A data management system and method for AR glasses production quality traceability

Abstract

The application relates to the technical field of intelligent manufacturing data management, in particular to a data management system and method for AR glasses production quality tracing. The method comprises the following steps: in the active alignment curing process of an AR glasses module, position deviation, torque current and curing process data are synchronously collected at a high frequency; based on the logarithmic coupling relationship between the position deviation and the torque current, tremor energy density representing micro vibration is calculated; based on the inverse weighted relationship between the position deviation change rate and the reciprocal of the remaining curing time, curing drift damping index representing late stability is calculated; the two indexes are fused by using a normalized weight coefficient to obtain a comprehensive quality score, and the module is dynamically classified and subjected to differential process treatment according to the comprehensive quality score. The method evaluates the quality by analyzing the micro stress of the AR glasses module in the active alignment curing process, and the accuracy of quality management tracing is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent manufacturing data management, in particular to a data management system and method for AR glasses production quality traceability. BACKGROUND

[0002] In the field of precision manufacturing of augmented reality (AR) and virtual reality (VR) devices, the assembly precision of the optical-mechanical module is a key factor that determines the optical performance of the final product and the user's visual experience. In order to achieve micron-level assembly precision, six-axis active alignment equipment is widely used in the industry for production. The working process of this type of equipment is usually to use a mechanical hand to hold the optical element, find the best optical position through image algorithm, and once the best coordinate is locked, the ultraviolet lamp is turned on to irradiate the photosensitive glue for curing to fix the position of the optical element.

[0003] The existing quality traceability management system mostly only records the final coordinate position and optical score at the moment of curing completion as result data when dealing with such precise processes. As long as these two values are within the range required by the specification, the system determines that the product is qualified.

[0004] However, this management method ignores the dynamic changes during the curing process and cannot capture the microscopic process of the motor continuously outputting torque to resist the volume shrinkage of the photosensitive glue under ultraviolet light irradiation. Although this microscopic change may not cause displacement to exceed the tolerance at the moment of curing completion, it will store a huge internal stress inside the module.

[0005] These products with high residual internal stress will slowly release internal stress after leaving the factory, causing a slight drift of the optical axis, which will eventually cause dizziness to the user during use. The existing technology lacks deep analysis of curing process data and cannot distinguish between low-stress good products and high-stress hidden danger products, resulting in the flow of such potentially defective products into the market, increasing the rate of after-sales repair.

[0006] Therefore, there is an urgent need for a method that can quantify the dynamic stability during the curing process, realize product life prediction and differentiated process management, and solve the problem of being unable to identify hidden high-stress products in the prior art. SUMMARY

[0007] To solve the problem of the existing quality traceability management system ignoring the dynamic stress changes during the curing process and being unable to identify hidden products with high residual internal stress, the present application provides a data management system and method for AR glasses production quality traceability.

[0008] In a first aspect, the present application provides a data management method for AR glasses production quality traceability, comprising:

[0009] In the active alignment curing process of the AR glasses module, for each sampling time, the position deviation, torque current and curing progress data representing the current curing remaining time of the AR glasses module are synchronously collected, and a position deviation sequence, a torque current sequence and a curing progress sequence are obtained respectively;

[0010] The coupling relationship between the position deviation sequence and the torque current sequence at the same sampling time is determined to calculate the tremor energy density;

[0011] The curing drift damping index is determined by analyzing the change of the position deviation of adjacent sampling times in the position deviation sequence with the curing progress data in the curing progress sequence;

[0012] Based on the tremor energy density and the curing drift damping index, combined with a preset normalized weight coefficient, the comprehensive quality score of the active alignment curing process is determined, and the AR glasses module is dynamically classified according to the comprehensive quality score and automatically pushed to the corresponding differentiated processing process route, so as to realize data management of AR glasses production quality traceability.

[0013] The technical solution proposes a deep quality management system based on process data flow, which is different from the existing technology that only focuses on the static result at the end of curing. The present solution goes deep into the time dimension of the curing process, can sensitively capture the microscopic oscillation when the glue shrinkage and the motor holding force are in fierce confrontation, and makes the invisible internal stress accumulation process explicit. By calculating the curing drift damping index, the small displacement of the glue near solidification in the later curing period is punished, and the stability risk of the product in the whole life cycle is accurately evaluated. The potential high stress hidden danger product is intercepted or specially processed before leaving the factory, which not only effectively reduces the optical axis drift and user dizziness risk caused by stress release, but also improves the accuracy of production quality traceability.

[0014] Preferably, for each sampling time, the method of synchronously collecting the position deviation, torque current and curing progress data representing the current curing remaining time of the AR glasses module is: the position deviation of the AR glasses module is collected by reading the encoder feedback register value and the set target register value of the motion control card, and the position deviation is used to represent the spatial microscopic displacement of the AR glasses module relative to the target position at the current sampling time; the torque current of the motor is read through the bus, and the torque current is used to maintain the electromagnetic torque output by the motor to resist the glue shrinkage when the position of the AR glasses module is maintained; the curing progress data of the AR glasses module is collected by calculating the time difference value between the current sampling time and the preset curing end time, and the curing progress data is used to represent the remaining time of the active alignment curing process.

[0015] Preferably, the tremor energy density is determined based on the following relationship:

[0016]

[0017] In the formula, The vibration energy density during the active alignment and curing process of the AR glasses module reflects the degree of high-frequency jitter during the active alignment and curing process. This represents the total number of sampling moments during the active alignment and curing process. This represents the sequence number of the sampling time. For the first The absolute value of the positional deviation at each sampling time. For the first Torque current at each sampling time, This is the preset base current correction constant. .

[0018] This technical solution constructs a physically meaningful energy density model, which amplifies the influence of micro-displacement by using the square term of the position deviation, and unifies the vibration in the positive and negative directions into energy potential. At the same time, it introduces logarithmically processed torque current as a weighting factor, reflecting the cost that the motor pays to maintain its position. When the motor outputs a large torque and the position still fluctuates, this indicator will rise sharply, thereby accurately identifying high-risk moments when the glue shrinks severely and internal stress remains seriously, providing an accurate basis for quality assessment.

[0019] Preferably, the curing drift damping index is determined based on the following relationship:

[0020]

[0021] in, The curing drift damping index is used to reflect the positional instability during the active alignment and curing process of the AR glasses module. and The first The absolute value of the position deviation at the sampling time and the th sampling time The absolute value of the positional deviation at each sampling time. It is the absolute value symbol. For the first The remaining time of each sampling moment relative to the end time of the active alignment and curing process. The preset time decay smoothing factor is determined based on the curing rate characteristics of the photosensitive adhesive. This represents the total number of sampling times.

[0022] This technical solution is based on the physical law of the change of the curing state of glue with time in polymer chemistry. It constructs a time-reciprocal weighted cumulative model. As the curing process progresses, the remaining time gradually decreases and the denominator becomes smaller. This allows the model to give higher weight to the small displacements in the later stage of curing. This is because the displacement when the glue is close to solid can cause permanent lattice damage or microcracks. This index can effectively distinguish between the normal displacement in the fluid self-healing stage in the early stage of curing and the harmful displacement that causes structural damage in the late stage of curing, thereby accurately predicting the stability of the product.

[0023] Preferably, the basic current correction constant satisfies the following constraint: the sum of the basic current correction constant and the minimum value of the torque current is not less than 1, and the minimum value of the torque current is obtained by statistically analyzing historical production data of multiple AR glasses modules in advance.

[0024] Preferably, the overall quality score of the active alignment curing process is determined based on the following relationship: ;in, The overall quality score for the active alignment curing process. and All are preset normalized weighting coefficients. and These are the vibration energy density and curing drift damping index of the active alignment curing process of the AR glasses module, respectively.

[0025] This technical solution introduces weighting coefficients to uniformly map the vibration energy and drift index of different physical dimensions into score deduction items, thus constructing an intuitive quality score evaluation system. The more intense the vibration during the curing process or the more severe the drift in the later stage, the more points are deducted and the lower the score. This scoring mechanism transforms complex physical process data into quality levels that are easy to understand and implement, facilitating rapid decision-making on the production site and logical judgment of automated systems.

[0026] Preferably, the method for dynamically classifying the AR glasses module based on the comprehensive quality score is as follows: A first threshold, a second threshold, and a third threshold are preset in an increasing order, and first-class, second-class, third-class, and fourth-class products are preset in order of quality from best to worst; if the comprehensive quality score is greater than the first threshold, the AR glasses module is determined to be a first-class product; if the comprehensive quality score is not greater than the first threshold but greater than the second threshold, the AR glasses module is determined to be a second-class product; if the comprehensive quality score is not greater than the second threshold but greater than the third threshold, the AR glasses module is determined to be a third-class product; if the comprehensive quality score is not greater than the third threshold, the AR glasses module is determined to be a fourth-class product.

[0027] Preferably, the automatic pushing to the corresponding differential processing process route is to realize the data management of the AR glasses production quality traceability, including: for the first-class product, the high-temperature aging process is exempted, and the packaging line is directly entered; for the second-class product, the aging process of the standard time length is executed; for the third-class product, the product is marked as the key observation object, and the aging test of the doubled time length is forced to be executed; for the fourth-class product, the direct scrap processing is executed, and the product is not shipped.

[0028] The technical scheme realizes the risk grading management by executing the differential processing process route according to the comprehensive quality score of the AR glasses module, for the ordinary product with a slightly low score but still in the safe domain, the standard process is executed to ensure compliance, and for the high-risk product with obvious shock but not to be scrapped, the aging test of the doubled time length is forced to be executed, and the strategy can eliminate the quality hidden danger in the factory in time, and avoid the product with the implicit defect from flowing into the market to cause complaints.

[0029] Preferably, after the position deviation sequence, the torque current sequence and the curing progress sequence are obtained, the position deviation sequence, the torque current sequence and the curing progress sequence are further smoothed by using a sliding window mean filtering algorithm.

[0030] In a second aspect, the present application also provides a data management system for AR glasses production quality traceability, the system comprising a memory and a processor, the memory storing a computer program, and the processor executing the computer program to realize the steps of the data management method according to any one of the preceding aspects.

[0031] The present application has the following effects:

[0032] The present application makes the invisible internal stress in the production process of the AR glasses module explicit by evaluating the shock energy density and the curing drift damping index in the active alignment curing process of the AR glasses module, and performs differential processing based on dynamic grading, effectively intercepts the implicit defect products that cannot be identified by traditional detection means, and improves the accuracy of production quality traceability. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 is a flowchart of the present application;

[0034] Figure 2 is a real-time monitoring schematic diagram of the active alignment curing process of the AR glasses module of the present application;

[0035] Figure 3 is a product quality screening capability comparison schematic diagram of the present application and the prior art. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application.

[0037] According to the technical scheme of the application, a data management method for AR glasses production quality traceability specifically comprises the following steps:

[0038] S1: Obtain the position deviation sequence, torque current sequence and curing progress sequence in the active alignment curing process of the AR glasses module.

[0039] In the assembly process of the optical-mechanical module of the AR glasses, in order to pursue micron-level optical performance, active alignment equipment must be used for precise assembly. The active alignment curing process refers to the process of turning on the ultraviolet light source to irradiate the photosensitive glue and make it undergo crosslinking polymerization reaction while the mechanical hand clamps the optical element and adjusts it to the best optical position (active alignment stage) and keeps it in a state of not moving (UV curing stage).

[0040] When the photosensitive glue changes from a liquid state to a solid state, it will shrink in volume and generate a huge internal pulling force, trying to pull the optical element away from the optimal position. In order to maintain the absolute position coordinates of the optical element unchanged, the six-axis motor (high-precision servo motor group driving the six-degree-of-freedom adjustment platform) must continuously output the opposite electromagnetic torque through the closed-loop control system to resist the force.

[0041] The existing quality management system often only records the final coordinates after curing, ignoring the implicit details in the curing process, such as microscopic tremors and stress accumulation in the process, which are the root cause of optical axis drift and user dizziness after the product is shipped.

[0042] Therefore, this step aims to establish a complete capture of the dynamic data flow of the curing process, convert the invisible mechanical resistance process into visible digital signals, and provide accurate physical evidence for subsequent quantification of internal stress.

[0043] Specifically, the system first obtains the unique identity code of the current AR glasses module through a code scanning gun or RFID reader. This identity code will serve as the primary key for all subsequent process data, used to establish an index in the database, ensuring that each set of collected position deviation sequence and torque current sequence can be accurately associated with a specific physical product.

[0044] Set the sampling frequency to 1000 Hz (i.e., collect 1000 times per second), and within the entire time window from the start of the UV lamp to the end of curing, the following three sets of time series data are collected and locked simultaneously:

[0045] The first group is a position deviation sequence, which reflects the micro-displacement of the module in space. The specific acquisition method is: at each sampling time, the difference between the encoder feedback register value of the motion control card and the target theoretical position register value of the system lock is obtained, which accurately reflects the degree of deviation of the optical element from the ideal position under the action of the shrinkage force of the glue.

[0046] The second group is a torque current sequence, which directly reflects the motor to maintain the current position. The specific acquisition method is: at each sampling time, the current value of the motor is directly read through the EtherCAT bus, and the greater the current value, the greater the reaction force on the motor, i.e. the stronger the shrinkage force of the glue. EtherCAT is an industrial real-time Ethernet fieldbus protocol. Since EtherCAT has microsecond-level synchronous communication capability, it can bypass the multi-task scheduling delay of the operating system of the host computer and directly access the bottom register of the servo driver, thereby ensuring the strict alignment of the current and position deviation data on the time axis.

[0047] The third group is a curing progress sequence, which records the remaining time from the current sampling time to the end of curing. At each sampling time, the curing progress data of the AR glasses module is collected by calculating the time difference between the current sampling time and the preset curing end time, and the curing progress data is used to represent the remaining time of the active alignment curing process.

[0048] After collecting the original data, due to the complex electromagnetic environment in the industrial field, high-frequency burr noise is often mixed in the original signal. In order to ensure the accuracy of subsequent analysis, the sliding window mean filtering algorithm is used to smooth the original data in this step, and the window size is set to 5 sampling times. For each dimension of data at each sampling time, the average value of the data and the data of the previous and next 2 sampling times is calculated as the data of the sampling time. Through this smoothing operation, the noise interference without physical meaning is eliminated, and the pure data stream which can truly reflect the physical process is obtained.

[0049] S2: Based on the position deviation sequence, the torque current sequence and the curing progress sequence, the tremor energy density and the curing drift damping index of the active alignment curing process of the AR glasses module are determined.

[0050] After obtaining the pure process data stream, the original data alone is not enough to directly judge the product quality, because the large position deviation may be caused by a large external force, or the motor may be insufficient in rigidity, and the large current may be caused by a large shrinkage force of the glue, or the load itself may be heavy.

[0051] In order to accurately evaluate the internal stress of the module in the active alignment curing process, two core features are extracted from the high-frequency data from the perspective of physics: one is an energy index reflecting the intensity of micro-shaking in the process, which is used to reflect the intensity of the confrontation between the motor torque and the glue shrinkage force; the other is a trend index reflecting the position instability in the late curing stage, which is used to evaluate the risk size in the shaping stage. These two indexes can comprehensively reveal the internal health status of the module.

[0052] Specifically, it includes:

[0053] S21: Determine the tremor energy density by analyzing the coupling relationship between the position deviation sequence and the torque current sequence at the same sampling time.

[0054] The tremor energy density is used to evaluate whether there is high-frequency shaking at the micro level in the curing process. The physics logic shows that when the glue shrinkage force is intense and the motor rigidity is insufficient to resist, double shaking of position and current will occur, which is the main source of stress residue.

[0055] In one embodiment, the tremor energy density is determined based on the following relationship:

[0056]

[0057] In the formula, is the tremor energy density of the active alignment curing process of the AR glasses module, which is used to reflect the high-frequency shaking degree in the active alignment curing process. The larger the value is, the more unstable the module is in the active alignment curing process, is the total number of sampling time of the active alignment curing process, is the serial number of the sampling time, is the absolute value of the position deviation of the sampling time (unit: microns). Here, the square is taken to convert the positive and negative shaking deviation into energy potential, and the square operation amplifies the influence of large deviation on the result, so that the tremor energy density is more sensitive to large shaking deviation, is the torque current of the sampling time (unit: ampere), is a preset basic current correction number, The basic current correction number satisfies the following constraint condition: wherein, is a preset basic current correction number, is the minimum value of the torque current obtained by statistically analyzing the historical production data of a plurality of AR glasses modules, is the logarithmic transformation of current value. It should be noted that all parameters in this formula are scalar values when participating in calculation, in order to eliminate dimensional differences and avoid calculation anomalies.

[0058] This formula multiplies the square of the position deviation with the logarithmic current value, constructing a weighted energy model. If the torque current is large, it means that the motor is struggling against the huge shrinkage force of the glue. At this time, if the position deviation is also small, it means that the motor is struggling but holding the position, and the oscillation energy is relatively controllable. However, if the torque current is large and the position deviation is also large, it means that the motor has been difficult to resist the shrinkage of the glue, and the system has fallen into a state of severe oscillation. At this time, the product term will increase sharply, resulting in a significant increase in the final tremor energy density.

[0059] This formula accurately captures the high-stress sampling moment through the coupling calculation of position deviation and torque current, and introduces Because the range of current change is usually large, logarithmic transformation can smooth the numerical difference, and The setting of ensures that the logarithm is always greater than or equal to 1, avoiding mathematical errors and ensuring the physical meaning of the calculation.

[0060] In order to more clearly illustrate the calculation process of tremor energy density, the following gives a simple calculation example:

[0061] Assuming that the sampling frequency is 1000Hz, the total curing time is 5 seconds (the curing time is 5 seconds), and the total number of sampling moments , select the data of the th sampling moment: assume that the position deviation at this moment is microns, the torque current is amperes, and the base current correction constant is .

[0062] The contribution value of this sampling moment in the cumulative summation is calculated as follows:

[0063]

[0064] In order to simplify the demonstration, assume that the contribution values of all other sampling moments in the cumulative summation are also 4, then , if this index increases sharply, it means that the motor current at a certain sampling moment is large, the force is large, and the vibration occurs, resulting in a position deviation.

[0065] S22: determining a curing drift damping index by analyzing the change of the position deviation of adjacent sampling moments in the position deviation sequence with the curing process data in the curing progress sequence.

[0066] The curing drift damping index is used to measure the positional instability in the later stages of curing. According to the principles of polymer chemistry, in the early stages of curing, the glue is in a liquid state, and small positional deviations can be self-healed by the fluid. However, in the later stages of curing, the glue is close to a solid state, and any small positional deviation will cause permanent lattice damage or microcracks. Therefore, the later the curing time, the heavier the penalty for positional deviations should be.

[0067] In one embodiment, the curing drift damping index is determined based on the following relationship:

[0068]

[0069] in, The curing drift damping index is used to reflect the degree of failure of the module's ability to resist positional drift during the curing stage. In other words, it reflects the positional instability of the module during the active alignment and curing process. The larger the value, the higher the risk of displacement in the later stage. and The first The position deviation at the sampling time and the first sampling time Positional deviation at each sampling time It is the absolute value symbol. Reflects the first The instantaneous drift change of the position deviation at each sampling time. For the first The remaining time (in seconds) of the sampling time relative to the end time of the active alignment and curing process, that is, the time remaining in the curing process sequence at the sampling time. The solidification process data collected at each sampling time, over time... It will gradually approach 0. The preset time decay smoothing factor is a constant greater than 0 to prevent the denominator from being 0. N is the total number of sampling times. This represents the sampling time number. It should be noted that all parameters in this formula are used in the calculation as scalar values ​​to eliminate dimensional differences and avoid calculation errors.

[0070] In this relation, This is determined based on the curing rate characteristics of photosensitive adhesives. Since the transition of the adhesive from a liquid state to a gel state and then to a glassy state is not instantaneous, there is a characteristic transition period. The introduction of ...

[0071] This relationship is a time-reciprocal weighted cumulative model, where the numerator is the change in positional deviation and the denominator is the remaining time. In the initial stage of curing... The denominator is relatively large, meaning the positional deviation occurring at this point contributes little to the curing drift damping exponent, consistent with the self-healing physical characteristic of liquid adhesives. As the curing process nears its end... As the denominator becomes very small, it decreases sharply, and the reciprocal increases sharply. At this point, if the same positional deviation occurs, its contribution to the solidification drift damping index will be greatly amplified.

[0072] By using time-reciprocal weighting, a non-linear penalty mechanism is implemented: the later the positional deviation occurs, the more severe the risk the system determines in terms of product quality. The introduction of this indicator not only prevents calculation errors with a denominator of zero, but also adjusts the growth curve of the time weight to better match the physical hardening process of the glue changing from liquid to solid. In this way, the indicator can keenly identify potential products that have undergone minor displacement during the critical curing stage (usually the later stage).

[0073] To more clearly illustrate the calculation process of the curing drift damping index, a simple calculation example is given below:

[0074] Assuming in the first The sampling time and the first sampling time At each sampling time, the same positional deviation occurred, and the displacement... micrometers, setting ;

[0075] If the first The sampling time is in the early stage of the active alignment and solidification process, and there are still [times] remaining until the end. Second, , No. The contribution of each sampling time point to the curing drift damping index of the active alignment curing process is: .

[0076] If the first The sampling time was at the end of the active alignment and solidification process, with [number] days remaining until the end. Second, , No. The contribution of each sampling time point to the curing drift damping index of the active alignment curing process is: .

[0077] It can be seen that for the same amount of displacement, the contribution to the curing drift damping index at the end of curing is 25 times that at the beginning of curing. This is consistent with the physical law that positional deviation at the end of curing will cause permanent lattice damage.

[0078] like Figure 2As shown, the real-time monitoring of the active alignment curing process is shown, the left vertical axis represents the position deviation, the right vertical axis represents the torque current, observing the middle area of the diagram (about 1.5 seconds to 3.5 seconds), the process of the photosensitive glue undergoing a violent shrinkage reaction can be clearly seen, at this time, the position deviation curve presents a sinusoidal high-frequency oscillation, indicating that the module produces micro-displacement under the pulling force of the glue; At the same time, in order to resist this pulling force and try to pull the module back to zero position, the torque current curve synchronously rises sharply, the coupling phenomenon of the large oscillation of the offset position and the high load output of the current coexists, which is the physical root of the sharp increase of the tremor energy density, which reveals the internal stress accumulating in the module, and the static coordinate of the curing end point cannot capture this dynamic resistance process. Observing the right end area of the diagram, as the curing process approaches the end, the remaining time tends to 0, the glue is close to the glass state, at this time, the position deviation curve appears a small one-way deviation trend. Although its absolute value (about 0.05 microns) is much smaller than the oscillation amplitude in the middle, but since it is in the key period of curing forming, according to the curing drift damping index calculation model of the application, this small displacement at the end will be given a high weight penalty, because the small displacement in the glue hardening stage will cause permanent structural damage or optical axis angle risk.

[0079] S3: Based on the tremor energy density and the curing drift damping index, combined with the preset normalized weight coefficient, the comprehensive quality score of the active alignment curing process is determined.

[0080] After obtaining the energy index reflecting the intensity of the process oscillation and the trend index reflecting the stability in the later period, these two indexes can convert the invisible internal stress and potential risk into specific numerical indexes, so as to accurately capture the high stress moment and the end drift risk.

[0081] This step further considers that the physical dimensions and numerical ranges of the two indexes are different, and directly using them for quality judgment is neither intuitive nor convenient for management. In order to realize standardized quality control, this step aims to establish a fusion model to map the two different dimensional physical characteristics into a unified percentage evaluation system, and by introducing a normalized weight coefficient, the complex physical data is converted into a simple and clear quality score, so that each produced module has a quantifiable health degree label, providing a unique decision basis for subsequent automatic grading processing.

[0082] In one embodiment, the comprehensive quality score of the active alignment curing process is determined based on the following relationship:

[0083]

[0084] Wherein, The comprehensive quality score of the active alignment curing process, which is a dimensionless value ranging from 0 to 100, with a full score of 100, and are preset normalization weight coefficients, and their roles include mapping physical quantities of different dimensions to dimensionless deduction values, and are the tremor energy density and the curing drift damping index of the active alignment curing process of the AR glasses module, respectively. It should be noted that and are numerical indicators, that is, they are indicators after de-dimensioning, only the numerical scalar size is retained. It should be noted that all parameters in this relationship are calculated according to their scalar values to eliminate dimensional differences and avoid calculation abnormalities.

[0085] In this relationship, is used to convert the physical value of the tremor energy density into the corresponding deduction value, is used to map the physical value of the curing drift damping index into a deduction value. These two weight coefficients are obtained through historical training. One specific way is: 100 modules are obtained in advance, the tremor energy density and the curing drift damping index of each module are obtained, the contribution of the two indicators to the quality is determined by principal component analysis or linear discriminant analysis, and the normalized contribution is set to and , respectively.

[0086] This relationship adopts a subtraction model, that is, taking perfect 100 points as the starting point, and deducting points according to the existing defects, represents the quality deduction caused by process oscillation. The more intense the oscillation, the greater the deduction, represents the quality deduction caused by late drift. The more serious the late drift, the greater the deduction, reflects the total quality loss. 100 is subtracted to obtain the final comprehensive quality score.

[0087] In the active alignment curing process of the AR glasses module, whether the tremor energy density or the curing drift damping index increases, the total deduction will increase, resulting in a decrease in the final comprehensive quality score, which accurately reflects the decline in product quality. This scoring mechanism is not only intuitive, but also can dynamically adjust and to flexibly adapt to the sensitivity differences of different product models to oscillation or drift.

[0088] To make the calculation process of the solidification drift damping index clearer, a simple calculation example is given below:

[0089] Using the example data in step S2, suppose after calculation, , , the weight coefficient is set , ;

[0090] Quality deduction due to process shock: Quality deduction due to late drift: Final comprehensive quality score: .

[0091] As shown in Figure 3 , the production quality screening ability of the present application and the prior art is shown. The background column chart represents the prior art judgment result. For most samples, the prior art judges them as qualified products without distinction, and the comprehensive quality score is 100. The foreground line chart represents the comprehensive quality score of the present application. Taking sample No. 9 as an example, under the detection system of the prior art, the final position coordinate of the sample at the end of solidification falls within the tolerance range, so it is judged as a qualified product by the system. However, under the monitoring of the present application, the comprehensive quality score of sample No. 9 is only 58, which is marked as a hidden defect. This is because the present application traces back to its whole process data and finds that sample No. 9, although it returns to the origin finally, experiences extremely violent shock resistance in the middle of solidification, and has unstable drift tendency at the end of solidification. These dynamic instability conditions occurring in the active alignment solidification process cause the comprehensive quality score of the AR glasses module to decrease significantly. The prior art will regard this sample as a qualified product, affecting normal use. The present application performs fault interception in advance, improving the accuracy of quality management.

[0092] In this way, through the comprehensive quality score model, complex multi-dimensional physical characteristics are converted into intuitive single scores, which can accurately reflect the true level of product quality, especially revealing the hidden quality decline that the prior art cannot detect.

[0093] S4: According to the comprehensive quality score, dynamically classify and automatically push to the corresponding differentiated processing process route to realize data management of AR glasses production quality traceability.

[0094] After obtaining the comprehensive quality score of each AR glasses module, the traditional method is often simply to determine whether it is qualified or not. However, this binary determination method ignores the continuous distribution characteristics of product quality. Although products with scores of 99 and 61 are both qualified products, their internal stress states and potential lifespans are completely different. If the same post-processing process is used for all qualified products, such as uniform aging test for a certain period of time, it is easy to cause waste of time resources for high-quality products and improper handling for potential high-risk products.

[0095] Therefore, the present step aims to implement a dynamic and refined grading management strategy. According to the high or low of the comprehensive quality score, the products are divided into different grades, and the most reasonable processing method is matched for each grade, so as to maximize production efficiency while ensuring the quality of shipment.

[0096] Specifically, it includes:

[0097] The first threshold 90, the second threshold 75 and the third threshold 60 are preset in order of increasing, and the first-class product, the second-class product, the third-class product and the fourth-class product are preset in order of quality from good to bad, and are respectively labeled as excellent, ordinary, high risk and scrap.

[0098] The system labels each produced AR glasses module according to the comprehensive quality score and automatically pushes it to the MES system to realize quality data management, which specifically includes:

[0099] If , the label is excellent, indicating that the module is extremely stable during the curing process and has extremely low internal stress, so the high-temperature aging process of the module is exempted, and it directly flows into the packaging line, reducing unnecessary production cycle and reducing inventory cost.

[0100] If , the label is ordinary, indicating that the module has slight oscillation but is within the safety domain, and the standard aging process (e.g. 12 hours) is performed, which is to ensure its performance stability and meet the regular shipment standard.

[0101] If , the label is high risk, indicating that the optical index of the module is qualified, but the process oscillation is severe, and the doubled aging test (e.g. 48 hours cold and hot impact) is forced to perform, which is to force the potential stress to release in advance through harsh environment. If the parameters drift after aging, it will be intercepted, and if it is still stable, it will be downgraded for shipment or used for other purposes.

[0102] If , the label is scrap, representing that the process is severely unstable, and it is directly scrapped, even if the current optical test is passed, it will not be shipped to avoid potential hidden dangers in the market.

[0103] According to the calculation result of S3, the comprehensive quality score of the AR glasses module in the active alignment curing process is 66, which is labeled as high-risk product, so the system will automatically route it to the double aging test process, instead of directly shipping or only doing standard aging as in the traditional way.

[0104] In this way, through dynamic grading management, both the production efficiency is improved by quickly releasing high-quality products, and the after-sales repair rate is effectively reduced by strictly intercepting high-risk products, realizing the dual optimization of quality and efficiency.

[0105] Finally, after determining the comprehensive quality score and completing the grading, the system also performs data archiving operation: the unique identity code of the AR glasses module, the calculated tremor energy density, the curing drift damping index, the comprehensive quality score, the determined grade label, and the original position deviation sequence and torque current sequence are compressed and packaged to generate the data history of the curing process of the AR glasses module, the data history is uploaded to the MES (Manufacturing Execution System) of the factory through industrial Ethernet and is persistently stored, so as to be used for subsequent production quality traceability.

[0106] The application also provides a data management system for AR glasses production quality traceability, which comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the steps of a data management method for AR glasses production quality traceability, so as to realize the production quality data management of AR glasses.

[0107] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A data management method for AR glasses production quality traceability, characterized by, Comprise: In the active alignment curing process of the AR glasses module, for each sampling time, the position deviation, torque current and curing progress data representing the current curing remaining time of the AR glasses module are synchronously collected to obtain the position deviation sequence, torque current sequence and curing progress sequence respectively; The coupling relationship between the position deviation sequence and the torque current sequence at the same sampling time is determined to calculate the tremor energy density, which satisfies: ； The vibration energy density during the active alignment and curing process of the AR glasses module reflects the degree of high-frequency jitter during the active alignment and curing process. This represents the total number of sampling moments during the active alignment and curing process. This represents the sequence number of the sampling time. For the first The absolute value of the positional deviation at each sampling time. For the first Torque current at each sampling time, This is the preset base current correction constant. ; By analyzing the change of the position deviation of adjacent sampling time in the position deviation sequence with the curing progress data in the curing progress sequence, the curing drift damping index is determined, which satisfies: ， The curing drift damping index is used for the active alignment and curing process of AR glasses modules to reflect the positional instability during the active alignment and curing process. and The first The absolute value of the position deviation at the sampling time and the th sampling time The absolute value of the positional deviation at each sampling time. It is the absolute value symbol. For the first The remaining time at each sampling moment relative to the end of the active alignment and curing process. The preset time decay smoothing factor is determined based on the curing rate characteristics of the photosensitive adhesive; Based on the tremor energy density and the curing drift damping index, combined with the preset normalized weight coefficient, the comprehensive quality score of the active alignment curing process is determined, which satisfies: ; to evaluate the overall quality of the active alignment curing process, and are preset normalization weight coefficients; According to the comprehensive quality score, the AR glasses module is dynamically classified and automatically pushed to the corresponding differentiated processing process route to realize data management of AR glasses production quality traceability.

2. The data management method according to claim 1, characterized by, For each sampling time, the method for synchronously collecting the position deviation, torque current and curing progress data representing the current curing remaining time of the AR glasses module is: the position deviation of the AR glasses module is collected by reading the encoder feedback register value and the set target register value, and the position deviation is used to represent the spatial micro-displacement of the AR glasses module relative to the target position at the current sampling time; the torque current of the motor is read through the bus, and the torque current is used to maintain the electromagnetic torque output by the motor to resist the shrinkage of the glue; the curing progress data of the AR glasses module is collected by calculating the time difference between the current sampling time and the preset curing end time, and the curing progress data is used to represent the remaining time of the active alignment curing process.

3. The data management method of claim 1, wherein, The base current correction constant satisfies the following constraint condition: the sum of the base current correction constant and the minimum value of the torque current is not less than 1, and the minimum value of the torque current is obtained by pre-acquiring historical production data of multiple AR glasses modules.

4. The data management method of claim 1, wherein, The method for dynamically classifying the AR glasses module according to the comprehensive quality score is: The first threshold value, the second threshold value and the third threshold value are preset in the order of increasing, and the first-grade product, the second-grade product, the third-grade product and the fourth-grade product are preset in the order of quality from good to bad; If the comprehensive quality score is greater than the first threshold value, the AR glasses module is determined to be a first-grade product; if the comprehensive quality score is not greater than the first threshold value and greater than the second threshold value, the AR glasses module is determined to be a second-grade product; if the comprehensive quality score is not greater than the second threshold value and greater than the third threshold value, the AR glasses module is determined to be a third-grade product; if the comprehensive quality score is not greater than the third threshold value, the AR glasses module is determined to be a fourth-grade product.

5. The data management method according to claim 4, characterized by, Automatically pushing to the corresponding differentiated processing process route to realize data management of AR glasses production quality traceability, comprising: For the first-grade product, the high-temperature aging process is exempted, and it directly enters the packaging line; For the second-grade product, the aging process of standard length is executed; For the third grade, marked as the key observation object, the aging test with doubled time length is enforced; For the fourth grade, the direct scrapping treatment is performed, and the product is not shipped.

6. The data management method of claim 1, wherein, After the position deviation sequence, the torque current sequence and the curing progress sequence are obtained, the position deviation sequence, the torque current sequence and the curing progress sequence are further smoothed by using a sliding window mean filtering algorithm.

7. A data management system for AR glasses production quality traceability, characterized by, The system comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the steps of the data management method in any one of claims 1-6.

Images