputting the collected data set and the labeling information of the abnormal electrical parameters into a CNN model as a training set, and identifying the relationship between the abnormal parameters and expected values by the CNN model through a supervised learning strategy so as to obtain the CNN model with an abnormal region identification function;
The marked training set data are input into a CNN model for training, and the relationship between the abnormal electrical parameters and the defects is identified through supervised learning, so that the defect area can be predicted; by learning and adjusting the internal weights, the model can learn how to accurately infer the fault condition from the electrical parameters.
T105, through repeated iterative training of the CNN model, optimizing model parameters, performing cross verification by adopting a data set which does not participate in training, and adjusting the model parameters of the CNN model according to a cross verification result; model parameters include learning rate, number of layers, and filter size.
Performing repeated iterative training on the CNN model, and continuously optimizing model parameters such as learning rate, layer number, filter size and the like; cross-validation is performed with the data set not involved in training to detect the generalization ability of the model and performance in practical applications. Iterative training and cross-validation are important links for ensuring that the model not only performs well on a training set, but also can make accurate predictions on unknown data, and final parameters of the model can be adjusted and determined through the steps, so that the effectiveness and accuracy of the model in practical application are ensured.
In this embodiment, it is specifically described that step S3 specifically includes:
s31, imaging the abnormal region by using a high-definition camera to obtain a high-definition image of the abnormal region.
S32, uploading the obtained high-definition image of the abnormal region to an image processing unit of the detection system;
The obtained high definition image is uploaded to an image processing unit of the detection system, which is to transfer the image data to a system that is capable of further processing for detailed image analysis and processing. The method is an important link for ensuring smooth detection flow and is also used for preparing for subsequent image analysis.
S33, setting the amplification ratio of the high-definition image of each abnormal region according to the size of the PCB to be tested and the defect detection precision requirement, and amplifying the uploaded high-definition image according to the corresponding amplification ratio through a display unit; and the display unit is provided with a zoom-in and zoom-out function and a local view function, so that a technician can conveniently operate and view image information, and the defect type can be conveniently screened.
According to the size of the PCB and the accurate requirement of defect detection, the image magnification ratio of each abnormal area is set, and the image is correspondingly magnified through the display unit, so that defect details are better observed and analyzed. The proper magnification can help the technician see the tiny defects more clearly, especially on a complex or small-sized PCB board, where the magnified image is critical for accurately identifying the defect type.
S34, preprocessing the high-definition image through an image processing unit, performing visual detection of the defect type on the preprocessed high-definition image, and marking the identified defect position and defect type on the high-definition image to obtain a visual detection result.
The high-definition image is preprocessed by the image processing unit, such as contrast adjustment, brightness adjustment, filtering and the like, and then the visual detection of the defect type is performed. In the detection process, the identified defect position and type are marked on the image by utilizing an image processing technology, and finally, a visual detection result is obtained. The image preprocessing can enhance the visibility of defects, improve the detection accuracy, and the marked defect positions and types provide direct basis for subsequent analysis, report generation and repair work, so that the practicability and operability of detection results are ensured.
In this embodiment, it is specifically described that step S4 specifically includes:
S41, analyzing the relevance between the electrical parameter abnormality and the visual detection result through a data analysis model, and relating the electrical parameter abnormality to the type of the defect detected visually;
Exploring and analyzing the association between the electrical parameter anomalies and the visual inspection results by using a data analysis model, this step involving matching and associating the anomalies of the electrical parameters with the types of defects identified by the visual inspection; the connection between different types of electrical anomalies and specific defects can be revealed, the root cause of the defects can be understood, and the accuracy and the depth of diagnosis can be improved.
S42, judging the influence degree of the abnormal electrical parameters and the visually detected defect types on the performance of the PCB, and evaluating the overall performance index of the PCB to obtain the performance index of the PCB to be tested.
Based on the abnormality of the electrical parameters and the defect types detected visually, evaluating the influence of the problems on the overall performance of the PCB, comprehensively considering the detection data, and comprehensively evaluating the performance of the PCB; the specific influence of the defects on the functions of the PCB is clarified, so that the PCB can be helped to know whether the PCB meets the performance requirement or what improvement measures need to be taken.
In this embodiment, it is further explained that step S4 further includes:
And S5, formulating a repair suggestion aiming at each defect type of the PCB to be tested based on the performance indexes, and summarizing to form a production guidance report.
Based on the results of the performance evaluation, formulating repair recommendations for each identified defect type, and assembling all data and recommendations into a detailed production guidance report; the method not only provides specific guidance for repair work, but also provides basis for improvement of production flow, and can help to know the current state of the PCB and the measures to be taken through detailed reports, so that repair efficiency and product quality are improved.
Embodiment two:
The present invention also provides an intelligent recognition device for a PCB defect, for implementing the intelligent recognition method for a PCB defect according to the first embodiment, as shown in fig. 4, where the intelligent recognition device for a PCB defect includes:
The electrical testing equipment comprises a testing platform and a testing probe, and is used for testing the electrical characteristics of the PCB to be tested to obtain an electrical parameter set;
The image capturing module comprises a high-definition camera, an imaging light source and a positioning device, wherein the high-definition camera is used for imaging the abnormal region to obtain a high-definition image of the abnormal region;
The detection system comprises an image processing unit and a display unit, wherein the display unit is used for amplifying and presenting the high-definition image according to a preset proportion;
The data analysis module comprises a data analysis model and is used for integrating the electrical parameter set and the visual detection result of each abnormal region and evaluating the performance index of the PCB to be tested;
and the recognition software module comprises a CNN model, and is used for carrying out model training on the CNN model and preliminarily positioning an abnormal region with a defect through the trained CNN model.
The user interface module comprises a data display unit and a report output unit, wherein the data display unit is used for displaying the electric parameter set, and the report output unit is used for displaying the production guidance report.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (3)
1. An intelligent recognition method for a PCB defect is characterized by comprising the following steps:
Connecting a PCB to be tested to electrical testing equipment, and performing electrical characteristic test on the PCB to be tested to obtain an electrical parameter set;
analyzing the electrical parameter set, identifying abnormal parameters which are different from expected values, inputting the electrical parameter set and the abnormal parameters into a trained CNN model for analysis, and primarily positioning an abnormal region with defects;
Imaging the abnormal region by using a high-definition camera to obtain a high-definition image of the abnormal region, magnifying and presenting the high-definition image according to a preset proportion by a display unit, and identifying a specific defect type to obtain a visual detection result;
the electrical parameter set and the visual detection result of each abnormal area are synthesized, and the performance index of the PCB to be tested is evaluated;
Connecting a PCB to be tested to electrical testing equipment, and performing electrical characteristic test on the PCB to be tested to obtain an electrical parameter set; the method specifically comprises the following steps:
Fixing a PCB to be tested on a test platform, and connecting a test probe of electrical test equipment with a designated test point of the PCB to be tested according to test requirements;
Setting test parameters on the electrical test equipment, and adjusting according to the specification and test requirements of the PCB to be tested;
starting the electrical test equipment, executing a preset test program, automatically performing comprehensive electrical performance test on the PCB to be tested according to set test parameters by the test program, and monitoring test data in real time in the test process;
After the electrical performance test is completed, collecting all relevant electrical parameter sets of the set, wherein the electrical parameter sets comprise resistance, capacitance, inductance and leakage current;
Wherein the analyzing the electrical parameter set identifies an abnormal parameter that differs from an expected value; the method specifically comprises the following steps:
The resistance, the capacitance, the inductance and the leakage current of the electrical parameter set are respectively compared with corresponding expected values, the electrical parameter set with abnormality is identified, and the electrical parameter set with abnormality is arranged to form an obtained abnormal parameter;
Based on the types and the abnormal amplitudes of the abnormal electrical parameter sets, primarily identifying the defect types, and presuming a primary defect area according to the primarily identified defect types;
Inputting the electrical parameter set and the abnormal parameters into a trained CNN model for analysis, and primarily positioning an abnormal region with defects; the method specifically comprises the following steps:
Calling a CNN model, and performing model training of an abnormal region identification function on the CNN model;
Carrying out data formatting processing on the abnormal parameters, and inputting the formatted abnormal parameters into the CNN model after training as input features;
the trained CNN model carries out parameter analysis on the abnormal parameters and outputs one or more predicted abnormal areas;
Combining the predicted abnormal region and the preliminary defect region to preliminarily locate the abnormal region with the defect;
the training process of the CNN model is as follows:
Collecting a preset number of data sets for testing the electrical characteristics of the PCB, wherein the data sets comprise normal electrical parameters, abnormal electrical parameters and corresponding PCB defect area information;
Marking each data point of the abnormal electrical parameter, and indicating the corresponding component or circuit defect type and defect position;
designing a framework of a CNN model, wherein the framework comprises an input layer, a feature extraction layer, a function prediction layer and an output layer;
inputting the collected data set and the labeling information of the abnormal electrical parameters into the CNN model as a training set, and identifying the relationship between the abnormal parameters and expected values by the CNN model through a supervised learning strategy so as to obtain the CNN model with an abnormal region identification function;
Through repeated iterative training of the CNN model, optimizing model parameters, performing cross verification by adopting a data set which does not participate in training, and adjusting the model parameters of the CNN model according to a cross verification result; the model parameters comprise learning rate, layer number and filter size;
The high-definition image is amplified and presented through a display unit according to a preset proportion, specific defect types are identified, and a visual detection result is obtained; specifically comprises
Uploading the obtained high-definition image of the abnormal region to an image processing unit of a detection system;
Setting the amplification ratio of the high-definition image of each abnormal region according to the size of the PCB to be tested and the defect detection precision requirement, and amplifying the uploaded high-definition image according to the corresponding amplification ratio through a display unit;
Preprocessing the high-definition image through the image processing unit, visually detecting the defect type of the preprocessed high-definition image, marking the identified defect position and defect type on the high-definition image, and obtaining a visual detection result;
the electrical parameter set and the visual detection result of each abnormal area are synthesized, and the performance index of the PCB to be tested is evaluated; the method specifically comprises the following steps:
analyzing the relevance between the electrical parameter abnormality and the visual detection result through a data analysis model, and correlating the electrical parameter abnormality with the defect type detected visually;
And judging the influence degree of the electrical parameter abnormality and the visually detected defect type on the performance of the PCB, and evaluating the overall performance index of the PCB to obtain the performance index of the PCB to be tested.
2. The intelligent recognition method of a PCB defect according to claim 1, wherein the evaluating the performance index of the PCB to be tested further comprises:
And based on the performance indexes, formulating a repair suggestion aiming at each defect type of the PCB to be tested, and summarizing to form a production guidance report.
3. An intelligent recognition device of a PCB defect, for implementing the intelligent recognition method of a PCB defect according to any one of claims 1 to 2, the intelligent recognition device of a PCB defect comprising:
The electrical testing equipment comprises a testing platform and a testing probe, and is used for testing the electrical characteristics of the PCB to be tested to obtain an electrical parameter set;
The image capturing module comprises a high-definition camera, an imaging light source and a positioning device, wherein the high-definition camera is used for imaging the abnormal region to obtain a high-definition image of the abnormal region;
The detection system comprises an image processing unit and a display unit, wherein the display unit is used for amplifying and presenting the high-definition image according to a preset proportion;
The data analysis module comprises a data analysis model and is used for integrating the electrical parameter set and the visual detection result of each abnormal region and evaluating the performance index of the PCB to be tested;
and the recognition software module comprises a CNN model, and is used for carrying out model training on the CNN model and preliminarily positioning an abnormal region with a defect through the trained CNN model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410646815.0A CN118226236B (en) | 2024-05-23 | 2024-05-23 | Intelligent recognition method and device for PCB defects |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410646815.0A CN118226236B (en) | 2024-05-23 | 2024-05-23 | Intelligent recognition method and device for PCB defects |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118226236A CN118226236A (en) | 2024-06-21 |
| CN118226236B true CN118226236B (en) | 2024-08-06 |
Family
ID=91507819
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410646815.0A Active CN118226236B (en) | 2024-05-23 | 2024-05-23 | Intelligent recognition method and device for PCB defects |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118226236B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118641931B (en) * | 2024-06-28 | 2024-12-03 | 青岛展诚科技有限公司 | Method for locating integrated circuit design defects based on electrical rule detection |
| CN118608514B (en) * | 2024-07-01 | 2025-04-04 | 东莞市群益电子有限公司 | An intelligent method and system for implementing PCBA board testing |
| CN118858326A (en) * | 2024-07-08 | 2024-10-29 | 深圳市宏联电路有限公司 | A multi-layer circuit board detection system |
| CN119027393B (en) * | 2024-08-14 | 2025-03-07 | 东莞市方展实业有限公司 | Performance detection method and system for data line |
| CN119023677A (en) * | 2024-08-15 | 2024-11-26 | 镇江俊晨精密加工有限公司 | A highly accurate intelligent detection system for wiring harness defects |
| CN119048483B (en) * | 2024-09-03 | 2025-03-04 | 湖南新韶光智能有限公司 | Quality inspection method for packaged products based on fusion of X-ray and visible light |
| CN119178985A (en) * | 2024-09-10 | 2024-12-24 | 立昱电子(惠州)有限公司 | Test system and test method for circuit board |
| CN119574582B (en) * | 2024-11-25 | 2025-07-25 | 江苏顺海科技有限公司 | SMT surface defect visual inspection device based on AI image recognition |
| CN119893178B (en) * | 2025-03-26 | 2025-06-03 | 四川天邑康和通信股份有限公司 | State evaluation method, device, medium and equipment of set top box |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10154900A (en) * | 1996-11-25 | 1998-06-09 | Hitachi Ltd | Failure analysis method and system for printed circuit board for mounting electronic components |
| KR20050041357A (en) * | 2003-10-30 | 2005-05-04 | 엘지.필립스 엘시디 주식회사 | System for testing a pattern of a liquid crystal display device and method of testing and repairing the pattern using thereof |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5832200B2 (en) * | 2011-08-29 | 2015-12-16 | マイクロクラフト株式会社 | Apparatus and method for acquiring defect analysis image of wiring board using X-ray |
| CN114429445B (en) * | 2021-11-09 | 2025-05-23 | 南京师范大学 | A PCB defect detection and identification method based on MAIRNet |
| KR102504598B1 (en) * | 2022-05-19 | 2023-03-02 | 이성씨엔아이 주식회사 | A device for detecting poor contact in PCB slots for Nuclear Power Plant |
| CN115372793B (en) * | 2022-08-09 | 2023-04-07 | 铜陵威博电子科技有限公司 | Circuit detection method for integrated circuit |
| CN117665546B (en) * | 2024-02-02 | 2024-04-30 | 深圳市鑫恒天电子有限公司 | Automatic test method and system for PCBA (printed circuit board assembly) |
-
2024
- 2024-05-23 CN CN202410646815.0A patent/CN118226236B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10154900A (en) * | 1996-11-25 | 1998-06-09 | Hitachi Ltd | Failure analysis method and system for printed circuit board for mounting electronic components |
| KR20050041357A (en) * | 2003-10-30 | 2005-05-04 | 엘지.필립스 엘시디 주식회사 | System for testing a pattern of a liquid crystal display device and method of testing and repairing the pattern using thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118226236A (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118226236B (en) | Intelligent recognition method and device for PCB defects | |
| US11449980B2 (en) | System and method for combined automatic and manual inspection | |
| CN112053318B (en) | Two-dimensional PCB defect real-time automatic detection and classification device based on deep learning | |
| JP6922168B2 (en) | Surface mount line quality control system and its control method | |
| CN117269731B (en) | PCBA performance automatic test system based on Internet of things | |
| CN117665546B (en) | Automatic test method and system for PCBA (printed circuit board assembly) | |
| CN117572209B (en) | Intelligent door lock main board PCBA testing method | |
| CN104730078A (en) | Thermal infrared imager-based AOI circuit board detection method | |
| CN118429270B (en) | Electronic component damage test method and system based on image recognition | |
| CN118297274B (en) | A data analysis system for ICT testing tooling of automotive parts | |
| CN106980099B (en) | A kind of calibration method and system of Automatic Testing System of Circuit Board | |
| CN117889943B (en) | Gas ultrasonic flowmeter inspection method and system based on machine learning | |
| CN118244088B (en) | Circuit board quality detection method, system and equipment | |
| CN118781097A (en) | Artificial intelligence-based PCB detection method, terminal equipment and storage medium | |
| CN118376864A (en) | System and method for diagnosing faults of tubing mechanism based on semiconductor punching equipment | |
| KR20150104766A (en) | Tracking method for badness history in inspection process | |
| CN114324382A (en) | Panel terminal cleanliness detection method and panel terminal cleanliness detection device | |
| CN117054517A (en) | Ultrasonic line detection method for display device | |
| CN115439459A (en) | Circuit board fault determination method based on infrared thermal imaging | |
| CN114355154A (en) | A high-frequency circuit module detection system and method based on electromagnetic scanning technology | |
| CN113245238A (en) | Elbow welding flaw detection method, device and system | |
| CN114820412A (en) | Automatic detection system and operation method thereof | |
| Abraham et al. | Circuit Scan: PCB Error Detection using Microcontroller | |
| CN111539933B (en) | Direct-insert element detection method and system | |
| CN119728836B (en) | Mobile phone screen dead pixel detection method, system and electronic equipment |
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 |