A phased quality detection method and system for SMT patch semi-finished products

Abstract

This invention discloses a staged quality inspection method and system for SMT (Surface Mount Technology) semi-finished products, specifically relating to the field of electronic manufacturing quality management technology. The method includes the following steps: quality inspection is performed at each stage of the production line, the inspection results are recorded and a result set is formed, and early signs of production rhythm imbalance are identified using time interval distribution analysis; if imbalance is present, the defect frequency fluctuation index and the production line dynamic disorder index are further calculated to quantify the dynamic fluctuations and systemic problems of non-conforming phenomena; the two indices are input into a pre-trained machine learning model, and the model output distinguishes the quality status of the production line as high quality or low quality; for low-quality production lines, optimization is performed according to a preset adjustment strategy; this method can monitor the quality status of the production line in real time, accurately distinguish between accidental and non-accidental problems, improve quality management efficiency, ensure stable operation of the production line, and reduce the proportion of non-conforming products.

Description

Technical Field

[0001] This invention relates to the field of electronic manufacturing quality management technology, and more specifically, to a staged quality inspection method and system for SMT (Surface Mount Technology) semi-finished products. Background Technology

[0002] As electronic products demand higher performance and smaller sizes, surface mount technology (SMT) has become a key process in modern electronics manufacturing. In an SMT production line, automated equipment completes a series of steps, from solder paste printing and component placement to soldering. However, quality problems (such as defective products) can frequently occur on the production line due to the complexity of the manufacturing process and the instability of equipment operation.

[0003] The main problems with current production lines include: Difficulty in dynamic quality management: SMT assembly lines require continuous, high-volume production. Product quality is affected by various factors such as equipment status, production rhythm, and material characteristics. Fluctuations in production rhythm and decreased equipment performance can easily lead to an increase in defective products. Inability to accurately determine the nature of quality problems: Defective products may be due to accidental issues (such as transient errors) or non-accidental issues (such as systemic defects). Traditional testing methods struggle to distinguish between the two, easily leading to over- or under-adjustment of the production line. Therefore, this paper proposes a staged quality inspection method and system for SMT semi-finished products to address the above problems. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A staged quality inspection method for SMT (Surface Mount Technology) semi-finished products includes the following steps:

[0006] Quality inspections are performed at each stage of the production of SMT semi-finished products, and the results of each stage are recorded to obtain a set of inspection results for each stage.

[0007] Perform time interval distribution analysis on the detection result set, and identify early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results;

[0008] If there are early signs of production rhythm imbalance in the corresponding production stage, defect analysis and disorder analysis are performed on the production line to generate a defect frequency fluctuation index and a production line dynamic disorder index, respectively.

[0009] The defect frequency fluctuation index and the production line dynamic disorder index are fed into the pre-trained machine learning model. Based on the output results, the corresponding production stages are divided into high-quality production lines and low-quality production lines. The low-quality production lines are adjusted according to the preset adjustment strategy.

[0010] The production stages of SMT semi-finished products include: printing stage, mounting stage, and soldering stage. Quality inspection is carried out at each production stage in accordance with the preset quality inspection strategy for the corresponding production stage.

[0011] In a preferred embodiment, performing time interval distribution analysis on the detection result set refers to:

[0012] During the production line inspection process, the set of inspection time points where all quality inspection results are marked as non-conforming products is obtained and arranged in chronological order to obtain the inspection time sequence of non-conforming products.

[0013] The time interval is calculated for adjacent time points in the detection time series to generate a time interval distribution dataset. The time interval represents the length of the time interval between two consecutive detections. Then, feature extraction is performed on the time interval distribution dataset to obtain statistical characteristics including the following:

[0014] The average time interval is used to reflect the average inspection rate of the production line. The average time interval refers to the average of all statistical time intervals.

[0015] The fluctuation range of the time interval is used to reflect the stability of the production rhythm. The fluctuation range of the time interval refers to the maximum time interval minus the minimum time interval.

[0016] The time interval anomaly ratio is used to identify the frequency of abnormal rhythmic behavior. The time interval anomaly ratio refers to the proportion of time intervals that are outside the preset time standard range to the total number of time intervals.

[0017] In a preferred embodiment, identifying early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results refers to:

[0018] An abnormal signal is generated when the average value of the time interval is not within the preset normal rhythm range;

[0019] An abnormal signal is generated when the fluctuation range of the time interval is less than a preset threshold.

[0020] When the proportion of abnormal points in the time interval exceeds a preset limit, an abnormal signal is generated;

[0021] When an abnormal signal is generated, it indicates that there are early signs of production rhythm imbalance in the corresponding production stage; when no abnormal signal is generated, it indicates that there are no early signs of production rhythm imbalance in the corresponding production stage.

[0022] In a preferred embodiment, the logic for obtaining the defect frequency fluctuation index is as follows:

[0023] Within a time interval consisting of N time windows, obtain the non-compliance frequency fi in time window i, calculate the mean of the non-compliance frequencies of all time windows as μDF, and calculate the standard deviation of the non-compliance frequencies of all time windows as σDF.

[0024] Calculate the rate of change of the non-conforming frequency fi', fi' = fi - fj; j = i - 1; then calculate the average absolute value of the rate of change fi'. Then substitute the values ​​into the following calculation formula:

[0025] C is a non-negative constant, γ is a preset non-zero adjustment factor, and IDF is the defect frequency fluctuation index.

[0026] In a preferred embodiment, the logic for obtaining the production line dynamic disorder index is as follows:

[0027] Within a time interval consisting of N time windows, the start time of each test, the actual number of tests conducted within a preset standard unit time, and the test results (pass / fail markings) are obtained from the test result set for all products in the test time series. Then, the time interval between the start times of each test is calculated to obtain the time interval test dataset. The standard deviation of the time interval test dataset and the standard deviation reflecting the fluctuation of the actual number of tests conducted within a standard unit time are calculated.

[0028] Calculate the ratio of qualified products to actual inspection quantity in each time window i to obtain the window pass rate of time window i, and calculate the standard deviation of the pass rates of all windows.

[0029] Perform the following calculations:

[0030] μΔt is the standard deviation of the time interval detection dataset, σHG is the standard deviation of the pass rate of all windows, σQ is the standard deviation reflecting the fluctuation of the actual number of tests per standard unit time, and ILD is the production line dynamic disorder index.

[0031] In a preferred embodiment, the logic for using the machine learning model is as follows:

[0032] The defect frequency fluctuation index and the production line dynamic disorder index are substituted into the pre-trained convolutional neural network model. The convolutional neural network model outputs either 1 or 0. An output of 1 indicates that the corresponding production stage is classified as a high-quality production line, which means that the non-conforming production situation of the production line in this production stage is accidental. An output of 0 indicates that the corresponding production stage is classified as a low-quality production line, which means that the non-conforming production situation of the production line in this production stage is not accidental.

[0033] In a preferred embodiment, a staged quality inspection system for SMT (Surface Mount Technology) semi-finished products includes:

[0034] The quality inspection module performs quality inspections on each stage of the SMT assembly semi-finished product production process and records the quality inspection results for each stage, thus obtaining a set of inspection results for each stage.

[0035] The early signs identification module performs time interval distribution analysis on the detection result set and identifies whether there are early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results.

[0036] The production line analysis module performs defect analysis and disorder analysis on the production line if there are early signs of production rhythm imbalance in the corresponding production stage, generating a defect frequency fluctuation index and a production line dynamic disorder index respectively.

[0037] The production line classification module inputs the defect frequency fluctuation index and the production line dynamic disorder index into a pre-trained machine learning model, and classifies the corresponding production stages into high-quality production lines and low-quality production lines based on the output results.

[0038] The production line adjustment module adjusts low-quality production lines according to a preset adjustment strategy.

[0039] The technical effects and advantages of this invention are as follows:

[0040] By analyzing time interval distribution, calculating and judging defect frequency fluctuation index, and production line dynamic disorder index, this invention can distinguish between accidental and systemic problems in non-conformities. Accidental problems: avoid excessive intervention, maintain normal production line operation, and improve production efficiency. Non-accidental problems: promptly identify systemic hidden dangers, locate the root cause, and reduce unnecessary quality losses. This invention, by monitoring the frequency and rhythm fluctuations of non-conformities on the production line in real time, can capture potential problems in production line operation, generate early warning signals, improve the dynamic response capability of the production line, avoid the accumulation of problems leading to large-scale shutdowns or rework, and achieve intelligent monitoring of the entire production line quality status.

[0041] By incorporating the defect frequency fluctuation index and dynamic disorder index into a machine learning model, this invention can intelligently determine the quality status of the production stage without human intervention, reducing the subjectivity and error of manual judgment, lowering the cost of manual intervention, and improving inspection efficiency. This invention achieves precise classification of high-quality and low-quality production lines: For high-quality production lines, minimal adjustments are needed, reducing unnecessary resource waste. For low-quality production lines, preset optimization strategies (such as shutdown for maintenance, process adjustments, etc.) are activated to concentrate resources on solving quality problems, achieving scientific resource allocation and avoiding losses from complete line shutdowns.

[0042] This invention significantly reduces the proportion of defective products by precisely monitoring and optimizing the production line quality management process: it improves the overall product qualification rate of the production line, ensures that the final product quality meets design requirements, reduces rework rates and raw material waste, and enhances production efficiency. This invention covers multiple stages of the SMT (Surface Mount Technology) production line, including solder paste printing, component mounting, and soldering. Quality inspection data at each stage is recorded and analyzed, ensuring quality controllability throughout the entire production process. It can quickly locate the source of defects and reduce troubleshooting time. This invention is applicable to various electronic product manufacturing scenarios (such as mobile phones, automotive electronics, and smart devices). Through intelligent management, it reduces downtime for maintenance, lowers rework costs and material losses due to defective products, improves production efficiency and yield, and reduces overall production line operating costs. Attached Figure Description

[0043] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0044] Figure 1 This is a schematic diagram of a staged quality inspection method for SMT semi-finished products according to the present invention.

[0045] Figure 2 This is a schematic diagram of a staged quality inspection system for SMT semi-finished products according to the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Reference Figure 1 - Figure 2 The following examples were obtained:

[0048] Example 1: A staged quality inspection method for SMT semi-finished products, comprising the following steps:

[0049] Quality inspections are conducted at every stage of the SMT (Surface Mount Technology) semi-finished product production process, and the results are recorded for each stage, resulting in a set of inspection results for each production stage. By independently inspecting all production stages, including printing, mounting, and soldering, potential problems can be precisely identified, ensuring that every stage of the production line meets quality requirements. The quality inspection results for each stage are systematically recorded, forming a set of inspection results that provides a reliable data foundation for subsequent analysis and hazard identification. Early detection and resolution of potential problems during production reduce finished product scrap and rework.

[0050] Interval distribution analysis is performed on the detection result set to identify early signs of production rhythm imbalance in the corresponding production stages. By analyzing the detection time series, the changing characteristics of the production rhythm are identified, and the trend of rhythm imbalance is determined. The statistical characteristics of the detection time intervals (such as average value, fluctuation range, and outlier ratio) can promptly reveal potential equipment problems, human error, or unreasonable process design, providing quantitative basis for production line adjustments and preventing problems from escalating.

[0051] If early signs of production rhythm imbalance appear in the corresponding production stage, defect analysis and disorder analysis are performed on the production line to generate a defect frequency fluctuation index and a production line dynamic disorder index, respectively. The defect frequency fluctuation index reflects the frequency and fluctuation of non-conforming products, helping to assess whether defects are accidental or systematic. The dynamic disorder index reflects the overall operational stability of the production line through a comprehensive analysis of time intervals, fluctuations in the number of inspections, and changes in the pass rate. The combination of the two indices can more comprehensively describe the current degree of abnormality of the production line and whether it is accidental, providing a basis for subsequent decision-making.

[0052] The defect frequency fluctuation index and the production line dynamic disorder index are fed into a pre-trained machine learning model. Based on the output, the corresponding production stages are divided into high-quality production lines and low-quality production lines. Low-quality production lines are then adjusted according to a pre-set adjustment strategy. Using the machine learning model, the quality status of the production line can be quickly and accurately identified, avoiding errors from subjective judgment or manual analysis. Classification Standardization: The output is 1 (high quality) or 0 (low quality), providing a clear classification for production line status assessment. High-quality production line: Indicates that the non-conformity is an occasional problem, requiring little intervention. Low-quality production line: Indicates that the non-conformity is systemic, requiring focused adjustment. Taking adjustment measures for low-quality production lines can effectively reduce the risk of large-scale production errors. Efficient Problem Solving: Through pre-set adjustment strategies (such as downtime maintenance, equipment calibration, and personnel operation improvements), identified problems can be quickly resolved, ensuring production efficiency and quality. Improved Production Line Stability: The implementation of adjustment measures ensures that the production line can operate continuously in a stable state, thereby improving overall production efficiency.

[0053] The production stages of SMT (Surface Mount Technology) semi-finished products include: printing, mounting, and soldering. Quality inspection is conducted at each stage according to the pre-defined quality inspection strategy for that stage. The processing steps for semi-finished products include the following:

[0054] Printing stage (solder paste printing): The product is a semi-finished product. At this stage, the product is just a bare PCB with solder paste printed on it. Components have not yet been mounted on it. It is a typical semi-finished product.

[0055] Component mounting stage: The state is a semi-finished product. At this stage, the components have been mounted on the PCB, but have not yet been soldered and fixed. The solder paste is in an unmelted state, and the electrical function has not yet formed a complete circuit.

[0056] Soldering stage (reflow soldering): After soldering is completed, the product is considered close to finished. After reflow soldering, the components are firmly soldered to the PCB, and the circuit has basic electrical functions. However, the product may still have defects at this stage (such as cold solder joints or solder bridges), requiring further testing and verification.

[0057] Semi-finished product status: Semi-finished products refer to products that have completed some of the processes, including accurately placing components on the PCB using a mounting machine, but the products have not yet fully met the finished product standards.

[0058] Key characteristics: Components have been mounted on the PCB, but have not yet been soldered in place using the reflow soldering process. Alternatively, reflow soldering may have been completed, but functional testing or other final inspections have not been performed. Before mounting or soldering is complete, the semi-finished product typically cannot achieve full electrical functionality because the circuit connections are not yet fully formed.

[0059] Performing time interval distribution analysis on the detection result set refers to:

[0060] During the production line inspection process, the set of inspection time points for all quality inspection results marked as non-conforming products is obtained and arranged in chronological order to obtain the inspection time series of non-conforming products. By extracting the inspection time points of non-conforming products, we can focus on analyzing the time distribution pattern of abnormal phenomena and avoid the analysis being out of focus due to interference from data of conforming products.

[0061] The time interval is calculated for adjacent time points in the detection time series to generate a time interval distribution dataset. The time interval represents the length of the time interval between two consecutive detections. Then, feature extraction is performed on the time interval distribution dataset to obtain statistical characteristics including the following:

[0062] The average time interval is used to reflect the average inspection rhythm of the production line. The average time interval refers to the average of all statistical time intervals. By calculating the time interval between adjacent time points, it reveals whether the rhythm of the production line is regular and whether it meets expectations. Sudden changes in time intervals may indicate unstable equipment operation, operational deviations, or abnormal production processes. Constructing a time interval distribution dataset provides a data foundation for subsequent statistical characteristic analysis.

[0063] The fluctuation range of the time interval is used to reflect the stability of the production rhythm. The fluctuation range of the time interval refers to the maximum time interval minus the minimum time interval. The larger the fluctuation range, the more unstable the production rhythm is, which may be due to fluctuations in equipment performance or inconsistencies in human operation. The difference between the maximum and minimum time intervals can intuitively identify whether there are sudden abnormalities in the production rhythm. If the fluctuation range is too large, it may be necessary to conduct targeted inspections of key equipment or adjust the operating procedures.

[0064] The time interval anomaly ratio is used to identify the frequency of abnormal rhythmic behavior. It refers to the proportion of time intervals that fall outside the preset time standard range out of the total number of time intervals. By statistically analyzing the anomaly ratio, it's possible to quickly determine whether the production rhythm imbalance is systemic. A low anomaly ratio indicates that the anomaly detection is accidental, and the overall production rhythm is relatively stable. A high anomaly ratio suggests a more serious rhythm imbalance on the production line, potentially requiring a shutdown for inspection.

[0065] Identifying early signs of production rhythm imbalance in corresponding production stages based on the time interval distribution analysis results refers to:

[0066] An abnormal signal is generated when the average time interval is outside the preset normal rhythm range; the average time interval reflects the frequency of defective products. If the average value is outside the preset normal rhythm range, it may indicate a systemic problem in the operation of certain links in the production line, such as improper process flow adjustments leading to the gradual accumulation of problems, or resource scheduling errors in a certain stage of operation, resulting in frequent defective products.

[0067] An abnormal signal is generated when the fluctuation range of the time interval is less than the preset threshold. The fluctuation range of the time interval can reveal whether the occurrence of defective products has periodic fluctuation characteristics. An excessively large fluctuation range may indicate that there is instability in the production process, which leads to aggravated quality fluctuations.

[0068] An abnormal signal is generated when the proportion of time interval anomalies exceeds a preset limit. The proportion of time interval anomalies reflects whether the occurrence of non-conforming products is systematic. If the proportion of anomalies is high, it may indicate that the quality problem is not an accidental phenomenon, but a common problem in the production stage.

[0069] The generation of anomaly signals indicates an early sign of production rhythm imbalance in the corresponding production stage; the absence of anomaly signals indicates the absence of such early signs in the corresponding production stage. Comprehensive analysis of average values, fluctuation ranges, and the proportion of outliers can capture quality fluctuations during production. The generation of anomaly signals directly points to early signs of quality problems, creating conditions for precise intervention. Dynamic analysis of time interval characteristics allows for timely warnings and identification of systemic quality risks, preventing problems from escalating to the entire production line and ensuring the controllability and consistency of quality across all production stages.

[0070] The logic for obtaining the defect frequency fluctuation index is as follows:

[0071] Within a time interval consisting of N time windows, obtain the non-compliance frequency fi in time window i, calculate the mean of the non-compliance frequencies of all time windows as μDF, and calculate the standard deviation of the non-compliance frequencies of all time windows as σDF.

[0072] Calculate the rate of change of the non-conforming frequency fi', fi' = fi - fj; j = i - 1; then calculate the average absolute value of the rate of change fi'. Then substitute the values ​​into the following calculation formula:

[0073] C is a non-negative constant, γ is a preset non-zero adjustment factor, and IDF is the defect frequency fluctuation index.

[0074] The mean of the nonconforming frequency represents the average frequency of nonconforming products across all time windows, reflecting the overall level of nonconformity. The standard deviation of the nonconforming frequency measures the degree of fluctuation in the nonconforming frequency across all time windows, indicating the stability or dispersion of quality problems in the production process. The rate of change of the nonconforming frequency represents the magnitude of change in the nonconforming frequency between adjacent time windows, reflecting dynamic fluctuations. The average absolute value of the rate of change is used to quantify the overall trend of these fluctuations.

[0075] The adjustment factor γ controls the weight of the rate of change, increasing or decreasing the impact of dynamic fluctuations in the non-conforming frequency. C avoids a zero denominator and smooths the overall formula output, ensuring the stability of the index. The Defect Frequency Fluctuation Index (IDF) comprehensively measures the overall level, fluctuation amplitude, and dynamic trend of the non-conforming frequency, reflecting the dynamic behavioral characteristics of non-conforming phenomena. A small IDF indicates stable changes in the non-conforming frequency, possibly indicating an occasional problem. For example, non-conforming products appearing within a certain time window may be due to temporary operational errors or equipment fluctuations. A large IDF indicates significant fluctuations in the non-conforming frequency, with a systematic trend, indicating strong non-randomness. This usually suggests quality problems in the production line caused by non-random factors, such as equipment aging, raw material batch issues, or unstable process parameters. The IDF value can be used as input to a machine learning model, combined with other indicators (such as the production line dynamic disturbance index), to further determine the nature of the non-conforming phenomenon through the model output: occasional anomalies indicate overall production line stability, requiring little intervention. Non-random anomalies require shutdown for maintenance or adjustments to the production process to prevent the problem from spreading.

[0076] The logic for obtaining the production line dynamic disorder index is as follows:

[0077] Within a time interval consisting of N time windows, the start time of each test, the actual number of tests conducted within a preset standard unit time, and the test results (pass / fail markings) are obtained from the test result set for all products in the test time series. Then, the time interval between the start times of each test is calculated to obtain the time interval test dataset. The standard deviation of the time interval test dataset and the standard deviation reflecting the fluctuation of the actual number of tests conducted within a standard unit time are calculated.

[0078] Calculate the ratio of qualified products to actual inspection quantity in each time window i to obtain the window pass rate of time window i, and calculate the standard deviation of the pass rates of all windows.

[0079] Perform the following calculations:

[0080] μΔt is the standard deviation of the time interval detection dataset, σHG is the standard deviation of the pass rate of all windows, σQ is the standard deviation reflecting the fluctuation of the actual number of tests per standard unit time, ILD is the production line dynamic disorder index, and α and β are preset adjustment coefficients.

[0081] The standard deviation of the time interval detection dataset represents the degree of fluctuation in the interval between adjacent detection time points. Small fluctuations indicate relatively uniform detection time and stable production rhythm, while large fluctuations may indicate uneven production processes. The standard deviation of the fluctuation in the number of products detected per unit time measures the degree of fluctuation in the number of products detected per standard unit time. Small fluctuations indicate relatively stable production line load, while large fluctuations suggest potential issues such as process adjustments, equipment bottlenecks, or unstable material supply. The standard deviation of the pass rate measures the volatility of the pass rate within the time window. Small fluctuations indicate stable production line quality, while large fluctuations suggest potential periodic quality problems or systemic failures in certain stages of the production line. α and β adjust the weights of the pass rate volatility and the quantity of products detected in the final index, respectively.

[0082] If the ILD is small, it indicates that the dynamic changes in the production line are small, and the fluctuations in inspection time intervals, pass rates, and inspection quantities are all within a reasonable range. The occurrence of non-conforming products is more likely to be an occasional, localized problem. A larger index indicates that the anomaly is not accidental. If the ILD is large, it indicates a systemic dynamic disorder in the production line, possibly caused by long-term instability in certain aspects (such as equipment, materials, or processes). These problems are more likely to be non-accidental. Combining the ILD with the defect frequency fluctuation index, a machine learning model can be used to determine whether the anomaly is accidental or not: Occasional anomalies: The overall production process is stable, with only localized problems; Non-accidental anomalies: Significant dynamic fluctuations, requiring focused adjustments to the production line to prevent the problem from escalating.

[0083] The logic for using machine learning models is as follows:

[0084] The defect frequency fluctuation index and the production line dynamic disorder index are substituted into the pre-trained convolutional neural network model. The convolutional neural network model outputs either 1 or 0. An output of 1 indicates that the corresponding production stage is classified as a high-quality production line, which means that the non-conforming production situation of the production line in this production stage is accidental. An output of 0 indicates that the corresponding production stage is classified as a low-quality production line, which means that the non-conforming production situation of the production line in this production stage is not accidental.

[0085] Two core indices are generated using data from the production line quality inspection process:

[0086] Defect frequency fluctuation index: used to reflect the frequency of non-conforming products and their dynamic fluctuation characteristics.

[0087] Production line dynamic disturbance index: This index measures the fluctuations in the time intervals, pass rates, and number of inspections during production line operation, reflecting overall dynamic stability. These indices serve as input features and are fed into a machine learning model for comprehensive analysis.

[0088] The processing logic of the convolutional neural network model: The machine learning model extracts and analyzes the features of the input defect frequency fluctuation index and production line dynamic disorder index to capture the regularity of the production line's quality status. The pre-training process of the convolutional neural network model has learned the quality anomaly patterns and regularities in a large amount of production line operation data. Through model calculation, it outputs a judgment result indicating whether the current production stage is in a high-quality or low-quality state.

[0089] Output result: "High quality": When the model output result is 1, it indicates that the production line in the current production stage is in a high-quality state. At this time, the occurrence of defective products is judged as an accidental phenomenon, a random non-systematic problem, and the overall operation of the production line is stable.

[0090] The output result is "low quality": When the model output result is 0, it indicates that the production line in the current production stage is in a low quality state. At this time, the occurrence of defective products is judged as a systemic problem, which has the characteristics of non-randomness, indicating that there may be long-term problems such as equipment abnormalities, process problems or operational errors in the production line.

[0091] Adjustment Mechanism: For low-quality production lines, further pre-set adjustment strategies will be implemented, which may include shutdown for maintenance, optimization of process parameters, or inspection of equipment status. For high-quality production lines, the status quo will be maintained without excessive intervention.

[0092] By comprehensively analyzing two core indicators—dynamic disorder and defect fluctuation—the model can accurately determine whether the production line's operational quality meets expectations and improve the intelligence of production management. The machine learning model utilizes patterns in historical data to intelligently classify the production line's quality status, reducing subjective errors and inefficient analysis processes inherent in manual judgment. It effectively distinguishes between accidental and non-accidental anomalies: Defective products in high-quality production lines are usually accidental; the model helps eliminate these low-probability events, avoiding unnecessary production line adjustments. Defective products in low-quality production lines are usually systemic problems; the model promptly identifies these non-accidental risks, providing a basis for production line optimization. It enhances the production line's quality control capabilities: Based on the model's classification results, managers can specifically optimize low-quality production lines, reducing the systematic generation of defective products. It reduces waste and downtime: Through the accurate classification by the machine learning model, resources can be concentrated on repairing and optimizing low-quality production lines without interfering with the normal operation of high-quality production lines, thereby reducing unnecessary downtime and production waste.

[0093] Example 2: A staged quality inspection system for SMT semi-finished products, comprising:

[0094] The quality inspection module performs quality inspections on each stage of the SMT assembly semi-finished product production process and records the quality inspection results for each stage, thus obtaining a set of inspection results for each stage.

[0095] The early signs identification module performs time interval distribution analysis on the detection result set and identifies whether there are early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results.

[0096] The production line analysis module performs defect analysis and disorder analysis on the production line if there are early signs of production rhythm imbalance in the corresponding production stage, generating a defect frequency fluctuation index and a production line dynamic disorder index respectively.

[0097] The production line classification module inputs the defect frequency fluctuation index and the production line dynamic disorder index into a pre-trained machine learning model, and classifies the corresponding production stages into high-quality production lines and low-quality production lines based on the output results.

[0098] The production line adjustment module adjusts low-quality production lines according to a preset adjustment strategy.

[0099] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0100] It should be understood that, in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0101] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A staged quality inspection method for SMT (Surface Mount Technology) semi-finished products, characterized in that, Includes the following steps: Quality inspections are performed at each stage of the production of SMT semi-finished products, and the results of each stage are recorded to obtain a set of inspection results for each stage. Perform time interval distribution analysis on the detection result set, and identify early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results; If there are early signs of production rhythm imbalance in the corresponding production stage, defect analysis and disorder analysis are performed on the production line to generate a defect frequency fluctuation index and a production line dynamic disorder index, respectively. The defect frequency fluctuation index and the production line dynamic disorder index are substituted into the pre-trained machine learning model. Based on the output results, the corresponding production stages are divided into high-quality production lines and low-quality production lines. The low-quality production lines are adjusted according to the preset adjustment strategy. The specific logic for analyzing the time interval distribution and identifying early signs of the detection result set is as follows: obtain the detection time series of non-conforming products, calculate the time interval between adjacent time points, and generate a time interval distribution dataset; extract statistical characteristics, including: the average time interval reflecting the average detection rhythm, the time interval fluctuation range reflecting the rhythm stability, and the proportion of time interval anomalies identifying the frequency of abnormal rhythms; when the average time interval is not within the preset normal rhythm range, or the time interval fluctuation range is less than the preset threshold, or the proportion of time interval anomalies exceeds the preset limit, an abnormal signal is generated, indicating that there are early signs of production rhythm imbalance. The logic for obtaining the defect frequency fluctuation index is as follows: Within a time interval consisting of N time windows, obtain the frequency of non-compliance in time window i. The mean of the non-compliance frequency across all time windows is calculated as follows: Calculate the standard deviation of the non-compliance frequency across all time windows. ; Calculate the rate of change of the non-conforming frequency , ; Then calculate the rate of change. Average absolute value Then substitute the values ​​into the following calculation formula: ; It is a non-negative constant. The preset non-zero adjustment factor, This is the defect frequency fluctuation index; The logic for obtaining the production line dynamic disorder index is as follows: Within a time interval consisting of N time windows, the start time of each test, the actual number of tests conducted within a preset standard unit time, and the test results (pass / fail markings) are obtained from the test result set for all products in the test time series. Then, the time interval between the start times of each test is calculated to obtain the time interval test dataset. The standard deviation of the time interval test dataset and the standard deviation reflecting the fluctuation of the actual number of tests conducted within a standard unit time are calculated. Calculate the ratio of qualified products to actual inspection quantity in each time window i to obtain the window pass rate of time window i, and calculate the standard deviation of the pass rates of all windows. Perform the following calculations: ; The average value of the dataset is detected at time intervals. The standard deviation of the pass rate for all windows. To reflect the standard deviation of the actual quantity fluctuation detected per standard unit of time, , These are the preset adjustment coefficients. This represents the dynamic disorder index of the production line.

2. The staged quality inspection method for SMT semi-finished products according to claim 1, characterized in that, The production stages of SMT semi-finished products include: printing stage, mounting stage, and soldering stage. Quality inspection is carried out at each production stage in accordance with the preset quality inspection strategy for the corresponding production stage.

3. The staged quality inspection method for SMT semi-finished products according to claim 1, characterized in that, The logic for using machine learning models is as follows: The defect frequency fluctuation index and the production line dynamic disorder index are substituted into the pre-trained convolutional neural network model. The convolutional neural network model outputs either 1 or 0. An output of 1 indicates that the corresponding production stage is classified as a high-quality production line, which means that the non-conforming production situation of the production line in this production stage is accidental. An output of 0 indicates that the corresponding production stage is classified as a low-quality production line, which means that the non-conforming production situation of the production line in this production stage is not accidental.

4. A staged quality inspection system for SMT (Surface Mount Technology) semi-finished products, implemented based on the staged quality inspection method for SMT semi-finished products according to any one of claims 1-3, characterized in that, include: The quality inspection module performs quality inspections on each stage of the SMT assembly semi-finished product production process and records the quality inspection results for each stage, thus obtaining a set of inspection results for each stage. The early signs identification module performs time interval distribution analysis on the detection result set and identifies whether there are early signs of production rhythm imbalance in the corresponding production stage based on the time interval distribution analysis results. The production line analysis module performs defect analysis and disorder analysis on the production line if there are early signs of production rhythm imbalance in the corresponding production stage, generating a defect frequency fluctuation index and a production line dynamic disorder index respectively. The production line classification module inputs the defect frequency fluctuation index and the production line dynamic disorder index into a pre-trained machine learning model, and classifies the corresponding production stages into high-quality production lines and low-quality production lines based on the output results. The production line adjustment module adjusts low-quality production lines according to a preset adjustment strategy.

Images