Liquid crystal display module assembly detection method and system

Abstract

The invention relates to the technical field of module assembly and detection, and discloses a liquid crystal display module assembly and detection method and system, and the method comprises the steps: calculating the station association relationship among a screen line connection station, a backlight system station, a display effect station, a touch function station and an appearance inspection station; when detecting that the liquid crystal display module is abnormal, tracking a first station of an abnormal source according to the station association relationship; based on the first station of the abnormal source, calculating the influence degree of the abnormity on a subsequent second station of the abnormal source, and obtaining risk information; according to the risk information, a detection threshold value and a judgment condition of a second station are adjusted, detection is executed by using the adjusted detection threshold value and the adjusted judgment condition, and a quality file of the liquid crystal display module is generated. Potential quality risks can be recognized in advance, prevention measures can be taken, the misjudgment rate and the omission ratio are effectively reduced, and the detection quality is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of module assembly detection, and in particular to a liquid crystal display module assembly detection method and system. BACKGROUND

[0002] The traditional liquid crystal display module production line adopts a segmented detection mode, and each station independently performs quality detection, including screen line connection detection, backlight system detection, display effect verification, touch function test and appearance inspection, and other key links. The most prominent problem in the prior art is the lack of an effective information exchange mechanism between stations, resulting in a serious data island phenomenon, which causes the detection abnormalities of the previous station to be unable to be timely transmitted to the subsequent station, so that the subsequent station cannot adjust the detection strategy according to the previous quality state. At the same time, due to the complex physical coupling relationship between the components of the liquid crystal module, screen line connection abnormalities will directly affect the backlight brightness distribution, and backlight non-uniformity will cause false judgments in display effect detection, but the existing detection method cannot identify and quantify the associated influence between stations, causing difficulty in tracing the quality problem and unclear abnormal propagation path. SUMMARY

[0003] The present application provides a liquid crystal display module assembly detection method and system, which can identify potential quality risks in advance and take preventive measures through inter-station correlation analysis and risk prediction, effectively reducing the false positive rate and the false negative rate, and improving the detection quality.

[0004] The first aspect of the present application provides a liquid crystal display module assembly detection method, which comprises: calculating the station correlation relationship between the screen line connection station, the backlight system station, the display effect station, the touch function station and the appearance inspection station; when detecting an abnormal liquid crystal display module, tracing the first station of the abnormal source according to the station correlation relationship; calculating the influence degree of the abnormal source on the second station after the abnormal source based on the first station of the abnormal source, to obtain risk information; adjusting the detection threshold and the determination condition of the second station according to the risk information, and performing detection using the adjusted detection threshold and the adjusted determination condition to generate a quality file of the liquid crystal display module.

[0005] In combination with the first aspect, in a first implementation manner of the first aspect of the present application, the calculation of the station correlation relationship between the screen line connection station, the backlight system station, the display effect station, the touch function station and the appearance inspection station comprises: The connection points of the LVDS screen lines and the liquid crystal panel in the screen line connection station are measured for resistance value to obtain resistance data; the LED backlight strips in the backlight system station are tested and measured for brightness data one by one; in the display effect station, a standard test image is inputted and contrast data is detected; in the touch function station, multi-point touch positioning test is performed and response data is measured; in the appearance inspection station, the surface of the liquid crystal display module is scanned by a CCD vision system and scratch, bright spot and dark spot defects are identified to obtain defect data; The resistance data, the brightness data, the contrast data, the response data and the defect data are synchronously integrated to obtain station real-time data; The station real-time data are analyzed for correlation between screen line resistance and backlight brightness, correlation between backlight uniformity and display contrast, correlation between display quality and touch precision and correlation between touch precision and appearance defects to obtain station correlation.

[0006] In combination with the first aspect, in a second implementation manner of the first aspect of the present application, the analysis of the station real-time data for correlation between screen line resistance and backlight brightness, correlation between backlight uniformity and display contrast, correlation between display quality and touch precision and correlation between touch precision and appearance defects to obtain station correlation comprises: The influence degree of the change amplitude of the resistance data in the station real-time data on the uniformity of the distribution of the brightness data is calculated to obtain the correlation between the screen line resistance and the backlight brightness; The influence degree of the uniformity deviation of the brightness data in the station real-time data on the measurement accuracy of the contrast data is calculated to obtain the correlation between the backlight uniformity and the display contrast; The influence degree of the abnormal state of the contrast data in the station real-time data on the accuracy of the response data is calculated to obtain the correlation between the display quality and the touch precision; The influence degree of the deviation degree of the response data in the station real-time data on the detection accuracy of the defect data is calculated to obtain the correlation between the touch precision and the appearance defects; The correlation between the screen line resistance and the backlight brightness, the correlation between the backlight uniformity and the display contrast, the correlation between the display quality and the touch precision and the correlation between the touch precision and the appearance defects are integrated to form the station correlation.

[0007] In combination with the first aspect, in a third implementation manner of the first aspect of the present application, the calculation of the influence degree of the change amplitude of the resistance data in the station real-time data on the uniformity of the distribution of the brightness data to obtain the correlation between the screen line resistance and the backlight brightness comprises: The resistance data and the brightness data in the station real-time data are paired to obtain synchronous resistance-brightness data pairs; A sliding window difference calculation is performed on the resistance data in the synchronous resistance and brightness data pair to obtain a resistance change amplitude sequence; A uniformity variance calculation is performed on the brightness data in the synchronous resistance and brightness data pair according to the LED backlight bar distribution area to obtain a brightness distribution uniformity index; Linear regression fitting is performed on the resistance change amplitude sequence as the independent variable and the brightness distribution uniformity index as the dependent variable to obtain the correlation between the screen line resistance and the backlight brightness.

[0008] In combination with the first aspect, in a fourth implementation manner of the first aspect of the present application, when the liquid crystal display module is detected to be abnormal, the first station of the abnormal source is tracked according to the station association relationship, including: When the liquid crystal display module is detected to be abnormal, the quality parameter value deviating from the normal range in the real-time data of the station is identified, and an abnormal station is determined; The quality parameters of the previous stations associated with the abnormal station in the station association relationship are queried, and the influence weight of the quality parameters of the previous stations on the current abnormality is calculated; An abnormal propagation path from the station with the maximum influence weight to the abnormal station is constructed according to the influence weight, and the station with the maximum influence weight in the abnormal propagation path is selected as the first station of the abnormal source.

[0009] In combination with the first aspect, in a fifth implementation manner of the first aspect of the present application, the quality parameters of the previous stations associated with the abnormal station in the station association relationship are queried, and the influence weight of the quality parameters of the previous stations on the current abnormality is calculated, including: The previous station numbers having a direct quality influence chain with the abnormal station in the station association relationship are searched to obtain a sequence of associated stations; The quality parameters of the previous stations of the stations corresponding to the sequence of associated stations are extracted from the real-time data of the stations; The deviation degree of the quality parameters in the quality parameters of the previous stations is calculated, and the deviation degree of the quality parameters is multiplied by the corresponding association coefficient in the station association relationship to obtain the influence weight.

[0010] In combination with the first aspect, in a sixth implementation manner of the first aspect of the present application, the influence degree of the abnormality on the second station subsequent to the abnormal source is calculated based on the first station of the abnormal source to obtain risk information, including: The target abnormal quality parameter of the first station of the abnormal source is obtained, and the direct subsequent station of the first station in the production process is determined as the second station; The abnormal severity of the first station is calculated based on the target abnormal quality parameter; obtain a quality influence transmission coefficient between the first station and the second station from the station association relationship, and perform a product calculation on the quality influence transmission coefficient and the abnormality severity of the first station to obtain an influence degree of the abnormality on the second station; divide a risk level according to the influence degree of the abnormality on the second station and generate corresponding risk information.

[0011] In a seventh implementation manner of the first aspect, the adjusting the detection threshold and the determination condition of the second station according to the risk information and performing detection using the adjusted detection threshold and the adjusted determination condition to generate the quality file of the liquid crystal display module comprises: numerically correcting the detection threshold currently used by the second station according to the risk level in the risk information, expanding a detection threshold tolerance range by a preset percentage when the risk level is high risk to obtain an adjusted detection threshold; adjusting the quality qualified determination logic of the second station based on the influence degree in the risk information, adding a secondary confirmation determination step for an expected abnormality type to obtain an adjusted determination condition; performing detection using the adjusted detection threshold and the adjusted determination condition to generate the quality file of the liquid crystal display module.

[0012] In an eighth implementation manner of the first aspect, the performing detection using the adjusted detection threshold and the adjusted determination condition to generate the quality file of the liquid crystal display module comprises: configuring the adjusted detection threshold and the adjusted determination condition to a detection system of the second station, and performing a quality detection operation on the liquid crystal display module to obtain a second station detection result; associating the second station detection result with historical quality data of the liquid crystal display module at a screen line connection station, a backlight system station, a display effect station, a touch function station, and an appearance inspection station; analyzing an evolution process of quality parameters of each station in the historical quality data, and marking a time point of abnormality occurrence and an abnormality type to obtain a quality state evolution record; storing the quality state evolution record together with product number and production batch information of the liquid crystal display module to obtain a quality file.

[0013] The second aspect of the present application provides a liquid crystal display module assembly detection system, comprising: a calculation module configured to calculate a station association relationship between a screen line connection station, a backlight system station, a display effect station, a touch function station, and an appearance inspection station; A tracking module is configured to track a first station of an abnormal source according to the station correlation relationship when an abnormality of the liquid crystal display module is detected; A risk analysis module is configured to calculate an influence degree of the abnormality on a second station subsequent to the abnormal source based on the first station of the abnormal source, and obtain risk information; A generation module is configured to adjust a detection threshold and a judgment condition of the second station according to the risk information, and perform detection using the adjusted detection threshold and the adjusted judgment condition to generate a quality profile of the liquid crystal display module.

[0014] Compared with the prior art, the present application has the following beneficial effects: by establishing the quality correlation relationship between stations, the mutual influence between screen line resistance, backlight brightness, display contrast, touch accuracy and appearance defects can be identified and quantified, the technical limitations of traditional independent detection of each station are broken through, the data island problem in traditional detection is solved through real-time data transmission and synchronous integration between stations, each station can obtain quality state information of the previous station in time, the source station and the propagation chain of the quality abnormality can be quickly located through the establishment of an abnormal propagation path tracking algorithm, and the technical problems of difficult quality problem tracing and low processing efficiency in the traditional method are solved. By dynamically adjusting the detection threshold and the judgment condition of the subsequent station according to the abnormal condition of the previous station, the passive mode of traditional fixed parameter detection is changed, and an adaptive intelligent detection mechanism is formed. By recording the complete quality change trajectory of each liquid crystal display module from the first station to the last station, a product-level quality profile is established, potential quality risks can be identified in advance and preventive measures can be taken through inter-station correlation analysis and risk prediction, the misjudgment rate and the missed detection rate are effectively reduced, and the overall detection quality is improved. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without any creative labor.

[0016] The structures, proportions, sizes, etc. shown in the drawings of the present specification are only used to cooperate with the content disclosed in the present specification, to enable those skilled in the art to understand and read, and are not used to limit the limiting conditions of the embodiments of the present application, and therefore do not have technical significance. Any modification of the structure, change of the proportion relationship or adjustment of the size, without affecting the effects and purposes that can be achieved by the present application, should still fall within the scope of the technical content disclosed by the present application.

[0017] Figure 1is a flowchart of a liquid crystal display module assembly detection method provided by an embodiment of the present application. Figure 2 is a structural schematic block diagram of a liquid crystal display module assembly detection system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0019] The flowchart shown in the drawings is only an example and does not necessarily include all the contents and operations / steps, nor does it necessarily be executed in the described order. For example, some operations / steps can be decomposed, combined or partially merged, so that the actual execution order can be changed according to the actual situation.

[0020] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, unless otherwise clearly indicated by the context, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0021] It should be further understood that the term "and / or" used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. Please refer to Figure 1 One embodiment of the liquid crystal display module assembly detection method in the embodiments of the present application includes: Step 100, calculating the station association relationship between the screen line connection station, the backlight system station, the display effect station, the touch function station and the appearance inspection station; Specifically, at the screen cable connection testing station, a resistance measurement module is used to measure the resistance value of each connection point between the LVDS screen cable and the LCD panel, and the measured multi-channel resistance data is used as a basic parameter to characterize the signal transmission integrity and connection quality. At the backlight system testing station, the brightness value of each light-emitting unit is obtained by lighting up the LED backlight strip one by one, and the backlight uniformity coefficient is calculated in combination with the brightness distribution characteristics to reflect the overall optical performance of the backlight system. At the display effect testing station, a standardized test image is input into the LCD panel and the contrast value of the image is collected by an optical inspection device to quantify the visual performance of the display layer. At the touch function testing station, the response time and positioning accuracy of each touch point are measured by a multi-point touch positioning test method to obtain response data. At the appearance inspection station, a high-resolution CCD vision system is used to perform a full-width scan of the surface of the LCD module, automatically identify and classify appearance defects such as scratches, bright spots and dark spots, and obtain defect data. The five types of detection data mentioned above are time-stamped during the acquisition process through a TCP / IP-based real-time information exchange network between workstations. A microsecond-level precision clock synchronization mechanism is then used to integrate the data in the central data processing unit, forming a real-time workstation dataset. In the analysis phase, a physical coupling characteristic model of the liquid crystal module is established to calculate the correlation coefficient between screen line resistance data and backlight brightness data, determining the impact of electrical connection quality on backlight luminous performance. The correlation between backlight uniformity coefficient and display contrast data is calculated to reveal the effect of optical uniformity on image clarity. The statistical dependence between display quality parameters and touch response accuracy is analyzed to determine the potential impact of display layer stability on touch recognition performance. Touch accuracy data and appearance defect data are cross-compared to assess whether touch anomalies are directly or indirectly related to surface defects. Through the correlation analysis of the above multi-dimensional data, a quantifiable workstation correlation matrix is ​​obtained.

[0022] Step 200: When an abnormality is detected in the LCD display module, trace the first workstation that is the source of the abnormality according to the workstation association relationship; Specifically, when an anomaly is detected in the LCD module, real-time data from all workstations throughout the process is rapidly analyzed and compared. Each quality parameter value is compared to a pre-defined normal range to identify specific parameters deviating from the normal threshold, thus determining the workstation where the anomaly was first detected. After identifying the abnormal workstation, the workstation association database is accessed to retrieve historical quality parameter records from all preceding workstations directly or indirectly associated with the abnormal workstation. Based on the unique cross-workstation anomaly propagation model for LCD modules, the cross-workstation propagation weight coefficient obtained from historical statistics is used to calculate the contribution of each preceding workstation's quality parameter to the current abnormal workstation's performance. This contribution value is quantified as an influence weight, and interfering workstations below the statistical significance threshold are automatically filtered out. A directed graph path construction algorithm is used, with the influence weight as the edge weight. Starting from the workstation with the largest influence weight, paths are sequentially connected to the current abnormal workstation, generating one or more possible anomaly propagation paths. These paths are then sorted according to the cumulative influence weight. The preceding workstation with the highest influence weight among all propagation paths is selected as the first workstation from which the anomaly originates.

[0023] Step 300: Calculate the impact of the anomaly on the subsequent second workstation based on the first workstation of the anomaly source to obtain risk information; Specifically, the target abnormal quality parameters of the first workstation at the source of the abnormality are obtained, and the difference between the numerical deviation and the standard value is determined based on the target abnormal quality parameters. Combining a pre-defined abnormality severity assessment model, the difference is converted into an abnormality severity value. The higher the abnormality severity, the greater the potential impact of the current quality status of the workstation on downstream workstations. The quality impact transmission coefficient between the first and second workstations is obtained from the workstation correlation relationship. This coefficient is statistically obtained based on long-term production data and physical coupling characteristics. The quality impact transmission coefficient is multiplied by the abnormality severity of the first workstation to obtain the actual impact of the abnormality at the first workstation on the second workstation. A time decay factor is introduced to ensure that the calculation result reflects the real-time status. The risk is classified according to the numerical range of the impact degree. For example, an impact degree below a set low-risk threshold is classified as blue risk, within the middle range as yellow risk, and above a high-risk threshold as red risk. A risk information package containing the risk level, corresponding workstation number, impact degree value, and timestamp is generated.

[0024] Step 400: Adjust the detection threshold and judgment conditions of the second station according to the risk information, and perform the detection using the adjusted detection threshold and the adjusted judgment conditions to generate the quality file of the liquid crystal display module.

[0025] Specifically, after receiving risk information from the upstream station, the second station analyzes the risk level and compares it with the detection threshold currently used by the second station. When the analysis result indicates a high risk level, the tolerance range of the current detection threshold is expanded by a predefined percentage according to a preset threshold correction rule. For example, in a high-risk situation, the brightness uniformity judgment range is widened from ±5% to ±8%, or the lower limit of contrast is appropriately lowered to obtain an adjusted detection threshold that can adapt to potential abnormal conditions. Simultaneously, based on the impact level value contained in the risk information, the quality acceptance judgment logic of the second station is dynamically modified, introducing a secondary confirmation judgment step for specific abnormal types caused by the impact level. For example, in touch function testing, when the risk information indicates a high impact level from the display effect station, an additional full-screen multi-touch stability test is added to avoid misjudgment due to the randomness of a single test result. After completing the dynamic adjustment of the detection threshold and judgment conditions, the second station uses the updated detection threshold and judgment logic to test the current LCD module and collects all quality parameter values, judgment results, and risk information reference records in real time during the testing process. After the testing is completed, the data is integrated to generate a quality profile for the LCD module. This profile includes the final values ​​and judgment status of each test item, as well as the extent of threshold adjustments, modifications to the judgment conditions, and related risk information sources and timestamps.

[0026] In one specific embodiment, the process of performing step 100 may specifically include the following steps: Resistance values ​​are measured at the connection points between the LVDS screen cable and the LCD panel in the screen cable connection station to obtain resistance data; LED backlight strips in the backlight system station are individually lit up and brightness data is measured; a standard test image is input and contrast data is detected in the display effect station; multi-touch positioning tests are performed and response data is measured in the touch function station; and the surface of the LCD module is scanned using a CCD vision system in the appearance inspection station to identify scratches, bright spots, and dark spots to obtain defect data. The resistance data, brightness data, contrast data, response data, and defect data are synchronously integrated to obtain real-time data for the workstation. The correlation between screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects in real-time data of the workstations were analyzed to obtain the workstation correlation relationships.

[0027] Specifically, a high-precision multi-channel resistance measurement device is deployed at the screen cable connection testing station to measure the resistance value of each connection point between the LVDS screen cable and the LCD panel. An automatic probe positioning and pressure control mechanism is introduced during the data acquisition process to ensure the stability and repeatability of each contact test, obtaining resistance data with micro-ohm resolution. This reflects the conductivity quality of the signal transmission path and potential electrical hazards such as poor soldering and contact. At the backlight system testing station, a point-to-point lighting control circuit sequentially drives each light-emitting unit in the LED backlight strip. Simultaneously, a high-precision photometer or imaging luminance meter measures the corresponding brightness value, and the backlight uniformity coefficient is calculated based on the brightness distribution. Multiple sampling and noise reduction algorithms improve the stability of optical measurements, obtaining quantified brightness data of backlight performance. At the display effect testing station, a standardized test image, such as grayscale bars, checkerboard patterns, and color gradients, is input to the LCD panel through a drive control system. Simultaneously, optical sensing equipment with resolution and color accuracy meeting the testing requirements measures the contrast value of the screen image. At the touch function station, the testing system sequentially triggers multi-point touch positioning tests within the effective touch area of ​​the LCD module. A high-sampling-rate touch controller records the response time, positioning accuracy, and trigger stability of each touch point, obtaining response data. At the appearance inspection station, a CCD vision system with high-resolution imaging capabilities performs a comprehensive scan of the entire surface of the LCD module. Image processing and defect recognition algorithms automatically detect and classify appearance defects such as scratches, bright spots, and dark spots, forming a defect dataset. All resistance data, brightness data, contrast data, response data, and defect data are transmitted to the central data processing unit via a real-time information exchange network between stations based on the TCP / IP protocol after acquisition. Synchronization with microsecond-level timestamps is achieved using the IEEE 1588 precise time protocol, realizing the synchronous integration of cross-station data and forming a real-time dataset for each station. During the data analysis phase, based on the physical coupling model of the liquid crystal module, the correlation coefficient between screen line resistance and backlight brightness was calculated to assess the impact of electrical connection quality on optical output performance. The correlation between backlight uniformity and display contrast was analyzed to quantify the contribution of light source uniformity to the display layer contrast. Statistical modeling methods were used to evaluate the relationship between display quality and touch accuracy, determining the potential impact of imaging stability and sharpness variations on touch positioning accuracy. Touch accuracy and appearance defect data were cross-analyzed to determine whether touch performance fluctuations are directly or indirectly related to defects such as scratches, bright spots, or dark spots on the screen surface. Through multi-dimensional, high-precision end-to-end data acquisition, time synchronization, and correlation analysis, a set of quantitative workstation correlations was obtained.

[0028] In one specific embodiment, the process of analyzing the correlation between screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects in real-time data of the workstation to obtain the workstation correlation relationships can specifically include the following steps: The influence of the change in resistance data in the real-time data of the calculation station on the uniformity of brightness data distribution is obtained to obtain the correlation between screen line resistance and backlight brightness. The influence of the uniformity deviation of brightness data in real-time data of the calculation station on the measurement accuracy of contrast data is obtained to obtain the correlation between backlight uniformity and display contrast. The impact of abnormal contrast data in real-time data of the calculation station on the accuracy of response data is determined, thus obtaining the correlation between display quality and touch accuracy. The degree of deviation of response data in real-time data of the calculation station affects the accuracy of defect data detection, and the correlation between touch accuracy and appearance defects is obtained; The correlation between screen line resistance and backlight brightness, the correlation between backlight uniformity and display contrast, the correlation between display quality and touch accuracy, and the correlation between touch accuracy and appearance defects are integrated to form the workstation correlation relationship.

[0029] Specifically, focusing on the correlation between screen line resistance and backlight brightness, a sequence of resistance data variation within the process window is constructed. Combining this with the brightness distribution matrix and uniformity coefficient obtained by lighting up each point, piecewise statistics and robust correlation are used to quantify the sensitivity of electrical connection degradation to the uniformity of optical output. Propagation weights from historical samples are introduced, and an empirical coefficient of 0.73 for the impact of screen line connection on backlight is used as a priori to ensure stable convergence of real-time evaluation under inter-batch and intra-batch fluctuations. When the LVDS signal integrity reported by the screen line connection detection is below 95%, the evaluation interval is sampled in a focused manner, and multiple scans are triggered in conjunction with the backlight station, upgrading a single scan to three scans to reduce random errors and outputting conclusions on the impact on brightness uniformity. To investigate the correlation between backlight uniformity and display contrast, standard test image contrast data from the imaging end is registered and compared with uniformity indicators from the light source side. Through region grouping and edge-center differential analysis, the contrast measurement bias caused by optical non-uniformity is characterized. An empirical coefficient of 0.86 is introduced from backlight to display to assign differentiated weights to different brightness blocks. Simultaneously, within a 15-second production cycle limit, the calculation time is compressed to less than 20 milliseconds to ensure real-time online judgment. Regarding the correlation between display quality and touch accuracy, display quality factors such as contrast, pixel response stability, and viewing angle are selected and jointly analyzed with indicators such as multi-touch response time, positioning deviation, and stability. A time sliding window is used to track the disturbance effect of display layer fluctuations on touch controller sampling, and a weighted evaluation is performed using an empirical coefficient of 0.51 for the cross-influence between display and touch. When a persistent anomaly occurs at the display end, the touch station automatically adds a full-screen stability retest as a secondary confirmation step to avoid misjudgments caused by the randomness of a single measurement. To investigate the correlation between touch accuracy and appearance defects, a spatial overlap rate detection method using touch error heatmaps and appearance defect coordinates is employed. This is combined with type labels for bright spots, dark spots, and scratches to assess the interference of touch deviation on defect identification. An empirical coefficient of 0.28 (touch-to-appearance) is used to suppress overfitting caused by weak correlations. Simultaneously, on the visual side, threshold adaptation and region refinement segmentation are used to improve the consistency of bright spot and dark spot detection. After quantifying the four correlation links, the results for screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects are integrated. A directed graph is used to represent the influence paths between workstations, and real-time evaluation values ​​are written into the edge weights and combined with historical priors to form a cross-workstation anomaly propagation model.

[0030] In this embodiment, the correlations between screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects are integrated to form workstation correlation relationships. This includes: establishing a Bayesian filter for real-time quality state transmission between LCD module workstations; setting the state transition probabilities from screen line resistance value to backlight brightness, backlight brightness to display contrast, display contrast to touch response time, and touch response time to appearance defect density to obtain a five-dimensional quality state transition equation; and calculating the information transmission error term between workstations and the current workstation detection based on the five-dimensional quality state transition equation. Input parameters are used to recursively update the quality state vector, resulting in a dynamic quality state sequence. Prior knowledge of the physical coupling relationship of the LCD modules is used to weight the dynamic quality state sequence. When the current station's detection confidence is higher than a preset threshold, the weight of historical state dependence is reduced; when the current station's detection confidence is lower than a preset threshold, the weight of historical state dependence is increased, resulting in adaptive weight coefficients. These adaptive weight coefficients are then weighted and fused with the dynamic quality state sequence to construct a quality influence transmission matrix between stations, yielding quantified station correlations. Based on these quantified station correlations, a dynamic adjustment mechanism is established for the detection threshold of subsequent stations based on the detection results of previous stations, resulting in the station correlation relationships.

[0031] In one specific embodiment, the process of calculating the influence of the change in resistance data in the real-time data of the workstation on the uniformity of brightness data distribution, and obtaining the correlation between screen line resistance and backlight brightness, can specifically include the following steps: Pair the resistance data and brightness data in the real-time data of the workstation to obtain synchronous resistance-brightness data pairs; A sliding window differential calculation is performed on the synchronous resistance brightness data and the centering resistance data to obtain the resistance change amplitude sequence; The uniformity variance of the brightness distribution is calculated by aligning the brightness data of the synchronous resistor with the brightness data according to the distribution area of ​​the LED backlight strip, and the brightness distribution uniformity index is obtained. By using the resistance variation amplitude sequence as the independent variable and the brightness distribution uniformity index as the dependent variable, a linear regression was performed to obtain the correlation between screen line resistance and backlight brightness.

[0032] Specifically, the resistance and brightness data in the real-time data from each workstation are paired and their correlation modeling is completed. A TCP / IP-based real-time information exchange network between workstations is enabled on the production line. Data packets output from the screen connection workstation and the backlight system workstation are aligned according to product serial number and microsecond-level timestamps. The IEEE 1588 precision time protocol ensures the comparability of acquisition times from different workstations. Simultaneously, a data access buffer is established within the central processing unit to discard or re-acquire data packets with delays exceeding a preset threshold, ensuring that pairing only occurs within valid observations of the same production cycle. After alignment, the resistance data and brightness data obtained from point-by-point illumination are paired using the product serial number as the key to form synchronous resistance-brightness data pairs. Consistent preprocessing is performed on both sides of the data, including removing range boundary anomalies, temperature drift compensation, probe contact jitter filtering, and optical noise smoothing, reducing bias caused by inconsistent sources in the analysis. On the resistance side, an adjustable sliding window is introduced to perform adjacent-time difference operations on the synchronous resistance sequence and suppress isolated spikes using the median absolute deviation, ensuring that the output resistance change amplitude sequence reflects the true conduction state fluctuations. The window length is adaptively set based on the production cycle time and sampling frequency. A shorter window is used to improve sensitivity when the cycle time is stable and the sampling density is high, while a longer window is used to enhance robustness when there is high noise or increased batch-to-batch variation. Corresponding to the brightness side, the original brightness matrix of each point is mapped to the physical partitions of the LED backlight strip. A partition mask is used to divide the image into regions such as the center, edges, and transition zones, and the uniformity variance of brightness within each region is calculated. Simultaneously, the results are weighted by region area and human eye sensitivity to obtain a single numerical value for brightness distribution uniformity. To reduce sampling randomness, when the LVDS signal integrity is below 95% or the backlight status indicator enters a warning state, the number of point-by-point scans is increased from one to three, and the results from the three scans are robustly summarized to improve the consistency of uniformity assessment. In the modeling phase, a linear regression relationship is established using the resistance variation amplitude sequence as the independent variable and the brightness distribution uniformity index as the dependent variable. A robust regression strategy is employed to mitigate the influence of a small number of outlier paired samples during parameter solving. K-fold cross-validation is used to assess generalization ability, and the coefficient of determination, residual distribution, and diagnostic results of influencing points are recorded. If the residuals exhibit significant heteroscedasticity or structural bias, a piecewise regression or variable transformation process is automatically triggered to restore linear interpretability. To avoid systematic biases introduced by batch and environmental conditions, the model training set is stratified by batch and environmental labels, and incremental updates are performed using a sliding time window after deployment, allowing the regression coefficients to adaptively adjust as the process and device aging slowly drift. When real-time data is received, the mean and variance of the independent variables for each batch are calculated immediately to ensure they fall within the training distribution range. If distribution drift occurs, the model confidence weight is reduced, and maintenance is prompted to perform calibration and retesting.On the output side, the linear regression results are written to the quality data bus as a correlation value between screen line resistance and backlight brightness, along with the sample size, window length, region weight, coefficient of determination, and timestamp. When the correlation value exceeds a preset threshold of the historical baseline, the display contrast and touch tests will adjust the thresholds and re-inspection process in tandem to reduce the risk of upstream electrical connection fluctuations misleading downstream judgments.

[0033] The study calculates the impact of uniformity deviation of brightness data in real-time data from the calculation station on the measurement accuracy of contrast data, thus establishing the correlation between backlight uniformity and display contrast. This includes: extracting brightness data from the backlight system station and contrast data from the display effect station from the real-time data, and matching them according to the LCD module serial number to obtain brightness-contrast pairing data; calculating the deviation between the brightness value of each area and the average brightness value in the brightness data of the brightness-contrast pairing data according to the distribution area of ​​the LED backlight strip to obtain the uniformity deviation value; performing repeated measurement error analysis and measurement stability calculation on the contrast data of the brightness-contrast pairing data to obtain the contrast measurement accuracy value; and performing correlation analysis and influence weight calculation on the uniformity deviation value and the contrast measurement accuracy value to obtain the correlation between backlight uniformity and display contrast.

[0034] The study calculates the impact of abnormal contrast data in real-time data from the workstations on the accuracy of response data, thus establishing the correlation between display quality and touch accuracy. This includes: extracting contrast data from the display effect workstation and response data from the touch function workstation from the real-time data, and associating them according to the detection sequence to obtain contrast response pairing data; determining the normal range and quantifying the degree of abnormality in the contrast data in the contrast response pairing data to obtain a contrast abnormality index; performing touch positioning error analysis and time delay stability calculation on the response data in the contrast response pairing data to obtain a touch response accuracy index; and conducting causal relationship analysis and influence intensity calculation between the contrast abnormality index and the touch response accuracy index to obtain the correlation between display quality and touch accuracy.

[0035] The calculation of the impact of the deviation of response data in the separation quality parameters on the accuracy of defect data detection yields the correlation between touch accuracy and appearance defects. This includes: extracting response data from the touch function station and defect data from the appearance inspection station from the separation quality parameters, and performing time-series pairing according to the detection order of the LCD module to obtain touch appearance pairing data; calculating the touch positioning coordinate deviation and response time delay of the response data in the touch appearance pairing data to obtain the touch accuracy deviation index; performing statistical analysis of defect identification accuracy and missed detection rate of the defect data in the touch appearance pairing data to obtain the appearance inspection accuracy index; and performing cross-influence analysis and correlation strength calculation between the touch accuracy deviation index and the appearance inspection accuracy index to obtain the correlation between touch accuracy and appearance defects.

[0036] In one specific embodiment, the process of performing step 200 may specifically include the following steps: When an abnormality is detected in the LCD module, the quality parameter values ​​that deviate from the normal range in the real-time data of the workstation are identified and the abnormal workstation is determined. Query the quality parameters of the preceding workstations associated with the abnormal workstation in the workstation association relationship, and calculate the influence weight of the preceding workstation quality parameters on the current abnormality; Based on the influence weight, an anomaly propagation path is constructed from the workstation with the largest influence weight to the abnormal workstation, and the workstation with the largest influence weight value in the anomaly propagation path is selected as the first workstation of the anomaly source.

[0037] Specifically, when an anomaly is detected in the LCD module, a real-time monitoring and dynamic comparison of real-time data from all workstations throughout the process is performed based on a real-time detection data acquisition and synchronization mechanism. During the inspection phase, when the parameter value output by the quality inspection module of any workstation exceeds the preset normal range threshold, an anomaly detection event is triggered. A quality parameter anomaly identification program compares the real-time data of all workstations corresponding to the current batch and product serial number with the process standard range item by item, filtering out parameter items whose deviation exceeds the tolerance limit. The workstation where the anomaly was first discovered is determined through an anomaly parameter-to-workstation mapping table, i.e., the initial anomaly workstation. The anomaly parameters are then reconfirmed, using the moving average and standard deviation of three to five adjacent historical sampling periods to determine the persistence and stability of the deviation. If the deviation is confirmed to be persistent, the workstation is marked as the current anomaly workstation. Based on the workstation association database, the quality parameters of the preceding workstations associated with the anomaly workstation are queried. The workstation association database is a cross-workstation quality influence map formed through statistical analysis and physical coupling modeling based on multi-dimensional quality parameters such as resistance, brightness, contrast, touch response, and appearance defects during long-term production. The system retrieves all preceding workstations that are directly or indirectly related to the current abnormal workstation and extracts their quality parameter values ​​for the current batch. To quantify the contribution of preceding workstations to the current abnormality, a weighted calculation is performed using pre-stored cross-workstation influence transmission coefficients and real-time deviation amplitudes, based on historical abnormality propagation data and real-time detection deviations, to obtain the influence weight of each preceding workstation. The calculation of influence weights considers not only the linear transmission relationship in numerical terms but also the physical correlation strength of the abnormality type. For example, the influence weight of a screen cable connection abnormality on backlight uniformity is greater than its influence weight on touch response. A time decay factor is also introduced to reflect the timeliness of information transmission, ensuring that preceding workstations closer to the current workstation have a higher immediate influence value in the calculation. After obtaining the influence weights of all preceding workstations, the abnormality propagation path construction module is activated. This module uses a directed graph, with each workstation as a node and the influence weight as the edge weight. Starting from the preceding workstation with the largest influence weight, it sequentially connects to the current abnormal workstation along the quality parameter propagation direction, forming one or more possible abnormality propagation paths. During path construction, a path search algorithm prioritizing the highest weight is executed, while weakly influential paths below the minimum significance threshold are excluded to ensure that the propagation chain reflects the effective anomaly transmission chain in the actual production process. When multiple path candidates exist, the cumulative influence weight of each path is calculated and ranked, with the path having the highest cumulative value selected as the primary propagation path. Within the primary propagation path, the workstation with the highest influence weight is selected as the first workstation of the anomaly source and marked as the initial fault location of the LCD module.

[0038] When an anomaly is detected in the LCD module, the system identifies quality parameter values ​​that deviate from the normal range in the real-time data of the workstation and determines the abnormal workstation. This includes: establishing a network for rapid location and propagation path tracking of LCD module anomaly sources; integrating electrical connection integrity index, backlight uniformity coefficient, display quality evaluation factor, touch accuracy parameter, and appearance defect level as a five-dimensional quality feature vector to obtain a quality feature space; real-time monitoring of quality parameters in each dimension of the quality feature space; triggering an anomaly identification algorithm when any dimension parameter exceeds the normal range; calculating the degree of parameter deviation and the duration of deviation to obtain an anomaly severity index; analyzing the anomaly severity index using an anomaly propagation path identification algorithm; calculating the probability distribution of each workstation as an anomaly source through maximum likelihood estimation; selecting the workstation with the highest probability as a candidate anomaly source to obtain a candidate anomaly source list; establishing a directed graph anomaly propagation model with workstations as nodes, anomaly propagation paths as edges, and propagation probabilities as weights; performing propagation path backtracking analysis on the candidate sources in the anomaly source candidate list to obtain an anomaly propagation chain; and comprehensively evaluating the anomaly contribution and propagation delay of each node in the anomaly propagation chain to determine the final anomaly source workstation, thus obtaining the abnormal workstation.

[0039] In one specific embodiment, the process of querying the quality parameters of the preceding workstation associated with the abnormal workstation in the workstation association relationship, and calculating the influence weight of the preceding workstation quality parameters on the current abnormality, can specifically include the following steps: Search the workstation number in the workstation association relationship that has a direct quality impact chain with the abnormal workstation to obtain the associated workstation sequence; Extract the quality parameters of the preceding workstations from the real-time data of the workstations; Calculate the degree of deviation of the quality parameters in the quality parameters of the preceding workstation, and multiply the degree of deviation of the quality parameters by the corresponding correlation coefficient in the workstation correlation relationship to obtain the influence weight.

[0040] Specifically, a workstation correlation database is established, derived from cross-workstation quality data collected during the long-term operation of the LCD module production line, and obtained through statistical analysis and physical mechanism modeling. The database is organized as a directed graph, where nodes represent each inspection workstation, edges represent direct quality influence chains between workstations, and edge weights are correlation coefficients used to quantify the impact of changes in the quality status of preceding workstations on the quality results of subsequent workstations. When a workstation is found to have quality parameters exceeding the preset normal range during the inspection process, this workstation is identified as an abnormal workstation. The workstation correlation database is then invoked, and a node-based adjacency list query is performed in the directed graph to quickly retrieve all preceding workstation numbers that have a direct quality influence chain with the abnormal workstation, resulting in a sequence of associated workstations arranged according to the production process. Based on the workstation numbers in the associated workstation sequence, the corresponding quality parameter values ​​are extracted from the real-time dataset one by one, ensuring that these parameter values ​​completely correspond to the current batch and the current product serial number to avoid calculation errors caused by the mixing of historical batch data. The extracted quality parameters include the LVDS signal integrity and resistance value at the screen cable connection station, the brightness uniformity coefficient at the backlight system station, the contrast ratio at the display effect station, the response latency at the touch function station, and the defect density at the appearance inspection station. The deviation degree of quality parameters in the preceding station's quality parameters is calculated. The difference between each quality parameter and the process standard value or the center value of the normal range is calculated, and then divided by the tolerance range of the corresponding parameter to obtain a normalized deviation degree index. The deviation degree of quality parameters is multiplied by the correlation coefficient from the preceding station to the abnormal station in the station association database to obtain the influence weight. The correlation coefficient reflects the transmission strength of the quality fluctuation of the preceding station to the quality anomaly of the abnormal station, and is obtained through regression analysis and physical coupling modeling of long-term production data. The normalized deviation degree is multiplied by the correlation coefficient to obtain a comprehensive index that simultaneously includes real-time deviation information and historical correlation strength, i.e., the influence weight, which quantifies the immediate impact of the current preceding station anomaly on the abnormal station.

[0041] In one specific embodiment, the process of performing step 300 may specifically include the following steps: Obtain the target abnormal quality parameters of the first station at the source of the abnormality, and determine the direct subsequent station in the production process as the second station. The severity of the anomaly at the first workstation is calculated based on the target anomaly quality parameters. The quality impact transmission coefficient between the first and second workstations is obtained from the workstation association relationship. The quality impact transmission coefficient is then multiplied by the severity of the anomaly at the first workstation to calculate the degree of impact of the anomaly on the second workstation. The risk level is determined based on the degree of impact of the anomaly on the second workstation, and corresponding risk information is generated.

[0042] Specifically, the target abnormal quality parameters of the first workstation at the source of the abnormality are obtained. For example, in the screen cable connection workstation, this is the LVDS signal integrity value; in the backlight system workstation, it is the brightness uniformity coefficient; in the display effect workstation, it is the contrast ratio; and in the touch function workstation, it is the touch response latency. The workstation sequence table of the production process is retrieved to determine the directly succeeding workstation of the first workstation as the second workstation. The difference between the target abnormal quality parameter and the corresponding standard value is calculated and divided by the process tolerance range of the target abnormal quality parameter to obtain a normalized deviation value. This deviation value, after numerical scaling, serves as an abnormality severity index, reflecting the potential impact of the abnormality at the first workstation on the production process. Based on the workstation relationship database, the quality influence transmission coefficient between the first and second workstations is extracted. The quality impact transmission coefficient is an empirical parameter derived from long-term production data accumulation and physical mechanism analysis. It reflects the intensity of the impact of changes in the quality status of the first station on the detection results of the second station. For example, the impact transmission coefficient of the screen cable connection station on the backlight system station is 0.73, and the transmission coefficient of the backlight system station on the display effect station is 0.86. The quality impact transmission coefficient is multiplied by the anomaly severity to obtain the actual impact degree of the anomaly on the second station. Based on the impact degree of the anomaly on the second station, a risk classification and risk information are generated. The risk classification rules are pre-set based on the numerical range of the impact degree. For example, an impact degree below 0.3 is classified as low risk (blue warning), between 0.3 and 0.6 as medium risk (yellow warning), and above 0.6 as high risk (red warning). A corresponding risk information package is generated based on the risk level. The risk information package includes the risk level, the impact degree value, the numbers of the first and second stations, the target anomaly quality parameter value, the anomaly severity, the quality impact transmission coefficient, and information such as the current production batch and timestamp.

[0043] The process involves classifying risk levels and generating corresponding risk information based on the impact of anomalies on the second workstation. This includes: establishing a real-time exchange and prediction / early warning protocol for LCD module quality information throughout the process; adopting a quality data packet transmission mechanism between workstations; defining standardized exchange formats for status information packets, measurement data packets, anomaly alarm packets, prediction / early warning packets, and control command packets to obtain an information exchange protocol; constructing a quality prediction model based on the information exchange protocol; using historical quality status sequences and current changes in detection parameters as input variables; employing an autoregressive structure to predict potential quality risks at subsequent workstations to obtain predicted quality risk values; and establishing a system based on the numerical range of the predicted quality risk values. The three-tiered risk level classification standard sets a blue alert when the predicted value deviates from the normal range by 10% to 15%, a yellow alert when it deviates by 15% to 25%, and a red alert when it deviates by more than 25%. This determines the risk level. The risk level is then mapped to the predicted time window and the scope of impact to generate comprehensive early warning information, including the risk level identifier, the affected workstation number, the warning duration, and recommended handling measures. This results in structured risk information. This structured risk information is then simultaneously broadcast to all relevant workstations and the host computer system via an information exchange protocol, with the response time controlled within a preset millisecond range.

[0044] In one specific embodiment, the process of performing step 400 may specifically include the following steps: Based on the risk level in the risk information, the detection threshold currently used in the second workstation is numerically corrected. When the risk level is high, the tolerance range of the detection threshold is expanded by a preset percentage to obtain the adjusted detection threshold. Based on the degree of impact in the risk information, the quality acceptance judgment logic of the second workstation was adjusted, and a secondary confirmation judgment step for the expected abnormality type was added to obtain the adjusted judgment conditions. The detection is performed using the adjusted detection threshold and the adjusted judgment conditions to generate a quality profile for the liquid crystal display module.

[0045] Specifically, the risk level in the risk information is compared and calculated with the detection threshold currently loaded at the second station, and the detection threshold correction rule library is called to automatically select the corresponding numerical correction strategy according to the different risk levels. When the risk level is determined to be high risk, the detection threshold tolerance range is widened by a preset percentage. For example, the brightness uniformity judgment tolerance is expanded from ±5% to ±8%, or the lower limit of the display contrast qualification value is reduced by a certain percentage. This ensures that in the case of significant anomalies upstream, the detection system will not produce a large number of false positives or false negatives due to overly strict thresholds, thereby ensuring the validity and stability of the detection results. At the same time, the quality qualification judgment logic of the second station is dynamically adjusted based on the degree of impact of anomalies in the risk information. An anomaly type judgment model is used to match the degree of impact with the specific anomaly type that may be caused. For example, when the high degree of impact from the upstream backlight system station indicates that backlight brightness unevenness will cause the contrast measurement distortion of the display effect station, a secondary confirmation judgment step for the contrast parameter is embedded in the original judgment logic. This includes repeated measurement, contrast comparison analysis under different image modes, or calling additional image analysis algorithms to verify the results when the measurement results are abnormal. For different risk levels, the execution conditions for secondary confirmation are flexibly configured. For example, for medium-risk levels, secondary confirmation is performed only for some key detection parameters, while for high-risk levels, additional re-inspection steps are performed for all core detection indicators. After the dynamic adjustment of the detection threshold and judgment conditions is completed, the second station immediately loads the adjusted parameter set into the detection execution process. During real-time detection, all detection signals, measured values, corrected threshold ranges, adjusted judgment logic steps, and secondary confirmation detection results are recorded in the detection process buffer. The detection execution outputs a pass / fail conclusion and retains the intermediate data of each judgment step. After the detection is completed, all information involved in this detection is integrated, including the comparison of detection thresholds before and after adjustment, the specific content of the judgment condition modification, a complete record of risk information, the measured value and judgment status of each detection parameter, the execution status and results of the secondary confirmation steps, and related auxiliary information such as production batch, timestamp, and operator number. This information is stored in the full-process quality tracking database in a structured data format and bound to the unique product serial number of the LCD module to form a product quality file.

[0046] In this embodiment, a multi-station interconnected nonlinear detection system model is established, with the screen cable connection station, backlight system station, display effect station, touch function station, and appearance inspection station as interconnected subsystems. The influence of detection equipment failures and unknown detection parameters between stations is considered to obtain the system state equation. A dynamic quality state observer is constructed based on dynamic detection threshold scaling technology to monitor and estimate the quality state of each station in the system state equation in real time, obtaining quality state estimates. An adaptive distributed quality feedback controller is designed using a high-order multi-station joint detection method. Based on the quality state estimates, the quality feedback control signals between stations are calculated to obtain a distributed control strategy. Compensation calculations are performed for detection efficiency loss, unknown detection gain, and unstructured uncertainties in the transmission delay between stations. The distributed control strategy is combined with the transmission delay compensation between stations to obtain a delay compensation control strategy. The stability of the delay compensation control strategy is analyzed using a quality stability evaluation function to eliminate the influence of the delay term on detection accuracy, ensuring that all quality signals of the multi-station detection system converge within a predetermined accuracy range, resulting in adjusted detection thresholds and adjusted judgment conditions.

[0047] In one specific embodiment, the process of performing detection using adjusted detection thresholds and adjusted judgment conditions to generate a quality profile for the liquid crystal display module may specifically include the following steps: Configure the adjusted detection threshold and the adjusted judgment conditions into the detection system of the second station, and perform quality inspection on the LCD module to obtain the detection results of the second station; The test results of the second station are correlated with the historical quality data of the LCD module at the screen cable connection station, backlight system station, display effect station, touch function station, and appearance inspection station. Analyze the evolution of quality parameters at each workstation in historical quality data, and mark the time points and types of anomalies to obtain a record of quality status evolution. The storage of quality status evolution records, along with the product number and production batch information of the LCD display module, yields a quality archive.

[0048] Specifically, the updated detection thresholds and judgment condition parameter packages are transmitted to the detection system at the second workstation via a real-time information exchange network. The parameter packages include adjusted numerical ranges, judgment logic modification instructions, execution priority identifiers, and associated risk information numbers. During transmission, a TCP / IP protocol combined with a checksum mechanism ensures data integrity and accuracy. Upon receiving the parameter packages, the detection system at the second workstation executes an online parameter loading program, writes the adjusted thresholds into the judgment module of the detection algorithm, and rearranges the detection process according to the updated logic table. For example, in high-risk situations, a secondary confirmation step or an increase in the number of samples may be introduced to ensure that the detection process fully reflects the potential impact of upstream anomalies. After parameter loading is complete, the detection system at the second workstation performs quality inspection operations on the LCD display modules entering the inspection area under actual production cycle time. The inspection content depends on the function of the workstation. For example, if the second workstation is a display effect inspection workstation, standardized test images are input into the LCD panel, and contrast, color saturation, and grayscale response time are measured. The system then determines whether the modules are qualified or not based on the adjusted threshold ranges and judgment conditions. The inspection system outputs the inspection results along with detailed measurement data, secondary confirmation results, executed inspection strategy numbers, and parameter adjustment timestamps as the second-station inspection results. Through a quality data association module, the second-station inspection results are correlated and matched with historical quality data from the LCD module's screen cable connection station, backlight system station, display effect station, touch function station, and appearance inspection station. The association process uses the LCD module's unique product number as the primary key and the production batch number as the grouping index. It retrieves all inspection records for the corresponding product at previous stations from the full-process quality tracking database, including the inspection parameter values, judgment status, executed inspection strategies, and collection timestamps for each station, and sorts and merges them according to the production process sequence. The evolution of quality parameters at each station in the historical quality data is analyzed, and time series analysis is used to model and visualize the changing trends of quality parameters at each station. The absolute values ​​of each parameter at different time points are analyzed, and the parameter change rate, fluctuation amplitude, and cross-station parameter correlation are calculated to identify potential quality degradation patterns. During this process, the time point of anomaly occurrence is automatically marked, i.e., the moment when the test result first deviates from the normal range. Anomaly type labels are determined by combining the anomaly parameter type, deviation magnitude, and correlation coefficient, such as "electrical connection degradation," "increased backlight unevenness," "decreased display contrast," "increased touch response delay," or "increased surface defects." After analysis, the resulting quality status evolution record is integrated with the product number and production batch information of the LCD module to generate a quality archive. This archive includes historical test data for each workstation, parameter change trend graphs, anomaly occurrence time points and type labels, the latest test results from the second workstation, records of executed test strategy adjustments and their effective dates, risk information citation numbers, and the timestamp of the analysis report generation.

[0049] The liquid crystal display module assembly and testing method in the embodiments of the present invention has been described above. The liquid crystal display module assembly and testing system in the embodiments of the present invention will be described below. Please refer to [link / reference]. Figure 2 One embodiment of the liquid crystal display module assembly and testing system of the present invention includes: Calculation module 21 is used to calculate the station relationships between the screen cable connection station, backlight system station, display effect station, touch function station and appearance inspection station; The tracking module 22 is used to track the first workstation that is the source of the abnormality when an abnormality is detected in the LCD display module, based on the workstation association relationship. Risk analysis module 23 is used to calculate the impact of the anomaly on the subsequent second workstation based on the first workstation of the anomaly source, and obtain risk information; The generation module 24 is used to adjust the detection threshold and judgment conditions of the second station according to the risk information, and to perform detection using the adjusted detection threshold and judgment conditions to generate a quality file of the liquid crystal display module.

[0050] Through the collaborative efforts of the aforementioned components and the establishment of quality correlations between workstations, the mutual influence between screen line resistance, backlight brightness, display contrast, touch accuracy, and appearance defects can be identified and quantified. This overcomes the technical limitations of traditional independent testing at each workstation. Real-time data transmission and synchronous integration between workstations break down data silos in traditional testing, enabling each workstation to promptly obtain quality status information from preceding workstations. By establishing an anomaly propagation path tracking algorithm, the source workstation and propagation chain of quality anomalies can be quickly located, solving the technical challenges of difficult quality problem tracing and low processing efficiency in traditional methods. By dynamically adjusting the detection thresholds and judgment conditions of subsequent workstations based on the anomalies of preceding workstations, the passive mode of traditional fixed-parameter testing is changed, forming an adaptive intelligent testing mechanism. By recording the complete quality change trajectory of each LCD module from the first to the last workstation, a product-level quality archive is established. Through inter-workstation correlation analysis and risk prediction, potential quality risks can be identified in advance and preventive measures can be taken, effectively reducing the false positive rate and the missed detection rate, and improving the overall testing quality.

[0051] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0052] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0053] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assembling and testing a liquid crystal display module, characterized in that, include: The workstation relationships between the screen cable connection workstation, backlight system workstation, display effect workstation, touch function workstation, and appearance inspection workstation are calculated. When an abnormality is detected in the liquid crystal display module, the first workstation that is the source of the abnormality is traced according to the workstation association relationship. Based on the first workstation of the anomaly source, the impact of the anomaly on the subsequent second workstation of the anomaly source is calculated to obtain risk information; The detection threshold and judgment conditions of the second station are adjusted according to the risk information, and the detection is performed using the adjusted detection threshold and judgment conditions to generate the quality file of the liquid crystal display module.

2. The liquid crystal display module assembly and testing method according to claim 1, characterized in that, The workstation relationships between the computing screen cable connection workstation, backlight system workstation, display effect workstation, touch function workstation, and appearance inspection workstation include: Resistance values ​​are measured at the connection points between the LVDS screen cable and the LCD panel in the screen cable connection station to obtain resistance data; LED backlight strips in the backlight system station are individually lit up and brightness data is measured; a standard test image is input and contrast data is detected in the display effect station; multi-touch positioning tests are performed and response data is measured in the touch function station; and the surface of the LCD module is scanned using a CCD vision system in the appearance inspection station to identify scratches, bright spots, and dark spots to obtain defect data. The resistance data, brightness data, contrast data, response data, and defect data are synchronously integrated to obtain real-time workstation data; The correlation between screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects in the real-time data of the workstation are analyzed to obtain the workstation correlation relationship.

3. The liquid crystal display module assembly and testing method according to claim 2, characterized in that, The analysis examines the correlations between screen line resistance and backlight brightness, backlight uniformity and display contrast, display quality and touch accuracy, and touch accuracy and appearance defects in the real-time data of the workstations. This yields the workstation correlation relationships, including: The influence of the change in resistance data in the real-time data of the workstation on the uniformity of brightness data distribution is calculated to obtain the correlation between screen line resistance and backlight brightness. The influence of the uniformity deviation of the brightness data in the real-time data of the workstation on the measurement accuracy of the contrast data is calculated to obtain the correlation between backlight uniformity and display contrast. The impact of abnormal contrast data in the real-time data of the workstation on the accuracy of the response data is calculated to obtain the correlation between display quality and touch accuracy. The impact of the deviation of the response data in the real-time data of the workstation on the accuracy of defect data detection is calculated to obtain the correlation between touch accuracy and appearance defects; The correlation between screen line resistance and backlight brightness, the correlation between backlight uniformity and display contrast, the correlation between display quality and touch accuracy, and the correlation between touch accuracy and appearance defects are integrated to form a workstation correlation relationship.

4. The liquid crystal display module assembly and testing method according to claim 3, characterized in that, The calculation of the influence of the change in resistance data in the real-time data of the workstation on the uniformity of brightness data distribution yields the correlation between screen line resistance and backlight brightness, including: The resistance data and brightness data in the real-time data of the workstation are paired to obtain synchronous resistance-brightness data pairs; The synchronous resistance brightness data is compared with the center resistance data by a sliding window differential calculation to obtain the resistance change amplitude sequence; The uniformity variance of the brightness distribution is calculated by aligning the brightness data of the synchronous resistor with the brightness data according to the distribution area of ​​the LED backlight strip, and the brightness distribution uniformity index is obtained. By performing linear regression fitting with the resistance variation amplitude sequence as the independent variable and the brightness distribution uniformity index as the dependent variable, the correlation between screen line resistance and backlight brightness is obtained.

5. The liquid crystal display module assembly and testing method according to claim 4, characterized in that, When an abnormality is detected in the liquid crystal display module, the first workstation that traces the source of the abnormality according to the workstation association relationship includes: When an abnormality is detected in the liquid crystal display module, the quality parameter values ​​that deviate from the normal range in the real-time data of the workstation are identified and the abnormal workstation is determined. Query the quality parameters of the preceding workstations associated with the abnormal workstation in the workstation association relationship, and calculate the influence weight of the preceding workstation quality parameters on the current abnormality; Based on the influence weights, an anomaly propagation path is constructed from the workstation with the largest influence weight to the abnormal workstation, and the workstation with the largest influence weight value in the anomaly propagation path is selected as the first workstation of the anomaly source.

6. The liquid crystal display module assembly and testing method according to claim 5, characterized in that, The process of querying the quality parameters of the preceding workstations associated with the abnormal workstation in the workstation association relationship, and calculating the influence weight of the preceding workstation quality parameters on the current abnormality, includes: Retrieve the preceding workstation number that has a direct quality impact chain with the abnormal workstation from the workstation association relationship to obtain the associated workstation sequence; Extract the quality parameters of the preceding workstation of the associated workstation sequence from the real-time data of the workstation; Calculate the degree of deviation of the quality parameters in the preceding workstation quality parameters, and multiply the degree of deviation of the quality parameters by the corresponding correlation coefficient in the workstation correlation relationship to obtain the influence weight.

7. The liquid crystal display module assembly and testing method according to claim 1, characterized in that, The process of calculating the impact of the anomaly on the subsequent second workstation based on the first workstation at the source of the anomaly, and obtaining risk information, includes: Obtain the target abnormal quality parameters of the first workstation of the abnormal source, and determine the direct subsequent workstation of the first workstation in the production process as the second workstation; The severity of the anomaly at the first workstation is calculated based on the target anomaly quality parameters. The quality influence transmission coefficient between the first workstation and the second workstation is obtained from the workstation association relationship, and the quality influence transmission coefficient is multiplied by the anomaly severity of the first workstation to calculate the degree of influence of the anomaly on the second workstation. The risk level is determined based on the degree of impact of the anomaly on the second workstation, and corresponding risk information is generated.

8. The liquid crystal display module assembly and testing method according to claim 1, characterized in that, The process of adjusting the detection threshold and judgment conditions of the second station based on the risk information, and performing detection using the adjusted detection threshold and judgment conditions to generate a quality profile for the liquid crystal display module includes: Based on the risk level in the risk information, the detection threshold currently used in the second workstation is numerically corrected. When the risk level is high risk, the tolerance range of the detection threshold is expanded by a preset percentage to obtain the adjusted detection threshold. Based on the degree of impact in the aforementioned risk information, the quality qualification judgment logic of the second workstation is adjusted, and a secondary confirmation judgment step for expected abnormality types is added to obtain the adjusted judgment conditions. The detection is performed using the adjusted detection threshold and the adjusted judgment conditions to generate a quality profile for the liquid crystal display module.

9. The liquid crystal display module assembly and testing method according to claim 8, characterized in that, The process of performing detection using adjusted detection thresholds and adjusted judgment conditions to generate a quality profile for the liquid crystal display module includes: Configure the adjusted detection threshold and the adjusted judgment conditions into the detection system of the second station, and perform quality inspection on the LCD module to obtain the detection results of the second station; The detection results of the second workstation are correlated with the historical quality data of the liquid crystal display module at the screen cable connection workstation, backlight system workstation, display effect workstation, touch function workstation, and appearance inspection workstation. The evolution process of quality parameters at each workstation in the historical quality data is analyzed, and the time points and types of anomalies are marked to obtain a record of the evolution of quality status. The quality status evolution record, along with the product number and production batch information of the liquid crystal display module, is stored to obtain a quality file.

10. A liquid crystal display module assembly and testing system, characterized in that, A method for performing the liquid crystal display module assembly and testing as described in any one of claims 1-9, comprising: The calculation module is used to calculate the workstation relationships between the screen cable connection workstation, the backlight system workstation, the display effect workstation, the touch function workstation, and the appearance inspection workstation. The tracking module is used to track the first workstation that is the source of the abnormality when an abnormality is detected in the liquid crystal display module, based on the workstation association relationship. The risk analysis module is used to calculate the impact of the anomaly on the subsequent second workstation based on the first workstation of the anomaly source, and to obtain risk information. The generation module is used to adjust the detection threshold and judgment conditions of the second station according to the risk information, and to perform detection using the adjusted detection threshold and judgment conditions to generate the quality file of the liquid crystal display module.

Images