Visual inspection and intelligent control system for PCBA production line

Abstract

The invention belongs to the technical field of visual inspection, and particularly relates to a visual inspection and intelligent control system for a PCBA (Printed Circuit Board Assembly) production line, which integrates a plurality of units such as image acquisition, image processing, defect identification, data analysis, control execution, feedback monitoring, learning optimization, operation interface and safety guarantee. Real-time intelligent control and decision making on a high-speed production line are realized; through efficient image processing and a rapid defect recognition algorithm, the system can realize real-time detection and control on a high-speed production line, adopts advanced image processing and defect recognition technologies, remarkably improves the accuracy and reliability of defect detection, and improves the accuracy and reliability of defect detection through real-time feedback and continuous learning optimization. The system can automatically adjust the control strategy according to the change of the production environment, the production efficiency and the product quality are improved, illegal access and operation are effectively prevented through multi-layer safety protection measures, and stable operation of the system is ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of visual inspection, and in particular relates to a visual inspection and intelligent control system for a PCBA production line. Background Art

[0002] In the modern electronic manufacturing industry, efficient operation and product quality control of PCBA (Printed Circuit Board Assembly) production lines are crucial. Traditional PCBA production lines mainly rely on manual visual inspection to find and deal with defects. This method is not only inefficient, but also prone to missed inspections and false inspections, resulting in unstable product quality. With the development of automation and intelligent technology, intelligent control systems based on visual inspection have gradually become a hot spot for research and application.

[0003] Although the existing PCBA production line visual inspection and intelligent control system has improved production efficiency and product quality to a certain extent, it still has deficiencies in real-time intelligent control and decision-making on high-speed production lines. Specifically,

[0004] Lack of real-time performance: The existing system has a slow response speed when processing large amounts of data on high-speed production lines and cannot achieve real-time control and decision-making.

[0005] Low accuracy: Existing defect identification and control methods are prone to misjudgment in complex environments, affecting product quality. Summary of the invention

[0006] The purpose of the present invention is to provide a visual inspection and intelligent control system for a PCBA production line, which has significant advantages in real-time intelligent control and decision-making on a high-speed production line, and can effectively improve the production efficiency and product quality of the PCBA production line to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention adopts the following technical solution: a visual inspection and intelligent control system for a PCBA production line, comprising:

[0008] Image acquisition unit, used to capture high-definition images of PCBA boards on the production line;

[0009] An image processing unit, connected to the image acquisition unit, preprocesses the captured image to remove noise and enhance image quality;

[0010] A defect recognition unit, connected to the image processing unit, recognizes potential defects on the PCBA board based on the processed image;

[0011] A data analysis unit, connected to the defect recognition unit, analyzes defect data and generates corresponding control instructions;

[0012] A control execution unit, connected to the data analysis unit, receives and executes control instructions to adjust the working state of the production line;

[0013] A feedback monitoring unit, connected to the control execution unit, monitors the state of the adjusted production line and feeds back the information to the data analysis unit;

[0014] A learning optimization unit, connected to the feedback monitoring unit, optimizes subsequent defect identification and control strategies according to feedback information;

[0015] An operation interface unit, connected to the learning optimization unit, provides a human-computer interaction interface to facilitate operators to monitor the system operation status and input adjustment parameters;

[0016] The security unit is connected to the operation interface unit to ensure the safe operation of the system and prevent illegal access and operation.

[0017] Preferably, the image acquisition unit comprises:

[0018] The light source adjustment submodule is used to dynamically adjust the light source intensity according to the ambient light conditions. The required light source intensity L is calculated by the formula L = αI + β, where I represents the current ambient light intensity, and α and β are coefficients pre-set according to the material characteristics of the PCBA board;

[0019] A high-definition camera submodule is connected to the light source adjustment submodule to capture a high-definition image of the PCBA board using the adjusted light source;

[0020] The image stabilization submodule is connected to the high-definition camera submodule and uses the image frame difference method to reduce the image blur caused by the vibration of the production line. The pixel difference between two consecutive frames of images is calculated by ΔP=|P t -P t-1 |, when ΔP>γ, γ is the preset threshold, and the image stabilization program is started until ΔP≤γ.

[0021] Preferably, the image processing unit comprises:

[0022] The noise filtering submodule is used to receive image data from the image acquisition unit and pass the formula Calculate the filtered pixel value I f (x,y), where w(i,j) is the Gaussian weight matrix, I(x,y) is the pixel value of the original image at position (x,y), and r is the filter radius;

[0023] The contrast enhancement submodule is connected to the noise filtering submodule to enhance the contrast of the denoised image. fCalculate the enhanced contrast C(x,y) by ((x,y) - μ) / δ, where μ and δ are the mean and standard deviation of the image I f (x,y), respectively;

[0024] An edge sharpening sub-module, connected after the contrast enhancement sub-module, sharpens the edges of the image with enhanced contrast to highlight the boundary information in the image. Calculate the final sharpening effect E(x,y) through the formula E(x,y) = C(x,y) + k * (C(x,y) - L(x,y)), where k is the sharpening factor and L(x,y) is the local average brightness.

[0025] Preferably, the defect recognition unit includes:

[0026] A feature extraction sub-module, used to extract possible defect features from the image output by the image processing unit. Determine the location of potential defects by calculating the gray level change rate within the local area:

[0027] G(x,y) = |I(x + 1,y) - I(x - 1,y)| + |I(x,y + 1) - I(x,y - 1)|, where I(x,y) is the gray value of the image at the position (x,y);

[0028] A region segmentation sub-module, connected after the feature extraction sub-module, divides the image into multiple candidate regions according to the extracted features. Mark the regions with a gray level change rate greater than the set threshold T as candidate defect regions, that is, when G(x,y) > T, mark this region;

[0029] A defect confirmation sub-module, connected after the region segmentation sub-module, analyzes the candidate defect regions by calculating the shape features of the candidate regions where A is the area of the region and P is the perimeter of the region. When S < S0, S0 is the preset shape feature threshold, then confirm this region as a defect region;

[0030] A defect classification sub-module, connected after the defect confirmation sub-module, classifies the confirmed defect regions.

[0031] Preferably, the data analysis unit includes:

[0032] A defect statistics sub-module, used to collect the defect information provided by the defect recognition unit, count the number of defects of each type, and calculate the defect rate where N d is the number of defects and N t is the total number of detections;

[0033] The trend analysis submodule is connected to the defect statistics submodule and predicts possible problems in the future based on the change trend of the defect rate. r (t) = at + b to fit the defect rate variation trend over time t, where a and b are regression coefficients;

[0034] The control instruction generation submodule is connected to the trend analysis submodule and generates control instructions according to the prediction results. r (t) When the preset threshold value Dth is exceeded, a shutdown inspection instruction or an instruction to adjust the production line parameters is generated;

[0035] The instruction optimization submodule is connected to the control instruction generation submodule to evaluate the effect of the generated control instructions by calculating the change in the defect rate before and after the control ΔD. r =D r (t1)-D r (t0), where t1 and t0 are the time points after and before control, respectively. If ΔD r If it is less than the preset value, the control strategy is optimized and the control instructions are regenerated.

[0036] Preferably, the control execution unit includes:

[0037] The instruction parsing submodule is used to receive the control instruction from the data analysis unit, parse the instruction content, and determine the operation to be performed;

[0038] The parameter adjustment submodule is connected to the instruction parsing submodule and adjusts the corresponding parameters of the production line according to the parsed instructions. new =P old +ΔP calculates the new parameter value P new , where P old is the current parameter value, ΔP is the adjustment calculated according to the instruction;

[0039] The execution monitoring submodule is connected to the parameter adjustment submodule to monitor the production line status after parameter adjustment in real time, record the adjustment effect, and calculate the parameter adjustment result through the formula E=|P actual -P target |Calculate the actual parameter value P actual With the target parameter value P target The error E between

[0040] The feedback adjustment submodule is connected to the execution monitoring submodule, and performs feedback adjustment according to the monitoring result. When the error E is greater than the preset threshold Eth, the adjustment amount ΔP is recalculated and the adjustment is performed again until E≤Eth.

[0041] Preferably, the feedback monitoring unit comprises:

[0042] Real-time data collection submodule, used to collect real-time data after production line adjustment;

[0043] The status evaluation submodule is connected to the real-time data acquisition submodule and evaluates the current status of the production line according to the collected data. Calculate the comprehensive status score S, where x i is the actual value of the i-th parameter, w i is the weight of the i-th parameter, n is the total number of parameters;

[0044] The change monitoring submodule is connected to the state evaluation submodule to monitor the change of the state score. new -S old Calculate the change in state score ΔS, where S new is the current status score, S old is the last status score;

[0045] The information feedback submodule is connected to the change monitoring submodule, and feeds back the state score and its change amount to the data analysis unit. When ΔS exceeds a preset threshold ΔSth, a warning message is generated and sent to the data analysis unit.

[0046] Preferably, the learning optimization unit includes:

[0047] The data accumulation submodule is used to receive data from the information feedback submodule of the feedback monitoring unit, including the state score and its change amount, and store these data in the database to form a historical data set;

[0048] The pattern recognition submodule is connected to the data accumulation submodule, and identifies the key factors affecting the production line status by analyzing the historical data set, and uses the correlation analysis formula Calculate the correlation R between each parameter and the status score, where x i is the i-th parameter value, y i is the corresponding status score, and are the average values ​​of parameter values ​​and status scores, respectively;

[0049] The strategy adjustment submodule is connected to the pattern recognition submodule and adjusts the defect recognition and control strategy according to the identified key factors. new =P old +αRΔS calculates the new parameter value P new , where P old is the current parameter value, α is the learning rate, and ΔS is the change in the state score;

[0050] The effect verification submodule is connected to the strategy adjustment submodule, applies the adjusted strategy to the production line, and monitors the effect through the feedback monitoring unit. new +S target |Calculate the actual status score S new and the target state score S target When E is less than the preset threshold Eth, the optimization is considered successful, otherwise the strategy continues to be adjusted.

[0051] Preferably, the operation interface unit includes:

[0052] The status display submodule is used to display the status information of the production line in real time in the form of charts;

[0053] The parameter input submodule is connected to the status display submodule and allows the operator to input adjustment parameters through the formula P input =P default +ΔP user Calculate the parameter value P entered by the user input , where P default is the default parameter value, ΔP user is the adjustment amount entered by the user;

[0054] The instruction generation submodule is connected to the parameter input submodule and generates control instructions according to the parameters input by the user. input ) calculates a control instruction C, where f is a mapping function that generates a control instruction according to a parameter value;

[0055] The operation recording submodule is connected to the instruction generation submodule to record all operations of the operator, including the input parameter values ​​and the generated control instructions, and the formula R = (t, P input ,C) Record each operation, where t is the operation time, P input is the input parameter and C is the generated control instruction.

[0056] Preferably, the safety assurance unit comprises:

[0057] The authentication submodule is used to verify the identity of the operator by entering the user name and password. Calculate the verification code V, where H(P) is the hash value of the password P, and K is the preset key. Represents XOR operation;

[0058] The authority management submodule is connected to the identity authentication submodule and allocates operation permissions according to the verification results through the formula:

[0059] A=IF(V=V expected ,Afull ,A limited ) Determine the operation authority A, where V expected is the expected verification code, A full Indicates full permissions, A limited Indicates limited authority;

[0060] The operation audit submodule is connected to the permission management submodule, records all operation behaviors, and records each operation through the formula L=(t,U,O), where t is the operation time, U is the user name of the operator, and O is the operation content;

[0061] The anomaly detection submodule is connected to the operation audit submodule to detect abnormal operation behavior in real time and determine whether there is an anomaly through the formula D=IF(F(L)>T,1,0), where F(L) is the anomaly scoring function of the operation log L, T is the preset anomaly threshold, and when D=1, an alarm is triggered and the abnormal event is recorded.

[0062] Technical effects and advantages of the present invention: Compared with the prior art, the visual inspection and intelligent control system for PCBA production line proposed by the present invention has the following advantages:

[0063] The present invention realizes real-time intelligent control and decision-making on a high-speed production line by integrating multiple units such as image acquisition, image processing, defect recognition, data analysis, control execution, feedback monitoring, learning optimization, operation interface and security assurance. Through efficient image processing and fast defect recognition algorithm, the system can realize real-time detection and control on a high-speed production line. The use of advanced image processing and defect recognition technology significantly improves the accuracy and reliability of defect detection. Through real-time feedback and continuous learning optimization, the system can automatically adjust the control strategy according to changes in the production environment to improve production efficiency and product quality. Multi-layer security protection measures effectively prevent illegal access and operation to ensure the stable operation of the system. The system of the present invention has significant advantages in real-time intelligent control and decision-making on high-speed production lines, and can effectively improve the production efficiency and product quality of PCBA production lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The block diagram of the visual inspection and intelligent control system of the present invention for PCBA production line. DETAILED DESCRIPTION

[0065] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0066] The present invention provides Figure 1 The visual inspection and intelligent control system for PCBA production line shown in the figure realizes real-time intelligent control and decision-making on high-speed production line by integrating multiple units such as image acquisition, image processing, defect recognition, data analysis, control execution, feedback monitoring, learning optimization, operation interface and security assurance. Through efficient image processing and fast defect recognition algorithm, the system can realize real-time detection and control on high-speed production line. The advanced image processing and defect recognition technology are adopted to significantly improve the accuracy and reliability of defect detection. Through real-time feedback and continuous learning optimization, the system can automatically adjust the control strategy according to the changes in the production environment, improve production efficiency and product quality, and multi-layer security protection measures effectively prevent illegal access and operation to ensure the stable operation of the system. The details are as follows:

[0067] In this embodiment, the above-mentioned image acquisition unit is used to capture a high-definition image of a PCBA board on a production line; further, the image acquisition unit includes:

[0068] The light source adjustment submodule is used to dynamically adjust the light source intensity according to the ambient light conditions. The required light source intensity L is calculated by the formula L=αI+β, where I represents the current ambient light intensity, and α and β are coefficients pre-set according to the material properties of the PCBA board; this allows the image brightness to remain within the optimal range under different lighting conditions, avoiding overbrightness or overdarkness, and improving image clarity and contrast.

[0069] The high-definition camera submodule is connected to the light source adjustment submodule and uses the adjusted light source to capture high-definition images of the PCBA board; since the light source has been optimized, the captured images have higher resolution and better quality, providing high-quality raw data for subsequent image processing and defect identification.

[0070] The image stabilization submodule is connected to the high-definition camera submodule and uses the image frame difference method to reduce the image blur caused by the vibration of the production line. The pixel difference between two consecutive frames of images is calculated by ΔP=|P t -P t-1|, when ΔP>γ, γ is the preset threshold, and the image stabilization program is started until ΔP≤γ. This effectively reduces the image blur caused by production line vibration and improves the stability and clarity of the image.

[0071] The cooperation of the light source adjustment submodule and the high-definition camera submodule ensures the high quality of the image under different ambient lighting conditions, avoids the situation of being too bright or too dark, and improves the clarity and contrast of the image. The real-time adjustment of the image stabilization submodule reduces the image blur caused by the vibration of the production line, ensuring the stability and clarity of the image. The high-quality and stable image provides a reliable foundation for subsequent image processing and defect recognition, and improves the detection accuracy and reliability of the entire system.

[0072] In this embodiment, the image processing unit is connected to the image acquisition unit to pre-process the captured image to remove noise and enhance image quality; further, the image processing unit includes:

[0073] The noise filtering submodule is used to receive image data from the image acquisition unit and pass the formula Calculate the filtered pixel value I f (x,y), where w(i,j) is the Gaussian weight matrix, I(x,y) is the pixel value of the original image at position (x,y), and r is the filter radius; this effectively suppresses the noise in the image and improves the clarity and detail retention of the image.

[0074] The contrast enhancement submodule is connected to the noise filtering submodule to enhance the contrast of the denoised image. f The enhanced contrast C(x,y) is calculated by (x,y)-μ) / δ, where μ and δ are the image I f The mean and standard deviation of (x,y); this makes details in the image more visible and improves the accuracy of defect identification.

[0075] The edge sharpening submodule is connected to the contrast enhancement submodule to sharpen the edges of the image after contrast enhancement to highlight the boundary information in the image. The final sharpening effect E(x,y) is calculated by the formula E(x,y)=C(x,y)+k*(C(x,y)-L(x,y)), where k is the sharpening factor and L(x,y) is the local average brightness. This makes the boundaries in the image clearer and helps to accurately locate and identify defects.

[0076] Random noise in the image is effectively removed by a Gaussian filter, improving the clarity and detail retention of the image. By enhancing the contrast of the denoised image, the visibility of details in the image is improved, making defects easier to identify. By sharpening the edges of the image with enhanced contrast, the boundary information in the image is highlighted, contributing to the precise positioning and identification of defects.

[0077] In summary, through the collaborative work of three sub-modules: noise filtering, contrast enhancement, and edge sharpening, the image processing unit significantly improves the preprocessing effect of the image, removes noise, and enhances the image quality, providing a high-quality image basis for subsequent defect identification, thereby improving the detection accuracy and reliability of the entire system.

[0078] In this embodiment, the above-mentioned defect recognition unit is connected after the image processing unit and identifies potential defects on the PCBA board based on the processed image; further, the defect recognition unit includes:

[0079] A feature extraction sub-module, which is used to extract possible defect features from the image output by the image processing unit. By calculating the gray-scale change rate within a local area:

[0080] G(x,y) = |I(x + 1,y) - I(x - 1,y)| + |I(x,y + 1) - I(x,y - 1)| to determine the position of potential defects, where I(x,y) is the gray-scale value of the image at position (x,y); this step can effectively identify possible defect areas in the image and provide preliminary positioning information for subsequent processing.

[0081] A region segmentation sub-module, connected after the feature extraction sub-module, divides the image into multiple candidate regions according to the extracted features. By setting a threshold T, regions with a gray-scale change rate greater than T are marked as candidate defect regions, that is, when G(x,y) > T, mark this region; this step further narrows the scope of potential defects through threshold screening, improving the efficiency and accuracy of subsequent processing.

[0082] A defect confirmation sub-module, connected after the region segmentation sub-module, analyzes the candidate defect regions by calculating the shape features of the candidate regions where A is the area of the region and P is the perimeter of the region. When S < S0, S0 is a preset shape feature threshold, then confirm this region as a defect region; this step further confirms whether the candidate region is a real defect through the analysis of shape features, reducing misjudgment and missed detection, and improving the accuracy of defect recognition.

[0083] A defect classification sub-module, connected after the defect confirmation sub-module, classifies the confirmed defect regions.

[0084] By analyzing the color distribution and texture features of the defect area, the pre-trained decision tree structure is used to determine the defect type. This step can distinguish different defect types and provide specific guidance for subsequent processing and repair.

[0085] By calculating the grayscale change rate in the local area, the potential defect features are effectively extracted, providing preliminary positioning information for subsequent processing. By setting a threshold, the area with a grayscale change rate greater than the threshold is marked as a candidate defect area, which narrows the scope of potential defects and improves processing efficiency. By calculating the shape features of the candidate area, it is further confirmed whether the candidate area is a real defect, reducing misjudgment and missed detection, and improving the accuracy of defect identification. By classifying the confirmed defective areas, different defect types are distinguished, providing specific guidance for subsequent processing and repair.

[0086] In this embodiment, the data analysis unit is connected to the defect recognition unit to analyze the defect data and generate corresponding control instructions; further, the data analysis unit includes:

[0087] The defect statistics submodule is used to collect defect information provided by the defect identification unit, count the number of defects of each type, and calculate the defect rate Where N d is the number of defects, N t is the total number of inspections; this step can fully understand the defect distribution on the production line and provide basic data for subsequent trend analysis and control instruction generation.

[0088] The trend analysis submodule is connected to the defect statistics submodule and predicts possible problems in the future based on the change trend of the defect rate. r (t)=at+b is used to fit the changing trend of defect rate over time t, where a and b are regression coefficients. This step can detect potential problems in advance and provide a scientific basis for preventive maintenance and timely adjustment of production line parameters.

[0089] The control instruction generation submodule is connected to the trend analysis submodule and generates control instructions according to the prediction results. r (t) When the preset threshold Dth is exceeded, a shutdown inspection instruction or an instruction to adjust the production line parameters is generated; this step enables timely measures to prevent the defect rate from further increasing and ensure the stable operation of the production line.

[0090] The instruction optimization submodule is connected to the control instruction generation submodule to evaluate the effect of the generated control instructions by calculating the change in the defect rate before and after the control ΔD. r =D r(t1)-D r (t0), where t1 and t0 are the time points after and before control, respectively. If ΔD r If it is less than the preset value, the control strategy is optimized and the control instructions are regenerated. This step can ensure the effectiveness of the control instructions and improve the system's adaptive ability through continuous optimization and adjustment.

[0091] In summary, the data analysis unit significantly improves the ability to analyze defect data and generate effective control instructions through the collaborative work of four sub-modules: defect statistics, trend analysis, control instruction generation, and instruction optimization, thereby optimizing the operating status of the PCBA production line.

[0092] In this embodiment, the control execution unit is connected to the data analysis unit to receive and execute control instructions to adjust the working state of the production line; further, the control execution unit includes:

[0093] The instruction parsing submodule is used to receive control instructions from the data analysis unit, parse the instruction content, and determine the operations to be performed; this step ensures that the control instructions can be correctly understood and executed, and provides accurate operational guidance for subsequent parameter adjustment and monitoring.

[0094] The parameter adjustment submodule is connected to the instruction parsing submodule and adjusts the corresponding parameters of the production line according to the parsed instructions. new =P old +ΔP calculates the new parameter value P new , where P old is the current parameter value, and ΔP is the adjustment amount calculated according to the instruction; this step can accurately adjust the production line parameters to ensure the execution effect of the control instruction.

[0095] The execution monitoring submodule is connected to the parameter adjustment submodule to monitor the production line status after parameter adjustment in real time, record the adjustment effect, and calculate the parameter adjustment result through the formula E=|P actual -P target |Calculate the actual parameter value P actual With the target parameter value P target This step can monitor the adjustment effect in real time to ensure that the parameter adjustment achieves the expected goal.

[0096] The feedback adjustment submodule is connected to the execution monitoring submodule, and performs feedback adjustment according to the monitoring results. When the error E is greater than the preset threshold Eth, the adjustment amount ΔP is recalculated and the adjustment is performed again until E≤Eth. This step ensures the accuracy and stability of parameter adjustment through a closed-loop feedback mechanism, avoiding problems caused by inaccurate initial adjustment.

[0097] The instruction parsing submodule ensures that the control instructions are correctly understood and executed, providing accurate operation guidance for subsequent parameter adjustment and monitoring. The parameter adjustment submodule accurately adjusts the production line parameters according to the parsed instructions to ensure the execution effect of the control instructions. The execution monitoring submodule monitors the status of the production line after parameter adjustment in real time, records the adjustment effect, and ensures that the parameter adjustment achieves the expected goal. The feedback adjustment submodule performs feedback adjustment based on the monitoring results to ensure the accuracy and stability of parameter adjustment and avoid problems caused by inaccurate initial adjustment.

[0098] In this embodiment, the feedback monitoring unit is connected to the control execution unit to monitor the state of the adjusted production line and feed back the information to the data analysis unit; further, the feedback monitoring unit includes:

[0099] The real-time data acquisition submodule is used to collect real-time data after the production line is adjusted; this step ensures the real-time and accuracy of the data and provides reliable data support for subsequent status evaluation.

[0100] The status evaluation submodule is connected to the real-time data acquisition submodule and evaluates the current status of the production line according to the collected data. Calculate the comprehensive status score S, where x i is the actual value of the i-th parameter, w i is the weight of the ith parameter, n is the total number of parameters;

[0101] This step can comprehensively consider the impact of multiple parameters, fully evaluate the current status of the production line, and provide a scientific basis for subsequent monitoring and feedback.

[0102] The change monitoring submodule is connected to the state evaluation submodule to monitor the change of the state score. new -S old Calculate the change in state score ΔS, where S new is the current status score, S old is the last status score; this step can timely discover the changing trend of the status score and provide early warning information for timely adjustment of production line parameters.

[0103] The information feedback submodule is connected to the change monitoring submodule, and feeds back the status score and its change amount to the data analysis unit. When ΔS exceeds the preset threshold ΔSth, a warning message is generated and sent to the data analysis unit. This step ensures that the data analysis unit can receive early warning information of production line status changes in a timely manner through a closed-loop feedback mechanism, so as to take corresponding control measures.

[0104] The real-time data acquisition submodule ensures the real-time and accuracy of the data, providing reliable data support for subsequent status evaluation. The status evaluation submodule comprehensively considers the influence of multiple parameters and comprehensively evaluates the current status of the production line, providing a scientific basis for subsequent monitoring and feedback. The change monitoring submodule promptly discovers the changing trend of the status score, providing early warning information for timely adjustment of production line parameters. The information feedback submodule feeds back the status score and its change amount to the data analysis unit. When the change amount exceeds the preset threshold, a warning message is generated, ensuring that the data analysis unit can take corresponding control measures in a timely manner.

[0105] In this embodiment, the above-mentioned learning optimization unit is connected to the feedback monitoring unit to optimize the subsequent defect identification and control strategy according to the feedback information; further, the learning optimization unit includes:

[0106] The data accumulation submodule is used to receive data from the information feedback submodule of the feedback monitoring unit, including the status score and its change amount, and store these data in the database to form a historical data set; this step ensures the integrity and traceability of the data, and provides rich data support for subsequent analysis and optimization.

[0107] The pattern recognition submodule is connected to the data accumulation submodule, and identifies the key factors affecting the production line status by analyzing the historical data set, and uses the correlation analysis formula Calculate the correlation R between each parameter and the status score, where x i is the i-th parameter value, y i is the corresponding status score, and are the average values ​​of parameter values ​​and status scores respectively; this step can identify the key factors that have the greatest impact on the status of the production line, providing a scientific basis for subsequent strategy adjustments.

[0108] The strategy adjustment submodule is connected to the pattern recognition submodule and adjusts the defect recognition and control strategy according to the identified key factors. new =P old +αRΔS calculates the new parameter value P new , where P old is the current parameter value, α is the learning rate, and ΔS is the change in the state score; this step can dynamically adjust the control strategy according to the influence of key factors to improve the adaptability and response speed of the production line.

[0109] The effect verification submodule is connected to the strategy adjustment submodule, applies the adjusted strategy to the production line, and monitors the effect through the feedback monitoring unit. new +Starget |Calculate the actual status score S new and the target state score S target When E is less than the preset threshold Eth, the optimization is considered successful, otherwise the strategy continues to be adjusted. This step ensures the effectiveness and continuous improvement of the optimization strategy through a closed-loop feedback mechanism, avoiding problems caused by inaccurate adjustments.

[0110] The learning optimization unit significantly improves the utilization efficiency of feedback information through the collaborative work of four sub-modules: data accumulation, pattern recognition, strategy adjustment, and effect verification, optimizes subsequent defect identification and control strategies, and further improves the performance of the production line and product quality.

[0111] In this embodiment, the above-mentioned operation interface unit is connected to the learning optimization unit to provide a human-computer interaction interface, which is convenient for operators to monitor the system operation status and input adjustment parameters; further, the operation interface unit includes:

[0112] The status display submodule is used to display the status information of the production line in real time in the form of charts. This step enables operators to intuitively understand the current status of the production line, including key indicators such as defect rate, production speed, temperature, humidity, etc. The display method in the form of charts makes the information easier to understand and improves the monitoring efficiency of operators.

[0113] The parameter input submodule is connected to the status display submodule and allows the operator to input adjustment parameters through the formula P input =P default +ΔP user Calculate the parameter value P entered by the user input , where P default is the default parameter value, ΔP user is the adjustment amount input by the user; this step enables the operator to flexibly adjust parameters according to the current status, improving the adaptability of the system and the convenience of operation.

[0114] The instruction generation submodule is connected to the parameter input submodule and generates control instructions according to the parameters input by the user. input ) calculates the control instruction C, where f is a mapping function that generates a control instruction based on the parameter value; this step ensures that the parameters input by the user can be converted into specific control instructions, realizes the operator's intention, and improves the response speed and accuracy of the system.

[0115] The operation recording submodule is connected to the instruction generation submodule to record all operations of the operator, including the input parameter values ​​and the generated control instructions, and the formula R = (t, P input,C) Record each operation, where t is the operation time, P input is the input parameter, and C is the generated control instruction. This step ensures the traceability of all operations, facilitates subsequent auditing and analysis, and improves the transparency and reliability of the system.

[0116] The operation interface unit significantly improves the friendliness of human-computer interaction and the convenience of operation through the collaborative work of four sub-modules: status display, parameter input, command generation and operation recording, ensuring that operators can easily monitor the system operation status and make necessary adjustments.

[0117] In this embodiment, the above-mentioned security unit is connected to the operation interface unit to ensure the safe operation of the system and prevent illegal access and operation. Further, the security unit includes:

[0118] The authentication submodule is used to verify the identity of the operator by entering the user name and password. Calculate the verification code V, where H(P) is the hash value of the password P, and K is the preset key. Represents XOR operation; this step ensures that only legitimate operators can access the system, preventing unauthorized access.

[0119] The authority management submodule is connected to the identity authentication submodule and allocates operation permissions according to the verification results through the formula:

[0120] A=IF(V=V expected ,A full ,A limited ) Determine the operation authority A, where V expected is the expected verification code, A full Indicates full permissions, A limited Indicates limited authority; this step ensures that operators at different levels can only perform operations consistent with their authority, preventing unauthorized operations.

[0121] The operation audit submodule is connected to the permission management submodule to record all operation behaviors and record each operation through the formula L=(t,U,O), where t is the operation time, U is the operator's user name, and O is the operation content; this step ensures the traceability of all operations, facilitates subsequent auditing and analysis, and improves the transparency and reliability of the system.

[0122] The anomaly detection submodule is connected to the operation audit submodule to detect abnormal operation behaviors in real time and determine whether there is an anomaly by using the formula D=IF(F(L)>T,1,0), where F(L) is the anomaly scoring function of the operation log L, and T is the preset anomaly threshold. When D=1, an alarm is triggered and the abnormal event is recorded. This step can detect and handle abnormal operations in a timely manner to prevent potential security risks.

[0123] The authentication submodule ensures that only legitimate operators can access the system, preventing unauthorized access. The authority management submodule allocates operation permissions based on the verification results, ensuring that operators at different levels can only perform operations consistent with their permissions, preventing unauthorized operations. The operation audit submodule records all operation behaviors, ensuring the traceability of all operations, facilitating subsequent audits and analysis, and improving the transparency and reliability of the system. The anomaly detection submodule detects abnormal operation behaviors in real time, promptly discovers and handles abnormal operations, and prevents potential security risks.

[0124] In summary, the security assurance unit significantly improves the security of the system through the collaborative work of four sub-modules: identity authentication, permission management, operation auditing, and anomaly detection, ensures the safe operation of the system, prevents illegal access and operation, and improves the stability and reliability of the system.

[0125] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A visual inspection and intelligent control system for PCBA production line, characterized in that: include: Image acquisition unit, used to capture high-definition images of PCBA boards on the production line; An image processing unit, connected to the image acquisition unit, preprocesses the captured image to remove noise and enhance image quality; A defect recognition unit, connected to the image processing unit, recognizes potential defects on the PCBA board based on the processed image; A data analysis unit, connected to the defect recognition unit, analyzes defect data and generates corresponding control instructions; A control execution unit, connected to the data analysis unit, receives and executes control instructions to adjust the working state of the production line; A feedback monitoring unit, connected to the control execution unit, monitors the state of the adjusted production line and feeds back the information to the data analysis unit; A learning optimization unit, connected to the feedback monitoring unit, optimizes subsequent defect identification and control strategies according to feedback information; An operation interface unit, connected to the learning optimization unit, provides a human-computer interaction interface to facilitate operators to monitor the system operation status and input adjustment parameters; The security unit is connected to the operation interface unit to ensure the safe operation of the system and prevent illegal access and operation.

2. According to claim 1, a visual inspection and intelligent control system for a PCBA production line is characterized in that: The image acquisition unit comprises: The light source adjustment submodule is used to dynamically adjust the light source intensity according to the ambient light conditions. The required light source intensity L is calculated by the formula L = αI + β, where I represents the current ambient light intensity, and α and β are coefficients pre-set according to the material characteristics of the PCBA board; A high-definition camera submodule is connected to the light source adjustment submodule to capture a high-definition image of the PCBA board using the adjusted light source; The image stabilization submodule is connected to the high-definition camera submodule and uses the image frame difference method to reduce the image blur caused by the vibration of the production line. The pixel difference between two consecutive frames of images is calculated by ΔP=|P t -P t-1 |, when ΔP>γ, γ is the preset threshold, and the image stabilization program is started until ΔP≤γ.

3. A visual inspection and intelligent control system for a PCBA production line according to claim 2, characterized in that: The image processing unit comprises: The noise filtering submodule is used to receive image data from the image acquisition unit and pass the formula Calculate the filtered pixel value I f (x,y), where w(i,j) is the Gaussian weight matrix, I(x,y) is the pixel value of the original image at position (x,y), and r is the filter radius; The contrast enhancement submodule is connected to the noise filtering submodule to enhance the contrast of the denoised image. f The enhanced contrast C(x,y) is calculated by (x,y)-μ) / δ, where μ and δ are the image I f The mean and standard deviation of (x,y); The edge sharpening submodule is connected to the contrast enhancement submodule to perform edge sharpening on the image after contrast enhancement to highlight the boundary information in the image. The final sharpening effect E(x,y) is calculated by the formula E(x,y)=C(x,y)+k*(C(x,y)-L(x,y)), where k is the sharpening factor and L(x,y) is the local average brightness.

4. A visual inspection and intelligent control system for a PCBA production line according to claim 3, characterized in that: The defect recognition unit comprises: The feature extraction submodule is used to extract possible defect features from the image output by the image processing unit by calculating the grayscale change rate in the local area: G(x,y)=|I(x+1,y)-I(x-1,y)|+|I(x,y+1)-I(x,y-1) to determine the location of the potential defect, where I(x,y) is the grayscale value of the image at position (x,y); The region segmentation submodule is connected to the feature extraction submodule, and the image is segmented into multiple candidate regions according to the extracted features. The region with a grayscale change rate greater than T is marked as a candidate defect region by setting a threshold T, that is, when G(x,y)>T, the region is marked; The defect confirmation sub-module, connected after the region segmentation sub-module, analyzes the candidate defect regions by calculating the shape features of the candidate regions where A is the area of the region, P is the perimeter of the region, and when S < S0, S0 being a preset shape feature threshold, then confirm that the region is a defect region; The defect classification submodule is connected to the defect confirmation submodule and classifies the confirmed defect areas.

5. A visual inspection and intelligent control system for a PCBA production line according to claim 4, characterized in that: The data analysis unit comprises: The defect statistics submodule is used to collect defect information provided by the defect identification unit, count the number of defects of each type, and calculate the defect rate Where N d is the number of defects, N t is the total number of tests; The trend analysis submodule is connected to the defect statistics submodule and predicts possible problems in the future based on the change trend of the defect rate. r (t) = at + b to fit the defect rate variation trend over time t, where a and b are regression coefficients; The control instruction generation submodule is connected to the trend analysis submodule and generates control instructions according to the prediction results. r (t) When the preset threshold value Dth is exceeded, a shutdown inspection instruction or an instruction to adjust the production line parameters is generated; The instruction optimization submodule is connected to the control instruction generation submodule to evaluate the effect of the generated control instructions by calculating the change in the defect rate before and after the control ΔD. r =D r (t1)-D r (t0), where t1 and t0 are the time points after and before control, respectively. If ΔD r If it is less than the preset value, the control strategy is optimized and the control instructions are regenerated.

6. A visual inspection and intelligent control system for a PCBA production line according to claim 5, characterized in that: The control execution unit comprises: The instruction parsing submodule is used to receive the control instruction from the data analysis unit, parse the instruction content, and determine the operation to be performed; The parameter adjustment submodule is connected to the instruction parsing submodule and adjusts the corresponding parameters of the production line according to the parsed instructions. new =P old +ΔP calculates the new parameter value P new , where P old is the current parameter value, ΔP is the adjustment calculated according to the instruction; The execution monitoring submodule is connected to the parameter adjustment submodule to monitor the production line status after parameter adjustment in real time, record the adjustment effect, and calculate the parameter adjustment result through the formula E=|P actual -P target |Calculate the actual parameter value P actual With the target parameter value P target The error E between The feedback adjustment submodule is connected to the execution monitoring submodule, and performs feedback adjustment according to the monitoring result. When the error E is greater than the preset threshold Eth, the adjustment amount ΔP is recalculated and the adjustment is performed again until E≤Eth.

7. A visual inspection and intelligent control system for a PCBA production line according to claim 6, characterized in that: The feedback monitoring unit comprises: Real-time data collection submodule, used to collect real-time data after production line adjustment; The status evaluation submodule is connected to the real-time data acquisition submodule and evaluates the current status of the production line according to the collected data. Calculate the comprehensive status score S, where x i is the actual value of the i-th parameter, w i is the weight of the i-th parameter, n is the total number of parameters; The change monitoring submodule is connected to the state evaluation submodule to monitor the change of the state score. new -S old Calculate the change in state score ΔS, where S new is the current status score, S old is the last status score; The information feedback submodule is connected to the change monitoring submodule, and feeds back the state score and its change amount to the data analysis unit. When ΔS exceeds a preset threshold ΔSth, a warning message is generated and sent to the data analysis unit.

8. The visual inspection and intelligent control system for a PCBA production line according to claim 7, characterized in that: The learning optimization unit comprises: The data accumulation submodule is used to receive data from the information feedback submodule of the feedback monitoring unit, including the state score and its change amount, and store these data in the database to form a historical data set; The pattern recognition submodule is connected to the data accumulation submodule, and identifies the key factors affecting the production line status by analyzing the historical data set, and uses the correlation analysis formula Calculate the correlation R between each parameter and the status score, where x i is the i-th parameter value, y i is the corresponding status score, and are the average values ​​of parameter values ​​and status scores, respectively; The strategy adjustment submodule is connected to the pattern recognition submodule and adjusts the defect recognition and control strategy according to the identified key factors. new =P old +αRΔS calculates the new parameter value P new , where P old is the current parameter value, α is the learning rate, and ΔS is the change in the state score; The effect verification submodule is connected to the strategy adjustment submodule, applies the adjusted strategy to the production line, and monitors the effect through the feedback monitoring unit. new +S target |Calculate the actual status score S new and the target state score S target When E is less than the preset threshold Eth, the optimization is considered successful, otherwise the strategy continues to be adjusted.

9. A visual inspection and intelligent control system for a PCBA production line according to claim 8, characterized in that: The operation interface unit comprises: The status display submodule is used to display the status information of the production line in real time in the form of charts; The parameter input submodule is connected to the status display submodule and allows the operator to input adjustment parameters through the formula P input =P default +ΔP user Calculate the parameter value P entered by the user input , where P default is the default parameter value, ΔP user is the adjustment amount entered by the user; The instruction generation submodule is connected to the parameter input submodule and generates control instructions according to the parameters input by the user. input ) calculates a control instruction C, where f is a mapping function that generates a control instruction according to a parameter value; The operation recording submodule is connected to the instruction generation submodule to record all operations of the operator, including the input parameter values ​​and the generated control instructions, and the formula R = (t, P input ,C) Record each operation, where t is the operation time, P input is the input parameter and C is the generated control instruction.

10. The visual inspection and intelligent control system for a PCBA production line according to claim 9, characterized in that: The safety assurance unit comprises: The authentication submodule is used to verify the identity of the operator by entering the user name and password. Calculate the verification code V, where H(P) is the hash value of the password P, and K is the preset key. Represents the exclusive OR operation; The authority management submodule is connected to the identity authentication submodule and allocates operation permissions according to the verification results through the formula: A=IF(V=V expected ,A full ,A limited ) Determine the operation authority A, where V expected is the expected verification code, A full Indicates full permissions, A limited Indicates limited authority; The operation audit submodule is connected to the permission management submodule, records all operation behaviors, and records each operation through the formula L=(t,U,O), where t is the operation time, U is the user name of the operator, and O is the operation content; The anomaly detection submodule is connected to the operation audit submodule to detect abnormal operation behavior in real time and determine whether there is an anomaly through the formula D=IF(F(L)>T,1,0), where F(L) is the anomaly scoring function of the operation log L, T is the preset anomaly threshold, and when D=1, an alarm is triggered and the abnormal event is recorded.

Images