LED display screen manufacturing process parameter optimization method and system

Abstract

The invention relates to the technical field of LED display screen manufacturing, in particular to an LED display screen manufacturing process parameter optimization method and system. The method comprises the following steps: acquiring performance attenuation state information; evaluating and identifying a local temperature anomaly region; when a first deviation obtained by comparing the transient temperature response data with preset reference response data exceeds a first preset threshold value, the welding spot microscopic concealment defect is determined; establishing an association relationship; judging whether a prospective quality risk exists or not according to the association relationship; when the prospective quality risk reaches a second preset threshold value, executing an intervention strategy: executing partition process parameter compensation on the heating assembly associated with the local temperature abnormal area, and / or scheduling maintenance operation to overhaul or replace the heating assembly; and obtaining a new surface heat radiation image and an updated statistical result of the welding spot microscopic concealment defect, and performing closed-loop verification. The technical problem that local temperature is abnormal due to performance degradation of a reflow oven heating assembly in the LED display screen manufacturing process is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of LED display screen manufacturing, in particular to an LED display screen manufacturing process parameter optimization method and system. BACKGROUND

[0002] Reflow soldering is a key process in LED display screen manufacturing. However, the heating assembly gradually ages in long-term operation, and the thermal efficiency and response decrease, resulting in local abnormalities in the temperature field of the furnace cavity. As a result, some solder joints form hidden defects with normal appearance but slightly increased contact resistance, which are difficult to identify by AOI. Brightness testing only shows a slight overall darkening, which is misinterpreted as process drift. Since the control logic does not take into account the "heating assembly degradation", the system mistakenly attributes the problem to other causes and increases the peak temperature. This measure not only fails to eliminate local low temperatures, but also exacerbates temperature unevenness and component wear, further accelerating aging, forming a vicious cycle of "brightness reduction - warming - faster aging", and leading to a continuous decline in yield and equipment performance and difficulty in locating the root cause.

[0003] The prior art needs to be improved in view of the above problems. SUMMARY

[0004] The present application discloses an LED display screen manufacturing process parameter optimization method and system, aiming to solve the technical problem of local temperature abnormalities caused by the performance degradation of the heating assembly of the reflow soldering furnace in the LED display screen manufacturing process, which further leads to microscopic hidden defects in solder joints, and existing detection methods cannot effectively identify and lead to a vicious cycle.

[0005] The technical solution of the present application is as follows:

[0006] In a first aspect, the present application discloses an LED display screen manufacturing process parameter optimization method, comprising:

[0007] Obtaining performance degradation state information of the heating assembly of the reflow soldering furnace;

[0008] According to the performance degradation state information, in combination with the surface thermal radiation image collected when the LED display screen substrate leaves the reflow soldering furnace, the temperature distribution data of the reflow soldering furnace is inversely calculated, and the local temperature abnormal area of the reflow soldering furnace is evaluated and identified;

[0009] Collecting transient temperature response data of the solder joint, and determining that the solder joint is a microscopic hidden defect in the solder joint when a first deviation between the transient temperature response data and the preset reference response data exceeds a first preset threshold;

[0010] Establishing a correlation between the local temperature abnormal area and the corresponding heating assembly and the corresponding microscopic hidden defects in the solder joint;

[0011] Within the first preset time window, based on the preset risk assessment rule set, it is determined whether there are any forward-looking quality risks according to the correlation.

[0012] When a forward-looking quality risk reaches the second preset threshold, an intervention strategy is executed based on the preset parameter compensation mapping relationship: performing zonal process parameter compensation on heating components associated with local temperature anomalies to maintain the target temperature distribution, and / or scheduling maintenance operations to repair or replace the heating components.

[0013] After completing the intervention strategy, new surface thermal radiation images and updated statistical results of microscopic hidden defects in the solder joints are obtained for closed-loop verification. The verification results are then used to update temperature distribution data, performance degradation status information, and risk assessment rule set.

[0014] Furthermore, transient temperature response data of the solder joint is collected. When the first deviation between the transient temperature response data and the preset reference response data exceeds a first preset threshold, the solder joint is determined to be a microscopic hidden defect, including:

[0015] A thermal reference area adjacent to the solder joint is determined on the LED display substrate. The solder joint and the thermal reference area are synchronously excited and their respective transient temperature response data are collected synchronously.

[0016] The transient temperature response data of the thermal reference area is compared with its preset baseline response data to obtain the environmental impact factor. Based on the environmental impact factor, the transient temperature response data of the solder joint is calibrated to obtain the calibrated transient temperature response data.

[0017] The first deviation between the calibrated transient temperature response data and the reference response data is used as the criterion for judgment. When the first deviation exceeds the first preset threshold, the solder joint is determined to be a microscopic hidden defect of the solder joint.

[0018] Furthermore, after performing zoned process parameter compensation on the heating components associated with localized temperature anomalies to maintain the target temperature distribution, the process includes:

[0019] Acquire real-time data from the production environment and use it as external input;

[0020] By combining performance degradation status information with local temperature anomaly areas, and based on a preset parameter compensation mapping relationship, the compensation correction amount associated with the local temperature anomaly areas is determined, and updated partition process parameter compensation is generated.

[0021] Within the second preset time window, monitor the second deviation between the updated partition process parameter compensation and temperature distribution data, as well as the occurrence of microscopic hidden defects in the solder joints;

[0022] When the second deviation continues to exceed the third preset threshold, a secondary calibration is triggered. The secondary calibration includes updating the parameter compensation mapping relationship based on real-time data of the current production environment, performance degradation status information and temperature distribution data, and outputting new partition process parameter compensation.

[0023] Furthermore, when the second deviation continues to exceed a third preset threshold, a secondary calibration is triggered, including:

[0024] Track and store long-term historical data on temperature distribution and the occurrence of microscopic hidden defects in solder joints;

[0025] A reference baseline trend reflecting slow, cumulative changes is determined based on long-term historical data;

[0026] The second deviation is compared with the trend of the reference baseline to determine whether the second deviation is consistent with the trend of the reference baseline.

[0027] When the second deviation continues to exceed the third preset threshold within the second preset time window and is consistent with the trend of the reference baseline, the baseline used for determining the second deviation is adjusted, and the determination rule for the second deviation within the second preset time window is updated.

[0028] When the second deviation continues to exceed the third preset threshold within the second preset time window and is inconsistent with the trend of the reference baseline, the second deviation is determined to be the actual deviation caused by poor parameter adjustment effect, and a secondary calibration is triggered to redetermine the compensation of the partition process parameters.

[0029] Furthermore, after determining whether there are forward-looking quality risks based on related relationships, this also includes:

[0030] When there are multiple forward-looking quality risks, these risks are prioritized based on their severity, the scope of their impact on product quality, and the anticipated urgency.

[0031] Identify the correlations between multiple forward-looking quality risks;

[0032] Assess the combined impact of intervention strategies on multiple prospective quality risks;

[0033] Based on priority ranking, correlation, and overall impact, the corresponding intervention strategy is selected.

[0034] Furthermore, the combined impact of intervention strategies on multiple prospective quality risks was assessed, including:

[0035] Based on correlation, a set of associated risks is determined for each candidate intervention strategy.

[0036] Quantify the synergistic effects of intervention strategies on each forward-looking quality risk in the associated risk set;

[0037] Quantify the offsetting effect of intervention strategies on each forward-looking quality risk in the associated risk set;

[0038] The combined impact of this intervention strategy is generated based on the synergistic and offsetting effects.

[0039] Furthermore, the synergistic effects of intervention strategies on each forward-looking quality risk in the associated risk set are quantified, including:

[0040] Before implementing the intervention strategy, record the baseline status of temperature distribution data and the occurrence of microscopic hidden defects in the solder joints;

[0041] During the implementation of the intervention strategy, temperature distribution data and the occurrence of microscopic hidden defects in the solder joints were continuously monitored.

[0042] Simultaneously acquire key parameters of the production environment;

[0043] By comparing the changes in baseline status before and after the implementation of the intervention strategy, and combining the real-time fluctuations of key parameters in the production environment, the accidental impact of non-intervention factors on temperature distribution data and the occurrence of microscopic hidden defects in solder joints can be identified and eliminated.

[0044] The change in baseline status after deducting the impact of chance will be used as the synergistic effect of the intervention strategy on each forward-looking quality risk in the associated risk set.

[0045] Furthermore, the offsetting effects of intervention strategies on each forward-looking quality risk in the associated risk set are quantified, including:

[0046] Based on the type and intensity of the intervention strategy, predict the secondary risks it will cause, including long-term cumulative secondary risks and / or explosive secondary risks.

[0047] Continuous monitoring of microstructural changes in equipment components involved in intervention strategies is used to characterize the secondary risk evolution state of material fatigue and / or microcrack initiation.

[0048] The secondary risk evolution index is calculated based on the secondary risk evolution state and compared with the fourth preset threshold.

[0049] When the secondary risk evolution index reaches the fourth preset threshold, the secondary risk is quantified as an offsetting effect.

[0050] Furthermore, within the first preset time window, when determining whether there is a forward-looking quality risk based on the preset risk assessment rule set and the correlation, the following is also included:

[0051] The performance degradation status information, the local temperature anomaly areas identified in the temperature distribution data, and the microscopic hidden defects of the solder joints are matched according to their spatial location.

[0052] When performance degradation status information, local temperature anomaly area, and microscopic hidden defects of solder joints appear consecutively in the same spatial area and meet the preset sequential relationship, the risk level of that spatial area is increased.

[0053] Based on the improvement of risk level, the trigger thresholds in the risk assessment rule set are adaptively tightened or relaxed according to spatial region.

[0054] Secondly, this application also discloses an LED display manufacturing process parameter optimization system, comprising:

[0055] The performance degradation status information acquisition module is used to acquire the performance degradation status information of the heating components in the reflow oven.

[0056] The local temperature anomaly assessment module is used to reverse-calculate the temperature distribution data of the reflow oven based on the performance degradation status information and the surface thermal radiation image collected when the LED display substrate leaves the reflow oven, and to assess and identify the local temperature anomaly areas of the reflow oven.

[0057] The microscopic hidden defect identification module is used to collect transient temperature response data of solder joints. When the first deviation between the transient temperature response data and the preset reference response data exceeds the first preset threshold, the solder joint is identified as a microscopic hidden defect.

[0058] The correlation analysis module is used to establish the correlation between local temperature anomaly areas and corresponding heating components and corresponding microscopic hidden defects in solder joints;

[0059] The forward-looking quality risk assessment module is used to determine whether there are forward-looking quality risks within a first preset time window, based on a preset set of risk assessment rules and correlations.

[0060] The intervention strategy scheduling and execution module is used to execute intervention strategies based on the preset parameter compensation mapping relationship when the prospective quality risk reaches the second preset threshold: perform zonal process parameter compensation for heating components associated with local temperature anomalies to maintain the target temperature distribution, and / or schedule maintenance operations to repair or replace heating components.

[0061] The closed-loop verification module is used to acquire new surface thermal radiation images and updated statistical results of microscopic hidden defects in solder joints after the intervention strategy is completed, to perform closed-loop verification, and to backfill the verification results to update temperature distribution data, performance degradation status information, and risk assessment rule set.

[0062] Beneficial effects

[0063] The LED display manufacturing process parameter optimization method disclosed in this application first obtains the performance degradation state of the heating components and, combined with the thermal radiation image of the substrate surface, reversely calculates the temperature distribution inside the furnace to accurately identify local temperature anomalies. Then, it collects the transient temperature response of the solder joints and compares it with a benchmark to identify microscopic hidden defects that are difficult to detect with AOI. This establishes a correlation between "abnormal area—corresponding heating component—defective solder joint," and assesses forward-looking quality risks according to a rule set within a preset time window. When the risk reaches a threshold, it performs partition compensation and / or maintenance according to parameter compensation mapping. Closed-loop verification is used to backfill the results, continuously updating data and rules to achieve adaptive optimization, avoiding vicious cycles caused by misjudgments, reducing defects, and improving welding quality, yield, and long-term reliability. Attached Figure Description

[0064] Figure 1 This application provides a flowchart illustrating a method for optimizing manufacturing process parameters of an LED display screen.

[0065] Figure 2 A flowchart of a process parameter optimization system for LED display manufacturing provided in this application.

[0066] The diagram shows: 1. Performance degradation status information acquisition module; 2. Local temperature anomaly assessment module; 3. Microscopic hidden defect identification module; 4. Correlation analysis module; 5. Proactive quality risk assessment module; 6. Intervention strategy scheduling and execution module; 7. Closed-loop verification module. Detailed Implementation

[0067] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0068] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0069] Reflow soldering is a critical step in LED display manufacturing, requiring precise and controllable process parameters. During long-term operation, heating components undergo progressive aging (such as high-temperature oxidation of resistance wires and the formation of heat-insulating films on the surface), leading to a decrease in heat transfer efficiency and response speed. Even with the same energy input, the furnace temperature field is difficult to maintain uniformity and stability, resulting in localized low temperatures or increased fluctuations. This easily leads to hidden solder joint defects (insufficient IMC, microvoids, uneven grain size, etc.), appearing normal but with slightly increased contact resistance. Existing AOI (Automated Optical Inspection) is biased towards visual inspection and struggles to identify such defects, causing risky products to be mistakenly deemed acceptable. While finished product brightness testing can capture slight overall decreases and interpret them as "process drift," the central control system does not incorporate "heating component attenuation" factors, often misjudging the root cause and choosing to increase peak temperatures. This practice exacerbates temperature field unevenness and component wear, accelerates aging, and worsens defects, creating a vicious cycle of "brightness decrease—continued heating—further equipment aging," ultimately causing a continuous decline in yield and equipment performance with difficulty in pinpointing the root cause.

[0070] Reference Figure 1 In response, this application proposes a method for optimizing process parameters in LED display manufacturing, including:

[0071] S1000: Obtain information on the performance degradation status of the heating components in the reflow oven;

[0072] S2000: Based on the performance degradation status information and combined with the surface thermal radiation image collected when the LED display substrate leaves the reflow oven, the temperature distribution data of the reflow oven is reverse-calculated to evaluate and identify local temperature anomaly areas in the reflow oven.

[0073] S3000: Collect transient temperature response data of solder joints. When the first deviation between the transient temperature response data and the preset reference response data exceeds the first preset threshold, the solder joint is determined to be a microscopic hidden defect.

[0074] S4000: Establish the correlation between local temperature anomaly areas and the corresponding heating components and the corresponding solder joint microscopic hidden defects;

[0075] S5000: Within the first preset time window, based on the preset risk assessment rule set, determine whether there are any forward-looking quality risks according to the correlation;

[0076] S6000: When a forward-looking quality risk reaches the second preset threshold, an intervention strategy is executed based on the preset parameter compensation mapping relationship: performing zonal process parameter compensation on heating components associated with local temperature anomalies to maintain the target temperature distribution, and / or scheduling maintenance operations to repair or replace heating components.

[0077] S7000: After completing the intervention strategy, acquire new surface thermal radiation images and updated statistical results of microscopic hidden defects in solder joints, perform closed-loop verification, and backfill the verification results to update temperature distribution data, performance degradation status information, and risk assessment rule set.

[0078] Specifically, the performance degradation information of the heating components in a reflow oven refers to real-time or historical data showing a decline in performance indicators such as heating efficiency, temperature control accuracy, and response speed due to factors such as material aging, wear, and contamination during long-term operation. This information can include changes in resistance, power output fluctuations, and thermocouple reading deviations.

[0079] A surface thermal radiation image refers to an image of the surface temperature distribution of an LED display substrate as it leaves the reflow oven, captured by equipment such as an infrared thermal imager. This image can visually reflect the uniformity of heating of the substrate during the soldering process.

[0080] Temperature distribution data refers to the temperature measurement or estimated data in different areas and at different times inside the reflow oven, used to characterize the thermal environment inside the oven cavity.

[0081] Localized temperature anomalies refer to specific areas within the reflow oven where the temperature distribution deviates from the target setting or uniformity standard, potentially leading to welding defects.

[0082] Transient temperature response data of solder joints refers to the dynamic curve of the temperature change of the solder joint over time when it is subjected to external stimuli (such as microcurrent heating, laser pulse, etc.). This data can reflect the thermal and electrical properties of the solder joint, and thus reveal its microstructural integrity.

[0083] The baseline response data refers to the transient temperature response data of the solder joint under ideal conditions (without defects), which serves as a reference standard for defect judgment.

[0084] Microscopic hidden defects in solder joints refer to tiny structural problems that are difficult to detect with the naked eye, such as voids, uneven grains, and incomplete intermetallic compound layers. These defects may lead to increased contact resistance and affect the long-term reliability of the product.

[0085] The parameter compensation mapping relationship refers to the preset correspondence between local temperature anomaly areas, heating component performance degradation and the required process parameter adjustment amount, which is used to guide the implementation of intervention strategies.

[0086] Zonal process parameter compensation refers to adjusting the power, time, or temperature settings of heating components in different areas of the reflow oven based on their performance degradation and local temperature anomalies, in order to achieve precise control of the temperature field inside the oven.

[0087] This method is typically implemented in an industrial automation environment that integrates sensors, a data acquisition system, a data processing unit (such as an industrial computer or server), an actuator (such as a reflow oven controller), and a human-machine interface. Sensors are responsible for collecting performance degradation status information, surface thermal radiation images, and transient temperature response data of the solder joints; the data processing unit is responsible for data analysis, calculation, identification, correlation, and risk assessment; and the actuator is responsible for adjusting parameters or scheduling and maintaining the reflow oven according to the intervention strategy.

[0088] Performance degradation of the heating element is one of the root causes of reflow soldering quality problems. Several methods can be used to obtain this information. For example, a temperature sensor can be installed on the heating element to continuously monitor its surface or internal temperature; deviations from theoretical values ​​in temperature readings can indicate performance degradation. Another method is to monitor the heating element's current, voltage, or power consumption; abnormal fluctuations in these electrical parameters under the same output requirements can also indicate performance degradation. Furthermore, periodic infrared thermal imaging of the heating element can analyze its heat distribution uniformity, identifying areas of localized overheating or poor heat dissipation, thereby obtaining information on performance degradation status.

[0089] Secondly, after obtaining the performance degradation status information of the heating components, it is necessary to combine it with the actual temperature distribution. For example, a thermocouple array can be used to measure the temperature at multiple points inside the reflow oven to obtain preliminary temperature distribution data. Simultaneously, when the LED display substrate leaves the reflow oven, an infrared thermal imager is used to capture its surface thermal radiation image, which reflects the final temperature field experienced by the substrate inside the oven. By inputting these surface thermal radiation images and the performance degradation status information of the heating components into a pre-established thermodynamic model, more refined temperature distribution data inside the oven can be calculated in reverse, thereby assessing and identifying potential localized temperature anomalies within the oven. For example, if the heating components in a certain area exhibit performance degradation, and the corresponding substrate surface thermal radiation image shows that the temperature in that area is lower than normal, then that area can be identified as a localized temperature anomaly.

[0090] Furthermore, to identify microscopic hidden defects that are difficult to detect with the naked eye, this application employs a method based on transient temperature response data. For example, a tiny thermal pulse (such as by laser or microcurrent) can be applied to the solder joint, and the temperature change curve of the solder joint under the thermal pulse can be acquired in real time using a high-precision infrared sensor or thermocouple array. Simultaneously, the same test is performed on an ideal, defect-free solder joint to obtain its baseline response data. The actual acquired transient temperature response data of the solder joint is compared with the baseline response data, and a first deviation between the two is calculated. If this first deviation (e.g., differences in peak temperature, heating rate, cooling rate, or thermal diffusivity) exceeds a preset first threshold, it can be determined that the solder joint has a microscopic hidden defect.

[0091] Next, this method establishes a correlation between localized temperature anomalies and corresponding heating components and microscopic hidden defects in solder joints. To achieve precise intervention, the root cause of the problem needs to be identified. For example, during the production process, the specific location and time of each LED display substrate passing through the reflow oven can be recorded and matched with the location and time of occurrence of localized temperature anomalies within the oven. Simultaneously, the location information of identified microscopic hidden defects in solder joints is correlated with these localized temperature anomalies. Through long-term data accumulation and statistical analysis, a correlation can be established that "which heating component's performance degradation led to which localized temperature anomaly area, and consequently, which solder joint exhibited microscopic hidden defects." This correlation can be based on a machine learning model trained on historical data or on an expert rule base.

[0092] Within a first preset time window, based on a preset risk assessment rule set, the system determines whether a forward-looking quality risk exists according to the correlation. After establishing the aforementioned correlation, the system needs to conduct a forward-looking assessment of potential quality risks. For example, a first preset time window can be set (such as the past 24 hours or one week). Within this time window, the system continuously monitors the performance degradation status of the heating components, the frequency and severity of local temperature anomalies, and the incidence of microscopic hidden defects in solder joints. This real-time data is input into the preset risk assessment rule set. The risk assessment rule set can contain a series of logical judgments, such as: "If the performance degradation index of a certain heating component exceeds the threshold for 3 consecutive hours, and the corresponding local temperature anomaly area persists, while the incidence of microscopic hidden defects in solder joints in that area increases by more than 10% within the first preset time window, then a forward-looking quality risk is determined to exist."

[0093] When a forward-looking quality risk reaches a second preset threshold, an intervention strategy is executed based on a preset parameter compensation mapping relationship: zoned process parameter compensation is performed on heating components associated with localized temperature anomalies to maintain the target temperature distribution, and / or maintenance operations are scheduled to repair or replace the heating components. Once a forward-looking quality risk is identified and reaches the preset second preset threshold (e.g., the risk level reaches "high" or the risk probability exceeds 80%), the system will immediately trigger the intervention strategy. The intervention strategy can be executed in several ways. For example, if the risk is due to a slight performance degradation of a heating component causing a localized low temperature, the system can perform zoned process parameter compensation on the area where the heating component is located, based on a preset parameter compensation mapping relationship. For example, it can appropriately increase the heating power or extend the heating time in that area to maintain the target temperature distribution within the furnace. Alternatively, if the performance degradation of the heating component is severe, or if zoned process parameter compensation cannot effectively solve the problem, the system can schedule maintenance operations, for example, automatically generating a maintenance work order to notify technicians to repair or replace the heating component.

[0094] Finally, after implementing the intervention strategy, new surface thermal radiation images and updated statistical results of microscopic hidden defects in the solder joints are acquired for closed-loop verification. The verification results are then used to update temperature distribution data, performance degradation status information, and the risk assessment rule set. The effectiveness of the intervention strategy needs to be confirmed through closed-loop verification. For example, after performing zonal process parameter compensation or maintenance operations, the system will again acquire surface thermal radiation images of the LED display substrate as it leaves the reflow oven and continuously monitor the occurrence of microscopic hidden defects in the solder joints. By comparing the surface thermal radiation images and defect statistics before and after the intervention, it is possible to assess whether the intervention strategy has successfully eliminated local temperature anomalies and reduced defects. If the verification results show that the intervention is effective, this positive feedback information is fed back into the system to update temperature distribution data, performance degradation status information of the heating components, and the risk assessment rule set, enabling the system to continuously learn and optimize its decision-making capabilities. If the verification results show that the intervention is ineffective, the system may trigger deeper analysis or adjustments to ensure that the problem is completely resolved.

[0095] The LED display manufacturing process parameter optimization method proposed in this application achieves "early identification, accurate diagnosis, rapid intervention, and closed-loop control" for key reflow soldering processes through multi-source data fusion. The process involves: continuously acquiring performance degradation information of the heated components; combining this with surface thermal radiation images of the substrate upon removal from the furnace to reverse-calculate the furnace temperature distribution, thereby accurately locating localized temperature anomalies; simultaneously acquiring transient temperature responses of solder joints and comparing them with benchmarks to identify microscopic hidden defects that are difficult to detect with traditional AOI. Subsequently, a spatial correlation is established between "component degradation—localized temperature anomaly—defect location," and pre-defined rules are used to assess forward-looking quality risks within the first time window and provide graded early warnings.

[0096] When the risk reaches a threshold, the system performs partitioned process parameter compensation based on the parameter compensation mapping relationship, or schedules maintenance to repair / replace attenuation components, preventing defect propagation and further equipment aging. After intervention, the results are backfilled through closed-loop verification (re-sampling thermal radiation maps and defect statistics), dynamically updating temperature distribution data, attenuation status, and risk rule sets to form an adaptive and self-optimizing control loop. Compared to post-processing correction relying on AOI and finished product brightness, this method can detect the root cause of attenuation early, accurately locate anomalies, and implement targeted compensation, avoiding the vicious cycle of "heating to cover up problems—accelerating aging," and significantly improving welding quality and long-term reliability.

[0097] In another embodiment of this application, step S3000 is further proposed to include:

[0098] S3100: Determine a thermal reference area adjacent to the solder joint on the LED display substrate;

[0099] S3200: Synchronously excites the solder joint and the thermal reference area and simultaneously collects their respective transient temperature response data;

[0100] S3300: The transient temperature response data of the thermal reference area is compared with its preset baseline response data to obtain the environmental impact factor. Based on the environmental impact factor, the transient temperature response data of the solder joint is calibrated to obtain the calibrated transient temperature response data.

[0101] S3400: The first deviation between the calibrated transient temperature response data and the reference response data is used as the judgment criterion. When the first deviation exceeds the first preset threshold, the solder joint is determined to be a microscopic hidden defect of the solder joint.

[0102] Specifically, the thermal reference area refers to a selected region on the LED display substrate adjacent to the solder joint to be tested. This region is typically designed to have known and stable thermal response characteristics. For example, it could be a known defect-free solder joint area or a blank area with standard thermal conductivity characteristics. Its purpose is to provide a benchmark for the environmental thermal response, distinguishing between the influence of environmental factors on the temperature response and abnormal temperature response caused by defects in the solder joint itself.

[0103] Synchronous excitation of the solder joint and the thermal reference area can be understood as applying the same heat source or energy input, such as laser pulses, current heating, or high-frequency acoustic wave excitation, so that the solder joint and the thermal reference area produce transient temperature responses at the same time point and under the same intensity. Synchronous acquisition of their respective transient temperature response data refers to using high-precision infrared thermal imagers, thermocouple arrays, or non-contact temperature sensors to record the temperature change curves of the solder joint and the thermal reference area over time, either simultaneously with or immediately after the excitation. The purpose is to ensure that data is acquired under the same environmental and excitation conditions for accurate subsequent environmental calibration.

[0104] In practical applications, the transient temperature response data of the thermal reference area is compared with its preset baseline response data to obtain the environmental impact factor. For example, the degree of environmental influence on the temperature response can be quantified by calculating the difference, ratio, or more complex statistical models between the two. The environmental impact factor can be a scalar, a function, or a vector, used to characterize the overall offset or distortion of the thermal response due to current environmental conditions. Environmental calibration is then performed on the transient temperature response data of the solder joint based on the environmental impact factor. For example, the environmental impact factor can be subtracted from or proportionally adjusted from the original transient temperature response data of the solder joint to eliminate or weaken the interference of environmental factors, thereby obtaining calibrated transient temperature response data that more accurately reflects the thermal characteristics of the solder joint itself.

[0105] The following is a specific example to illustrate this:

[0106] Suppose that on an LED display substrate, a thermal reference area with the same size and material properties as a standard solder joint but known to be defect-free is placed next to a solder joint to be inspected. During defect detection, a short-pulse thermal excitation is first applied simultaneously to both the solder joint and the thermal reference area using a high-precision laser. Then, a high-speed infrared thermal imager synchronously records the transient temperature decay curves of these two areas after excitation. For example, the transient temperature response data of the thermal reference area at a certain time point T1 may be 0.5°C higher than its preset reference response data; this 0.5°C deviation is determined as the environmental impact factor under the current environment. Subsequently, this 0.5°C environmental impact factor is subtracted from the original transient temperature response data of the solder joint to be inspected at the same time point T1 to obtain calibrated transient temperature response data. If the first deviation between the calibrated transient temperature response data of the solder joint and the preset reference response data (e.g., the temperature difference or decay rate difference at a critical time point) exceeds a first preset threshold (e.g., a 0.2°C or 10% decay rate difference), the solder joint is determined to have a microscopic hidden defect. This method ensures that even under conditions of large fluctuations in ambient temperature, minute defects inside the solder joint, such as voids, cracks, or poor wetting, can be accurately identified, thereby improving the reliability of defect detection.

[0107] In another embodiment of this application, a sub-step of S6000 is further proposed: after performing zonal process parameter compensation on the heating components associated with the local temperature anomaly region to maintain the target temperature distribution, it includes:

[0108] A1: Acquire real-time data from the production environment and use it as external input;

[0109] A2: Combining performance degradation status information with local temperature anomaly areas, determine the compensation correction amount associated with the local temperature anomaly amount based on the preset parameter compensation mapping relationship, and generate updated partition process parameter compensation.

[0110] A3: Within the second preset time window, monitor the second deviation between the updated partition process parameter compensation and temperature distribution data, as well as the occurrence of microscopic hidden defects in the solder joints;

[0111] A4: When the second deviation continues to exceed the third preset threshold, a secondary calibration is triggered. The secondary calibration includes updating the parameter compensation mapping relationship based on real-time data of the current production environment, performance degradation status information and temperature distribution data, and outputting new partition process parameter compensation.

[0112] Specifically, acquiring real-time data from the production environment and using it as external input refers to the system continuously collecting environmental parameters related to the operation of the reflow oven, such as ambient temperature, humidity, airflow speed, and power supply voltage fluctuations. This data is considered external input to reflect dynamic changes during the production process, providing more comprehensive contextual information for subsequent compensation and correction.

[0113] Specifically, by combining performance degradation status information with localized temperature anomaly areas, and based on a preset parameter compensation mapping relationship, the system determines the compensation correction amount associated with the localized temperature anomaly areas, generating updated partitioned process parameter compensation. This can be understood as the system not being permanently fixed after the initial compensation execution. It re-evaluates and calculates the required compensation correction amount based on the latest heating component performance degradation status information and the identified localized temperature anomaly areas. This correction amount is derived from the preset parameter compensation mapping relationship and aims to fine-tune or optimize the implemented partitioned process parameter compensation, thereby generating a more accurate and adaptable updated partitioned process parameter compensation.

[0114] In practical applications, within the second preset time window, the second deviation between the updated partition process parameter compensation, temperature distribution data, and the occurrence of microscopic hidden defects in solder joints refers to the system continuously tracking the actual temperature distribution data of the reflow oven and the occurrence of microscopic hidden defects in solder joints on the LED display substrate within a specific time period (the second preset time window) after executing the updated partition process parameter compensation. By comparing these real-time data with the desired target temperature distribution and the expected defect incidence rate, the second deviation can be calculated. This deviation quantifies the difference between the current compensation effect and the ideal state.

[0115] When the second deviation consistently exceeds the third preset threshold, a secondary calibration is triggered. "Consistently exceeding" here means that the second deviation is not a random, instantaneous fluctuation, but rather exhibits a certain persistence or trend, indicating that the current compensation effect may have systemic problems or deficiencies. Once this persistent deviation reaches or exceeds the preset third threshold, the system will trigger the secondary calibration mechanism. Secondary calibration involves updating the parameter compensation mapping relationship based on real-time data from the current production environment, performance degradation status information, and temperature distribution data, and outputting new partitioned process parameter compensations. This means that secondary calibration is not simply a repetition of previous compensation, but rather utilizes the latest and most comprehensive data (including real-time production environment data, the latest performance degradation status information, and current temperature distribution data) to learn and optimize the original parameter compensation mapping relationship, thereby generating a completely new, more accurate, and more robust partitioned process parameter compensation scheme.

[0116] The proposed solution effectively addresses the issues of insufficient timeliness and robustness in the compensation of partitioned process parameters in the basic scheme by introducing a dynamic and adaptive feedback and calibration mechanism.

[0117] Furthermore, A4 includes:

[0118] A41: Track and store long-term historical data on temperature distribution and the occurrence of microscopic hidden defects in solder joints;

[0119] A42: Determine reference baseline trends that reflect slow, cumulative changes based on long-term historical data;

[0120] A43: Compare the second deviation with the reference baseline trend to determine whether the second deviation is consistent with the reference baseline trend;

[0121] A44: When the second deviation continues to exceed the third preset threshold within the second preset time window and is consistent with the trend of the reference baseline, adjust the baseline used for determining the second deviation and update the determination rules for the second deviation within the second preset time window.

[0122] A45: When the second deviation continues to exceed the third preset threshold within the second preset time window and is inconsistent with the trend of the reference baseline, the second deviation is determined to be the real deviation caused by poor parameter adjustment effect, and a secondary calibration is triggered to redetermine the compensation of the partition process parameters.

[0123] The long-term historical data tracking and storing temperature distribution data and the occurrence of microscopic hidden defects in solder joints refers to the system continuously recording and saving detailed data on the internal temperature distribution of the reflow oven during long-term operation, as well as information such as the frequency, type, and location of microscopic hidden defects in LED display solder joints. This long-term historical data can cover production cycles of several weeks, months, or even years, with the aim of providing a sufficient data foundation for subsequent trend analysis.

[0124] Determining a reference baseline trend based on long-term historical data, reflecting slow and cumulative changes, refers to establishing a reference model or curve that characterizes the slow and cumulative changes in temperature distribution and solder joint defects of a reflow oven under normal operating conditions through statistical analysis, pattern recognition, or machine learning methods on the stored long-term historical data. This reference baseline trend can be understood as the natural evolution of the system under conditions without external intervention or major failures, such as normal aging of heating components and minor drifts in material properties.

[0125] Therefore, comparing the second deviation with the reference baseline trend to determine whether the second deviation is consistent with the reference baseline trend involves comparing the second deviation between the currently monitored updated zonal process parameter compensation, temperature distribution data, and the occurrence of microscopic hidden defects in solder joints with the predetermined reference baseline trend. This comparison can be performed by calculating the correlation and difference between the two or by using statistical hypothesis testing to assess whether the current deviation is part of the normal evolution of the system or deviates from the expected trajectory.

[0126] Specifically, when the second deviation consistently exceeds the third preset threshold within the second preset time window and its trend aligns with the reference baseline, the baseline used for determining the second deviation is adjusted, and the determination rules for the second deviation within the second preset time window are updated. This means that if the current deviation exceeds the immediate threshold, but its change pattern matches the long-term cumulative change trend of the system, it indicates that this deviation may be caused by normal system aging or slow environmental changes, rather than an immediate failure of the current parameter compensation strategy. In this case, the system will not immediately trigger a secondary calibration, but will adaptively adjust the baseline or determination rules used to determine the second deviation, such as appropriately relaxing the threshold or modifying the determination logic, to adapt to this expected, slow system evolution and avoid unnecessary frequent calibrations.

[0127] When the second deviation continuously exceeds the third preset threshold within the second preset time window and its trend is inconsistent with the reference baseline, the second deviation is determined to be a real deviation caused by poor parameter adjustment, and a secondary calibration is triggered to redetermine the compensation of the partition process parameters. This indicates that if the current deviation not only exceeds the immediate threshold, but its change pattern is also significantly inconsistent with the long-term cumulative change trend of the system, the system will determine that the current parameter compensation strategy has failed to effectively maintain the target temperature distribution or control the occurrence of solder joint defects, i.e., the parameter adjustment effect is poor. At this time, the system will immediately trigger a secondary calibration to update the parameter compensation mapping relationship based on the real-time data of the current production environment, performance degradation status information, and temperature distribution data, and output new partition process parameter compensation to correct the actual process problems.

[0128] In some preferred embodiments, the following specific example illustrates the situation:

[0129] This solution addresses temperature deviations caused by the slow aging of reflow oven heating components due to long-term use, proposing a historical trend-based discrimination and adaptive calibration mechanism. The system continuously tracks and stores temperature distribution data and the occurrence of microscopic hidden defects in solder joints within the target area, extracting a reference baseline trend reflecting normal aging (e.g., a gradual increase of approximately 0.1℃ per month). When a second deviation is detected to consistently exceed a third preset threshold within a second preset time window, the deviation is first compared with the reference baseline trend: if the change pattern matches the trend, it is determined to be normal aging, and a secondary calibration is not immediately triggered. Instead, the baseline and judgment rules for the second deviation are adjusted, and the third preset threshold for that spatial area is appropriately tightened / relaxed to avoid frequent, ineffective calibrations and overcompensation. If the second deviation is significantly inconsistent with the reference baseline trend (e.g., a sudden sharp increase or abnormal fluctuation), it is considered a genuine deviation caused by poor parameter adjustment, and a secondary calibration is immediately triggered: based on real-time data from the current production environment, performance degradation status information, and temperature distribution data, the parameter compensation mapping relationship is updated, and new zoned process parameter compensations are output to quickly correct process defects. This enables a precise distinction between "normal system evolution" and "abnormal process problems," reducing unnecessary interventions and improving calibration effectiveness and production stability.

[0130] In another embodiment of this application, after determining whether a forward-looking quality risk exists based on the correlation in sub-step S5000, the method further includes:

[0131] B1: When there are multiple forward-looking quality risks, prioritize them according to the severity of the risk, the scope of its impact on product quality, and the expected urgency.

[0132] B2: Identify the correlations between multiple forward-looking quality risks;

[0133] B3: Assess the combined impact of intervention strategies on multiple prospective quality risks;

[0134] B5: Select the corresponding intervention strategy based on priority ranking, correlation, and overall impact.

[0135] Specifically, when the system identifies multiple forward-looking quality risks, these risks need to be prioritized. Prioritization can be based on the severity of the risk; for example, risks with a greater impact on product functionality should have higher priority; the scope of the impact on product quality, such as risks affecting the entire batch of products rather than a single product, should be addressed first; and the anticipated urgency, such as risks that will lead to production disruptions or major defects, should be addressed immediately. These assessment criteria can be pre-set in a risk assessment rule set and dynamically adjusted based on expert experience or historical data.

[0136] Identifying the correlations between multiple forward-looking quality risks involves analyzing whether there are causal relationships, synergistic effects, or mutual influences among different risks. For example, the performance degradation of a heating component may simultaneously lead to multiple localized temperature anomalies, which in turn can cause microscopic, hidden defects in solder joints at different locations. By establishing these correlations, a more comprehensive understanding of the root causes and propagation paths of risks can be achieved.

[0137] Building on this, assessing the combined impact of intervention strategies on multiple forward-looking quality risks involves predicting the overall impact of an intervention strategy on all relevant risks before selecting one. This includes evaluating the potential positive synergistic effects of the strategy (e.g., an intervention addressing multiple related risks simultaneously) and the potential negative offsetting effects or secondary risks (e.g., addressing one risk may exacerbate another).

[0138] Ultimately, based on priority ranking, correlation, and overall impact, the corresponding intervention strategy is selected. This means that the selected intervention strategy must not only address high-priority risks but also consider the correlation between risks, ensuring that its overall impact is positive and optimal. For example, a strategy that can simultaneously address multiple high-priority related risks can be chosen, or a strategy that best reduces the overall risk level can be selected.

[0139] The solution proposed in this application effectively addresses the limitations that traditional single-risk handling methods may have when multiple risks coexist in a complex production environment by introducing a prioritization of multiple forward-looking quality risks, identification of their correlations, and a comprehensive impact assessment of intervention strategies.

[0140] In some preferred embodiments:

[0141] Suppose that during the LED display manufacturing process, the system identifies three forward-looking quality risks:

[0142] Risk A: Performance degradation of the first heating element in the reflow oven may lead to a lower temperature in Zone 1, potentially causing poor solder joints.

[0143] Risk B: Performance degradation of the second heating component in the reflow oven may lead to higher temperatures in Zone 2, potentially causing solder joint oxidation defects.

[0144] Risk C: Performance degradation of the third heating component in the reflow oven leads to large temperature fluctuations in zone three, which may cause fatigue defects in the solder joints.

[0145] First, prioritize the risks. Assuming that based on historical data and expert experience, cold solder joint defects (risk A) have the most severe impact on product reliability, oxidation defects (risk B) have a wider impact, and fatigue defects (risk C) have a lower urgency, then the priority might be: Risk A > Risk B > Risk C.

[0146] Secondly, risk correlations were identified. Analysis revealed that the performance degradation of both the first and second heating components might be related to quality issues with components from the same batch of suppliers, indicating a connection within the upstream supply chain. Furthermore, compensating for the process parameters of the first heating component in different zones might slightly affect the temperature distribution of the adjacent second heating component.

[0147] Next, the overall impact of the intervention strategy is assessed.

[0148] Strategy 1: Compensate the process parameters for the first heating component only. This may resolve risk A, but has no direct effect on risks B and C, and may even slightly exacerbate risk B.

[0149] Strategy Two: Simultaneously compensate the process parameters of the first and second heating components in separate zones. This may address both risks A and B, but requires more complex control and may indirectly affect the third heating component.

[0150] Strategy 3: Schedule maintenance operations and replace all three heating components. This completely eliminates all risks, but it is costly and involves long downtime.

[0151] Finally, based on priority ranking, correlation, and overall impact, the corresponding intervention strategy is selected. Considering that risks A and B have high priority and are correlated, and that strategy two can simultaneously address these two main risks while its impact on risk C is controllable, strategy two is chosen as the optimal intervention strategy. If the implementation cost and downtime of strategy two are within acceptable limits, it will be implemented first. If residual risks or new secondary risks remain after the implementation of strategy two, maintenance operations for risk C can be further considered.

[0152] In another embodiment of this application, step B3 is further proposed to include:

[0153] B31: Based on correlation, determine the corresponding set of associated risks for each candidate intervention strategy;

[0154] B32: Quantify the synergistic effects of intervention strategies on each forward-looking quality risk in the associated risk set;

[0155] B33: Quantify the offsetting effect of intervention strategies on each forward-looking quality risk in the associated risk set;

[0156] B34: Generate the combined impact of this intervention strategy based on synergistic and offsetting effects.

[0157] Specifically, correlation refers to the interrelationships between risks identified when recognizing the connections between multiple forward-looking quality risks. Based on this correlation, the set of risks that each candidate intervention strategy under consideration affects or is associated with can be determined. For example, suppose an intervention strategy aims to adjust the parameters of a heating component in a reflow oven. The performance degradation of this heating component may be associated with multiple localized temperature anomalies and the resulting microscopic defects in the solder joints. In this case, these localized temperature anomalies and microscopic defects in the solder joints constitute the associated risk set of the intervention strategy.

[0158] Furthermore, synergy refers to the positive and mutually reinforcing effects of an intervention strategy on multiple forward-looking quality risks within a set of associated risks. For example, an intervention strategy might simultaneously improve reflow oven temperature distribution data and reduce the occurrence of microscopic, hidden defects in solder joints in specific areas; these improvements reinforce each other. Quantifying synergy can involve analyzing the joint changes in risk indicators, for example, by using statistical models or machine learning algorithms to identify and measure the degree of simultaneous improvement of different risk factors under intervention.

[0159] Meanwhile, the offsetting effect refers to the situation where, when an intervention strategy is implemented, it may negatively impact certain forward-looking quality risks within the associated risk set, or introduce new risks, thereby offsetting some of the positive effects. For example, over-adjusting the parameters of a heating component may resolve localized temperature anomalies, but it could lead to uneven temperature distribution in other areas, or even introduce new solder joint defects. Quantifying the offsetting effect can involve predicting and monitoring secondary risks, for example, by analyzing the evolution of secondary risks such as changes in the microstructure of equipment components, material fatigue, and / or microcrack initiation that may result from the intervention strategy.

[0160] Ultimately, by comprehensively considering synergistic and offsetting effects, the overall impact of the intervention strategy can be generated. This result comprehensively reflects the net effect of the intervention strategy, providing a quantitative basis for subsequent intervention strategy selection. For example, the overall impact result can be a weighted score, where synergistic effects contribute positive scores and offsetting effects contribute negative scores, thus obtaining an overall risk mitigation or risk transfer assessment value.

[0161] The proposed approach quantifies the impact of intervention strategies on forward-looking quality risks by breaking them down into synergistic and offsetting effects, thereby enabling a more comprehensive and accurate assessment of the overall effectiveness of intervention strategies.

[0162] Furthermore, in another embodiment of this invention, B32 includes:

[0163] B321: Before implementing the intervention strategy, record the baseline status of temperature distribution data and the occurrence of microscopic hidden defects in the solder joints;

[0164] B322: During the implementation of the intervention strategy, continuously monitor temperature distribution data and the occurrence of microscopic hidden defects in the solder joints;

[0165] B323: Synchronously acquire key parameters of the production environment;

[0166] B324: Compare the changes in baseline status before and after the implementation of the intervention strategy, and combine them with the real-time fluctuations of key parameters in the production environment to identify and deduct the accidental impact of non-intervention factors on temperature distribution data and the occurrence of microscopic hidden defects in solder joints.

[0167] B325: The change in baseline status after removing the effects of chance is taken as the synergistic effect of the intervention strategy on each forward-looking quality risk in the associated risk set.

[0168] Specifically, before implementing the intervention strategy, it is necessary to record the baseline status of key indicators in the LED display manufacturing process. These key indicators mainly include the temperature distribution data of the reflow oven and the occurrence of microscopic hidden defects in the solder joints. Temperature distribution data can be collected in real time by multiple temperature sensors inside the reflow oven, generating detailed temperature distribution maps. The occurrence of microscopic hidden defects in the solder joints can be identified and statistically analyzed using non-destructive testing techniques such as X-ray inspection and ultrasonic testing, combined with image processing and machine learning algorithms. The purpose of recording these baseline statuses is to provide an initial reference point unaffected by the intervention strategy.

[0169] During the implementation of the intervention strategy, the aforementioned temperature distribution data and the occurrence of microscopic hidden defects in the solder joints need to be continuously monitored. Continuous monitoring can be achieved using high-frequency sampling to ensure that the dynamic changes of these key indicators after the implementation of the intervention strategy are captured. For example, data can be collected at regular time intervals (such as every few seconds or minutes) and stored in a database for subsequent trend analysis and effect evaluation.

[0170] Simultaneously, to more accurately assess the actual effectiveness of the intervention strategy, it is also necessary to acquire key parameters of the production environment. These key parameters may include, but are not limited to, ambient temperature, humidity, air pressure, power supply voltage fluctuations, and batch variations of raw materials. Real-time data on these parameters is crucial for distinguishing the effectiveness of the intervention strategy itself from interference from external environmental factors.

[0171] Subsequently, by comparing the changes in baseline conditions before and after the intervention strategy was implemented, and combining this with real-time fluctuations in key parameters of the production environment, the incidental impacts of non-intervention factors on temperature distribution data and the occurrence of microscopic, hidden defects in solder joints can be identified and deducted. For example, if the ambient temperature fluctuates significantly during the implementation of the intervention strategy, and such fluctuations are known to affect the temperature distribution of the reflow oven, then when evaluating the effectiveness of the intervention strategy, this portion of the temperature distribution change caused by the ambient temperature fluctuation needs to be deducted. This can be achieved by establishing statistical or machine learning models between environmental parameters and key indicators, thereby more accurately separating the true impact of the intervention strategy.

[0172] Ultimately, the change in baseline status, excluding the accidental impact of non-intervention factors, is taken as the synergistic effect of the intervention strategy on each forward-looking quality risk in the associated risk set. Synergistic effect refers to the positive or negative cascading impact of the intervention strategy on other related risks when addressing one or more forward-looking quality risks. This quantitative approach allows for a more accurate assessment of the overall effectiveness of the intervention strategy, avoiding misjudgments due to external interference.

[0173] The proposed solution aims to establish a comprehensive data analysis framework by recording and continuously monitoring the baseline status of key indicators before and after the implementation of intervention strategies, and simultaneously acquiring key parameters of the production environment.

[0174] Furthermore, B33 includes:

[0175] B331: Based on the type and intensity of the intervention strategy, predict the secondary risks it will cause, including long-term cumulative secondary risks and / or explosive secondary risks.

[0176] B332: Continuously monitor the microstructural changes of equipment components involved in the intervention strategy to characterize the secondary risk evolution state of material fatigue and / or microcrack initiation;

[0177] B333: Calculate the secondary risk evolution index based on the secondary risk evolution state and compare it with the fourth preset threshold;

[0178] B334: When the secondary risk evolution index reaches the fourth preset threshold, the secondary risk is quantified as an offsetting effect.

[0179] Specifically, before implementing an intervention strategy, it is necessary to predict potential secondary risks based on the type of intervention strategy (e.g., adjusting heating power, changing transmission speed, or scheduling maintenance) and its intensity (e.g., the magnitude of power adjustment, the frequency of maintenance). These secondary risks can be categorized into long-term cumulative secondary risks and / or explosive secondary risks. Long-term cumulative secondary risks refer to negative impacts that gradually emerge over time, such as accelerated aging of equipment components and accumulation of material fatigue; explosive secondary risks refer to negative events that suddenly occur under specific conditions, such as component damage due to overload. The prediction process can be based on historical data, expert experience, physical models, or simulation analysis.

[0180] To accurately assess the evolution of secondary risks, continuous monitoring of the microstructural changes in the equipment components involved in the intervention strategy is necessary. For example, for the heating components of a reflow oven, changes in the resistance of the heating wire, the degree of surface oxidation, and changes in the internal grain structure can be monitored. These microstructural changes are key indicators characterizing material fatigue and / or microcrack initiation, reflecting the component's health status and potential failure risk. Monitoring methods can include non-destructive testing techniques such as ultrasonic testing, eddy current testing, and infrared thermography, or real-time data acquisition through integrated sensors.

[0181] In practical applications, a secondary risk evolution index can be calculated based on the secondary risk evolution status. This index is the result of a quantitative and comprehensive assessment of microstructural changes. For example, a comprehensive score can be calculated by weighting factors such as material fatigue level, microcrack density, and component remaining life. This secondary risk evolution index is then compared with a fourth preset threshold. The fourth preset threshold is a pre-set critical value used to define whether the secondary risk has reached a level that needs to be quantified. This threshold can be set according to the equipment's design life, safety standards, historical failure data, and production quality requirements.

[0182] When the secondary risk evolution index reaches the fourth preset threshold, it indicates that the secondary risk has already significantly or is about to significantly affect the production process or product quality. At this point, the secondary risk is quantified as a mitigation effect. The quantification process can convert the potential losses of the secondary risk (e.g., maintenance costs, downtime, increased product scrap rate, etc.) into a numerical value and incorporate it into the comprehensive impact assessment of the intervention strategy to reflect the potential negative impact of the intervention strategy.

[0183] This application addresses the potential bias in assessing the offsetting effects of intervention strategies by introducing a mechanism for predicting, monitoring, and quantifying secondary risks. Traditional methods may focus only on the direct solution of the current problem while neglecting potential negative impacts. This approach first proactively predicts long-term cumulative and / or explosive secondary risks based on the type and intensity of the intervention strategy, thus incorporating potential negative factors into the initial assessment. Subsequently, by continuously monitoring the microstructural changes of equipment components, the evolution of secondary risks, such as material fatigue and microcrack initiation, can be accurately and in real-time, providing a solid data foundation for quantifying secondary risks. When the secondary risk evolution index reaches a preset fourth threshold, the secondary risk is considered to have reached a level requiring quantification, at which point it is quantified as an offsetting effect. This mechanism ensures that the assessment of the comprehensive impact of the intervention strategy considers not only its positive synergistic effects but also its potential negative offsetting effects, resulting in a more comprehensive and objective comprehensive impact outcome.

[0184] In some preferred embodiments, the following specific example illustrates the situation:

[0185] Suppose that during the LED display manufacturing process, a localized temperature anomaly is identified in a heating component of the reflow oven, increasing the risk of microscopic defects in the solder joints. To maintain the target temperature distribution, the intervention strategy determines to perform zonal process parameter compensation on that heating component, for example, increasing the heating power in that area by 5%.

[0186] Based on the type (power increase) and intensity (5%) of the intervention strategy, the system predicts potential secondary risks. For example, long-term cumulative secondary risks may include accelerated aging of the heating wire inside the heating assembly, increased material fatigue, and stress concentration in local areas of the furnace body; explosive secondary risks may include the risk of the heating wire overheating and melting under extreme operating conditions.

[0187] To monitor the evolution of these secondary risks, the system continuously monitors changes in the microstructure of the heating components. For example, embedded sensors acquire real-time data on the resistance of the heating wire, surface temperature gradient, and vibration spectrum. This data is used to characterize the state of material fatigue and / or microcrack initiation. For instance, a sustained increase in resistance may indicate aging and reduction in cross-section of the heating wire, while changes in the vibration spectrum may predict the initiation of microcracks.

[0188] Based on this monitoring data, a secondary risk evolution index is calculated. For example, a comprehensive index can be defined that combines the resistance change rate, temperature gradient anomaly index, and vibration spectrum characteristic values. Assume the fourth preset threshold for this comprehensive index is 0.8. When the secondary risk evolution index of the heating component rises from 0.5 to 0.85, exceeding the fourth preset threshold, the system quantifies the resulting secondary risk as a mitigation effect. For example, the quantification result might indicate that while power increase resolves the current temperature anomaly, the resulting risk of accelerated component aging is expected to lead to a 10% increase in maintenance costs or a 5-hour increase in downtime over the next three months. This quantified mitigation effect is then incorporated into the comprehensive impact assessment of the intervention strategy, allowing decision-makers to more comprehensively weigh the pros and cons; for example, it may be necessary to simultaneously implement preventative maintenance plans or consider more lenient parameter adjustments.

[0189] In another embodiment of this application, S5000 further includes:

[0190] S5100: Assigns corresponding information based on spatial location to the performance degradation status information, the local temperature anomaly areas identified in the temperature distribution data, and the microscopic hidden defects of the solder joints.

[0191] S5200: When performance degradation status information, local temperature anomaly area, and microscopic hidden defects of solder joints appear consecutively in the same spatial area and meet the preset order, the risk level of that spatial area is increased.

[0192] S5300: Based on the improvement of risk level, the trigger thresholds in the risk assessment rule set are adaptively tightened or relaxed according to spatial region.

[0193] Specifically, mapping performance degradation status information, localized temperature anomalies identified in temperature distribution data, and microscopic hidden defects in solder joints to spatial locations involves associating data from different sources, representing different physical phenomena, with specific physical areas of the LED display substrate or reflow oven through spatial coordinate mapping. For example, by establishing a three-dimensional coordinate system, performance degradation data of heating components, temperature anomalies obtained from surface thermal radiation image analysis, and defect locations in solder joint detection results can be precisely mapped to specific heating sections of the reflow oven or specific locations on the LED display substrate. The aim is to achieve spatial alignment of multi-source heterogeneous data, providing a unified geographical reference for subsequent comprehensive risk assessment.

[0194] Specifically, when performance degradation information, localized temperature anomalies, and microscopic hidden defects in solder joints occur consecutively within the same spatial region and meet a preset sequential relationship, the risk level of that spatial region is increased. Here, "consecutive occurrence" can be understood as these indicators being continuously monitored or repeatedly occurring within the same spatial region over a certain time period. "Preset sequential relationship" refers to the logical order in which these events occur; for example, performance degradation of heating components usually precedes localized temperature anomalies, which in turn may lead to microscopic hidden defects in solder joints. When this causal chain is verified within a specific spatial region, it indicates a high potential risk in that region, thus requiring an increase in its risk level. The aim is to more accurately determine the urgency and severity of risks by identifying the evolution path of risk events.

[0195] In practical applications, based on the improvement in risk level, the trigger thresholds in the risk assessment rule set are adaptively tightened or relaxed according to spatial regions. This means that for a specific spatial region whose risk level has been raised, its corresponding risk assessment trigger threshold will be dynamically adjusted. For example, if the risk level of a region is raised, the risk trigger threshold for that region may be tightened (i.e., a smaller deviation or a lower risk score will trigger intervention) to identify and respond to potential problems earlier. Conversely, if a region is stable in the long term and has a low risk level, its threshold can be appropriately relaxed to avoid over-intervention. The aim is to make the risk assessment system more flexible and adaptable, capable of dynamic optimization based on actual production conditions and regional characteristics.

[0196] The solution proposed in this application spatially correlates performance degradation status information, local temperature anomaly areas, and microscopic hidden defects of solder joints, and further identifies their continuous occurrence and predetermined sequence relationship within the same spatial area, thereby enabling a more refined capture of the evolution process of potential quality risks.

[0197] In some preferred embodiments:

[0198] Suppose that in the third heating zone of the reflow oven, monitoring shows a continuous decrease in heating efficiency due to performance degradation of the heating components. Simultaneously, surface thermal radiation image analysis reveals a persistent localized temperature anomaly at a specific location within this third heating zone, characterized by temperatures below the target value. Furthermore, on the LED display substrate passing through this third heating zone, solder joint detection results show a significant increase in the incidence of microscopic hidden defects (e.g., voids or poor wetting) at the spatial location corresponding to this localized temperature anomaly. When the system identifies that these three events—heating component performance degradation, localized temperature anomalies, and solder joint defects—occur consecutively within the same spatial area of ​​the third heating zone, and their order of occurrence conforms to a preset sequence of "performance degradation → temperature anomaly → solder joint defect," the system will automatically increase the risk level of this third heating zone. Based on this increased risk level, the system will adaptively tighten the trigger threshold for the third heating zone in the risk assessment rule set; for example, adjusting the threshold that originally required a risk index of 0.8 to trigger intervention to 0.6. In this way, even if the risk index of the region only reaches 0.65 later, the system can immediately identify it as a forward-looking quality risk and trigger corresponding intervention strategies, thereby achieving earlier risk warnings and more timely intervention measures.

[0199] Reference Figure 2 The specific embodiments of this application also disclose an LED display manufacturing process parameter optimization system, including:

[0200] Performance degradation status information acquisition module 1 is used to acquire performance degradation status information of the heating components of the reflow oven;

[0201] The local temperature anomaly assessment module 2 is used to reverse-calculate the temperature distribution data of the reflow oven based on the performance degradation status information and the surface thermal radiation image collected when the LED display substrate leaves the reflow oven, and to assess and identify the local temperature anomaly areas of the reflow oven.

[0202] The microscopic hidden defect identification module 3 is used to collect transient temperature response data of the solder joint. When the first deviation between the transient temperature response data and the preset reference response data exceeds the first preset threshold, the solder joint is identified as a microscopic hidden defect.

[0203] The correlation analysis module 4 is used to establish the correlation between local temperature anomaly areas and corresponding heating components and corresponding microscopic hidden defects of solder joints;

[0204] The forward-looking quality risk assessment module 5 is used to determine whether there is a forward-looking quality risk within a first preset time window, based on a preset risk assessment rule set and according to the correlation.

[0205] The intervention strategy scheduling and execution module 6 is used to execute intervention strategies based on the preset parameter compensation mapping relationship when the prospective quality risk reaches the second preset threshold: perform zonal process parameter compensation on the heating components associated with the local temperature anomaly area to maintain the target temperature distribution, and / or schedule maintenance operations to repair or replace the heating components.

[0206] The closed-loop verification module 7 is used to acquire new surface thermal radiation images and updated statistical results of microscopic hidden defects in solder joints after the intervention strategy is completed, to perform closed-loop verification, and to backfill the verification results to update temperature distribution data, performance degradation status information, and risk assessment rule set.

[0207] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing process parameters of an LED display manufacturing process, characterized in that, The method comprises the following steps: acquiring performance degradation state information of a heating assembly of a reflow soldering furnace; based on the performance degradation state information, combining surface thermal radiation images collected when an LED display screen substrate leaves the reflow soldering furnace, performing reverse calculation on temperature distribution data of the reflow soldering furnace, evaluating and identifying a local temperature abnormal area of the reflow soldering furnace; collecting transient temperature response data of a solder joint, and determining that the solder joint is a micro-concealed defect of the solder joint when a first deviation of a comparison between the transient temperature response data and preset reference response data exceeds a first preset threshold; establishing a correlation between the local temperature abnormal area and corresponding heating assemblies and corresponding micro-concealed defects of the solder joint; based on a preset risk evaluation rule set, determining whether there is a forward-looking quality risk according to the correlation within a first preset time window; when the forward-looking quality risk reaches a second preset threshold, performing an intervention strategy according to a preset parameter compensation mapping relationship: performing partition process parameter compensation on the heating assembly associated with the local temperature abnormal area to maintain a target temperature distribution, and / or scheduling a maintenance operation to repair or replace the heating assembly; after the intervention strategy is completed, collecting new surface thermal radiation images and updated statistical results of micro-concealed defects of the solder joint, performing closed-loop verification, and feeding back the verification results for updating the temperature distribution data, the performance degradation state information, and the risk evaluation rule set.

2. The LED display screen manufacturing process parameter optimization method according to claim 1, wherein, The method of collecting transient temperature response data of a solder joint, and determining that the solder joint is a micro-concealed defect of the solder joint when a first deviation of a comparison between the transient temperature response data and preset reference response data exceeds a first preset threshold, comprises the following steps: determining a thermal reference area adjacent to the solder joint on the LED display screen substrate, synchronously exciting the solder joint and the thermal reference area and synchronously collecting transient temperature response data of each of them; comparing the transient temperature response data of the thermal reference area with preset reference response data thereof to obtain an environmental influence factor, and performing environmental calibration on the transient temperature response data of the solder joint based on the environmental influence factor to obtain calibrated transient temperature response data; using a first deviation of a comparison between the calibrated transient temperature response data and the reference response data as a determination basis, and determining that the solder joint is the micro-concealed defect of the solder joint when the first deviation exceeds the first preset threshold.

3. The LED display screen manufacturing process parameter optimization method according to claim 1, wherein, After performing partition process parameter compensation on the heating assembly associated with the local temperature abnormal area to maintain a target temperature distribution, the method comprises the following steps: acquiring real-time data of a production environment as external input; based on a preset parameter compensation mapping relationship, determining a compensation correction amount associated with the local temperature abnormal area in combination with the performance degradation state information and the local temperature abnormal area, and generating updated partition process parameter compensation; monitoring a second deviation between the updated partition process parameter compensation and the temperature distribution data and the occurrence of micro-concealed defects of the solder joint within a second preset time window; When the second deviation continues to exceed the third preset threshold, triggering a secondary calibration, the secondary calibration includes updating the parameter compensation mapping relationship based on real-time data of the current production environment, the performance degradation state information and the temperature distribution data, and outputting new partition process parameter compensation.

4. The LED display screen manufacturing process parameter optimization method according to claim 3, wherein, The triggering of the secondary calibration when the second deviation continues to exceed the third preset threshold comprises: tracking and storing the temperature distribution data and long-term historical data of the occurrence of the solder joint microscopic hidden defects; determining a reference baseline trend reflecting slow and cumulative changes based on the long-term historical data; comparing the second deviation with the reference baseline trend to determine whether the second deviation is consistent with the reference baseline trend; when the second deviation continues to exceed the third preset threshold within a second preset time window and is consistent with the reference baseline trend, adjusting the baseline for second deviation determination and updating the determination rule of the second deviation within the second preset time window; when the second deviation continues to exceed the third preset threshold within a second preset time window and is inconsistent with the reference baseline trend, determining the second deviation as a real deviation caused by poor parameter adjustment effect, and triggering the secondary calibration to re-determine the partition process parameter compensation.

5. The LED display screen manufacturing process parameter optimization method according to claim 3, wherein, After determining whether there is a forward-looking quality risk according to the correlation relationship, the method further comprises: when there are multiple forward-looking quality risks, performing priority sorting on the multiple forward-looking quality risks, and the priority sorting is based on the severity of the risks, the influence range on product quality, and the expected urgency; identifying the correlation between the multiple forward-looking quality risks; evaluating the comprehensive influence of the intervention strategy on the multiple forward-looking quality risks; based on the priority sorting, the correlation and the comprehensive influence, selecting a corresponding intervention strategy.

6. The LED display screen manufacturing process parameter optimization method according to claim 5, wherein, The evaluation of the comprehensive influence of the intervention strategy on the multiple forward-looking quality risks comprises: based on the correlation, determining a corresponding associated risk set for each candidate intervention strategy; quantifying the synergistic effect of the intervention strategy on each forward-looking quality risk in the associated risk set respectively; quantifying the offsetting effect of the intervention strategy on each forward-looking quality risk in the associated risk set respectively; generating a comprehensive influence result of the intervention strategy according to the synergistic effect and the offsetting effect.

7. The LED display screen manufacturing process parameter optimization method according to claim 6, wherein, The quantification of the synergistic effect of the intervention strategy on each forward-looking quality risk in the associated risk set respectively comprises: before the intervention strategy is executed, recording the baseline state of the temperature distribution data and the occurrence of the solder joint microscopic hidden defects; during the execution of the intervention strategy, continuously monitoring the temperature distribution data and the occurrence of the solder joint microscopic hidden defects; synchronously acquiring key parameters of the production environment; comparing the baseline state changes before and after the execution of the intervention strategy, and combining the real-time fluctuations of the key parameters of the production environment, identifying and deducting the accidental influence of non-intervention factors on the temperature distribution data and the occurrence of the solder joint microscopic hidden defects; The baseline state change amount after deducting the accidental influence is taken as a synergistic effect of the intervention strategy on each prospective quality risk in the associated risk set.

8. The LED display screen manufacturing process parameter optimization method according to claim 6, wherein, The respective quantification of the offsetting effect of the intervention strategy on each prospective quality risk in the associated risk set comprises: Based on the type and intensity of the intervention strategy, predict the secondary risks it triggers, including long-term cumulative secondary risks and / or explosive secondary risks; Continuously monitor the microstructure changes of the device components involved in the intervention strategy to characterize the secondary risk evolution state of material fatigue and / or micro-crack initiation; Calculate a secondary risk evolution index based on the secondary risk evolution state and compare it with a fourth preset threshold value; When the secondary risk evolution index reaches the fourth preset threshold value, quantify the secondary risk as the offsetting effect.

9. The LED display screen manufacturing process parameter optimization method according to claim 1, wherein, In the first preset time window, based on the preset risk assessment rule set, whether there is a prospective quality risk is judged according to the association, further comprising: Corresponding the performance degradation state information, the local temperature abnormal area identified in the temperature distribution data and the micro-hidden defect of the welding point according to the spatial position; When the performance degradation state information, the local temperature abnormal area and the micro-hidden defect of the welding point appear continuously in the same spatial region and meet the preset sequence relationship, the risk level of the spatial region is improved; Based on the risk level improvement result, the trigger threshold in the risk assessment rule set is adaptively tightened or relaxed according to the spatial region.

10. An LED display screen manufacturing process parameter optimization system, characterized in that, Comprise: The performance degradation state information acquisition module is used for acquiring the performance degradation state information of the heating assembly of the reflow soldering furnace; The local temperature abnormality evaluation module is used for inversely calculating the temperature distribution data of the reflow soldering furnace according to the performance degradation state information, combining the surface thermal radiation image collected when the LED display screen substrate leaves the reflow soldering furnace, evaluating and identifying the local temperature abnormal area of the reflow soldering furnace; The micro-hidden defect identification module is used for collecting the transient temperature response data of the welding point, and when the first deviation of the transient temperature response data compared with the preset reference response data exceeds the first preset threshold value, the welding point is determined as a micro-hidden defect of the welding point; The correlation analysis module is used for establishing the association between the local temperature abnormal area and the corresponding heating assembly and the corresponding micro-hidden defect of the welding point; The prospective quality risk evaluation module is used for judging whether there is a prospective quality risk according to the association based on the preset risk assessment rule set in the first preset time window; The intervention strategy scheduling execution module is used for executing the intervention strategy according to the preset parameter compensation mapping relationship when the prospective quality risk reaches the second preset threshold value: executing the partition process parameter compensation on the heating assembly associated with the local temperature abnormal area to maintain the target temperature distribution, and / or scheduling the maintenance operation to overhaul or replace the heating assembly. A closed-loop verification module is configured to, after the intervention strategy is completed, acquire a new surface thermal radiation image and updated statistical results of micro-concealed defects of the solder joint, perform closed-loop verification, and backfill the verification results to update the temperature distribution data, the performance degradation state information, and the risk assessment rule set.

Images