Method and device for quickly estimating service life of LED chip and medium

Abstract

The invention discloses an LED chip service life rapid estimation method and device and a medium, and relates to the technical field of electrical digital data processing, and the method comprises the steps: designing a current pulse stress sequence for a calibration sample based on photoelectric thermal characteristic data, synchronously collecting photoelectric thermal response data before and after a pulse, and carrying out verification completion and mapping storage; generating a photoelectric thermal pulse response data set; according to the photoelectric thermal pulse response data set, photoelectric thermal change characteristics of the LED chip sensitive to the service life are extracted, and a photoelectric thermal pulse service life fingerprint of the LED chip is constructed; constructing a service life rapid estimation model, and predicting the service life of the LED chip according to the photoelectric thermal pulse service life fingerprint and the working condition information to generate the predicted service life of the LED chip; according to the method, the photoelectric thermal pulse life fingerprint of the LED chip is constructed, the photoelectric thermal evolution process under the pulse stress is compressed into a unified life characterization carrier, and a degradation feature basis capable of being directly called is provided for life modeling.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, and in particular to a method, device and medium for rapid estimation of LED chip lifespan. Background Technology

[0002] As a key component in solid-state lighting and display devices, the lifespan and reliability of LED chips directly affect the overall lighting flux maintenance and upkeep costs. Current lifespan assessments typically employ accelerated aging tests with constant current and temperature or increasing temperature and current. Under rated or stringent operating conditions, the luminous flux, color coordinates, electrical parameters, and junction temperature of the samples are continuously monitored as photoelectric and thermal characteristics. Combined with pre-defined failure criteria and statistical lifespan models, the lifespan distribution and reliability level of LED chips are inferred for device screening and product design. This is a commonly used reliability evaluation approach in the industry.

[0003] However, conventional accelerated aging and lifetime extrapolation methods mainly rely on the luminous flux decay curve under constant stress, making it difficult to fully utilize the photoelectric and thermal characteristics in short-term stress response to quickly estimate the lifetime of LED chips in the early degradation behavior. On the other hand, there is limited joint modeling of historical aging data and in-service fault records, and the model lacks systematic calibration and closed-loop correction between actual application conditions. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for rapid estimation of LED chip lifespan, which solves the problems of existing technologies that make it difficult to quickly estimate the lifespan of LED chips by utilizing the photoelectric and thermal response characteristics under short-term stress, and the limited joint modeling of historical aging data and in-service fault records.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for rapid estimation of LED chip lifespan, which includes collecting photoelectric and thermal characteristic data of LED chips under rated conditions, and verifying and screening them in combination with failure criteria to generate a set of calibration samples of LED chips. Based on photoelectric and thermal characteristic data, a current pulse stress sequence was designed for calibrating the sample. Photoelectric and thermal response data were collected synchronously before and after the pulse, and then verified, completed, mapped, and stored to generate a photoelectric and thermal pulse response dataset. Based on the photoelectric thermal pulse response dataset, the photoelectric thermal change characteristics of LED chips that are sensitive to lifetime are extracted, and the photoelectric thermal pulse lifetime fingerprint of LED chips is constructed. A rapid lifetime estimation model is constructed, and the lifetime of LED chips is predicted based on photoelectric thermal pulse lifetime fingerprint and operating condition information, generating the predicted lifetime of LED chips. Based on the photoelectric thermal pulse lifetime fingerprint, LED chips are clustered and representative samples of each type are extracted for accelerated aging verification. The deviation between the predicted lifetime and the measured lifetime is obtained, and the lifetime rapid estimation model is optimized.

[0007] As a preferred embodiment of the rapid LED chip lifetime estimation method of the present invention, the step of combining failure criteria for verification and screening to generate a set of calibrated LED chips includes the following steps: Each LED chip in the batch to be tested is subjected to photoelectric and thermal joint testing under rated conditions. Photoelectric and thermal characteristic data of the LED chips are collected and standardized, and a set of LED chip test data is generated. Based on historical aging data and in-service fault records, multivariate logistic regression analysis was used to statistically fit the relationship between different luminous flux maintenance rates, color coordinate offsets, upper limits of electrical parameters, and actual failure probabilities, generating failure criteria for LED chips. By combining LED chip test data sets and failure criteria, the photoelectric and thermal characteristic data of each LED chip are verified and screened, and a set of calibration samples for LED chips is generated.

[0008] As a preferred embodiment of the rapid LED chip lifetime estimation method of the present invention, the step of calibrating the sample design current pulse stress sequence includes the following steps. Based on the difference between the photoelectric and thermal characteristic data of each LED chip in the calibration sample set and the failure criteria, the allowable space of each LED chip from the failure boundary is estimated. The allowable space is statistically analyzed, and the low quantile value is selected as the safety margin of the stress level that the calibration sample can withstand. The safety margin is then converted into the target degradation excitation intensity according to the degradation excitation conversion factor. Under the constraints of safety margin and target degradation excitation intensity, the pulse width and pulse interval of the current amplitude are selected to generate a current pulse stress sequence.

[0009] As a preferred embodiment of the LED chip lifetime rapid estimation method of the present invention, the steps of performing verification completion and mapping storage to generate a photoelectric thermal pulse response dataset are as follows: Under controlled ambient temperature, current pulses are applied sequentially to each LED chip according to the current pulse stress sequence. Photoelectric thermal response data are collected synchronously before and after each current pulse and collected according to the LED chip number and the order of current pulses to form a set of original photoelectric thermal response data. The original photoelectric thermal response dataset is processed by filling in missing values, unifying the format, and standardizing the data to generate a photoelectric thermal pulse response dataset.

[0010] As a preferred embodiment of the LED chip lifetime rapid estimation method of the present invention, the steps of extracting the lifetime-sensitive photoelectric and thermal change characteristics of the LED chip and constructing the photoelectric and thermal pulse lifetime fingerprint of the LED chip are as follows: Based on the photoelectric and thermal pulse response dataset, the photoelectric and thermal characteristics of each LED chip before and after each current pulse relative to the previous moment and the rate of change per unit time are calculated to obtain a set of photoelectric and thermal change trajectories. Based on the set of photoelectric and thermal change trajectories and the correspondence between photoelectric and thermal changes and lifetime performance in historical aging data, we analyze the ability of different photoelectric and thermal characteristic changes and change rates to distinguish the degree of degradation, screen out the photoelectric and thermal change amounts and change rates that can distinguish different lifetime levels, and form a set of photoelectric and thermal change characteristics. Based on the set of photoelectric and thermal change characteristics, and in accordance with the preset characteristic order and fixed length, the photoelectric and thermal change amount and rate of change of each LED chip under each current pulse are spliced ​​together to construct the photoelectric and thermal pulse lifetime fingerprint of the LED chip.

[0011] As a preferred embodiment of the LED chip lifetime rapid estimation method of the present invention, the steps of constructing a rapid lifetime estimation model and predicting the lifetime of the LED chip based on photoelectric thermal pulse lifetime fingerprint and operating condition information to generate the predicted lifetime of the LED chip are as follows. A fast lifetime estimation model is constructed based on an input fusion layer, a physical index extraction layer, a feature representation layer, and a lifetime regression layer. The input fusion layer concatenates the photoelectric thermal pulse lifetime fingerprint of the LED chip with the operating condition information in a fixed order and performs standardization processing to generate a unified input feature vector. The physical index extraction layer extracts photoelectric and thermal response rate information (the rate of change of luminous flux per unit time, the increase of reverse leakage current, and the rate of junction temperature rise) according to the degradation rate rule, and transforms them into degradation rate index and thermal stress index through principal component analysis. The feature representation layer fuses the unified input feature vector with the degradation rate index and thermal stress index, and compresses and recombines them through nonlinear transformation to obtain the lifetime representation vector; The lifetime regression layer estimates the predicted lifetime of the LED chip based on the lifetime representation vector and uses gradient boosting regression tree. During the training phase, the error between the predicted lifetime and the measured lifetime is used to optimize the feature representation layer and the input fusion layer, thus forming a complete fast lifetime estimation model. By inputting photoelectric thermal pulse lifetime fingerprints and operating condition information into a rapid lifetime estimation model, the predicted lifetime of the LED chip can be obtained.

[0012] As a preferred embodiment of the LED chip lifetime rapid estimation method of the present invention, the steps of performing cluster analysis on the LED chip based on photoelectric thermal pulse lifetime fingerprint and extracting representative samples of each category for accelerated aging verification, obtaining the deviation between the predicted lifetime and the measured lifetime, and optimizing the lifetime rapid estimation model are as follows. Based on the photoelectric and thermal pulse lifetime fingerprint of LED chips, cluster analysis is performed on LED chips, and the photoelectric and thermal pulse lifetime fingerprints are divided into several fingerprint cluster categories according to the degree of similarity, generating lifetime fingerprint clustering results. Based on the lifetime fingerprint clustering results, several LED chips are extracted from the LED chips corresponding to each type of photoelectric thermal pulse lifetime fingerprint as accelerated aging verification objects to generate an accelerated aging verification sample set. The LED chips in the accelerated aging verification sample set were continuously operated under accelerated aging conditions until the failure criteria were met. The measured lifespan of each LED chip was recorded, and the measured lifespan was compared with the predicted lifespan of the corresponding LED chip to generate a lifespan deviation data set. Based on the lifetime deviation data set, the parameters of the lifetime regression layer and the photoelectric and thermal change characteristics in the feature characterization layer of the rapid lifetime estimation model are optimized and updated.

[0013] As a preferred embodiment of the LED chip lifetime rapid estimation method of the present invention, the photoelectric and thermal characteristic data include luminous flux, color coordinates, peak wavelength, forward voltage, reverse leakage current, dynamic on-resistance, estimated junction temperature, junction-to-shell thermal resistance, and temperature rise amplitude under pulse current.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the LED chip lifetime rapid estimation method as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the LED chip lifetime rapid estimation method as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By constructing the photoelectric and thermal pulse lifetime fingerprint of LED chips, the photoelectric and thermal evolution process under pulse stress is compressed into a unified lifetime characterization carrier, providing a directly callable degradation feature basis for lifetime modeling; by constructing a rapid lifetime estimation model and generating the predicted lifetime of LED chips based on the photoelectric and thermal pulse lifetime fingerprint and operating condition information, a processing link is realized that the lifetime characterization vector and lifetime result can be directly output from the short-term photoelectric and thermal test results under given operating conditions, which facilitates the rapid acquisition of LED chip lifetime information in production screening and reliability assessment. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart for a method to quickly estimate the lifespan of LED chips.

[0019] Figure 2 A flowchart for obtaining the lifetime deviation dataset.

[0020] Figure 3 A flowchart for generating a calibration sample set.

[0021] Figure 4 A flowchart for constructing photoelectric thermal pulse lifetime fingerprints. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4This is one embodiment of the present invention, which provides a method for rapidly estimating the lifetime of an LED chip, comprising the following steps: S1. Collect photoelectric and thermal characteristic data of LED chips under rated conditions, and verify and screen them in combination with failure criteria to generate a set of calibration samples of LED chips.

[0026] Each LED chip in the batch to be tested is subjected to a combined photoelectric and thermal test under rated conditions. The photoelectric and thermal characteristic data of the LED chips are collected and standardized, and then summarized to generate a set of LED chip test data.

[0027] Photoelectric and thermal characteristic data include luminous flux, color coordinates, peak wavelength, forward voltage, reverse leakage current, dynamic on-resistance, estimated junction temperature, junction-to-shell thermal resistance, and temperature rise under pulsed current.

[0028] Based on historical aging data and in-service fault records, multivariate logistic regression analysis was used to statistically fit the relationship between different luminous flux maintenance rates, color coordinate offsets, upper limits of electrical parameters, and actual failure probabilities, generating failure criteria for LED chips.

[0029] Furthermore, based on historical aging data and in-service fault records, the historical aging data and in-service fault records are organized according to LED chip number, operating conditions, and time axis. For each LED chip lifespan record, the luminous flux maintenance rate, color coordinate offset, and upper limit of electrical parameters are calculated, and an actual failure marker indicating whether failure has occurred is given, forming a sample set containing luminous flux maintenance rate, color coordinate offset, upper limit of electrical parameters, and actual failure marker. Based on the sample set, using luminous flux maintenance rate, color coordinate offset, and upper limit of electrical parameters as independent variables and actual failure marker as dependent variable, multivariate logistic regression analysis is used to calculate the actual failure probability under different combinations of luminous flux maintenance rate, color coordinate offset, and upper limit of electrical parameters. Based on the preset failure probability level (based on the maximum allowable failure rate in the target application, combined with a value of actual failure probability directly selected by relevant reliability standards or design specifications), the performance boundary parameters corresponding to luminous flux maintenance rate, color coordinate offset, and upper limit of electrical parameters are calculated inversely. The performance boundary parameters are summarized into a set of performance boundary parameters used to determine whether the LED chip has entered a failure state, generating the failure criteria for the LED chip.

[0030] It should be noted that historical aging data refers to complete test process data for LED chips from previous production batches, continuously recorded under controlled accelerated stress or long-term rated operating conditions, including changes in luminous flux, electrical parameters, junction temperature, and the time when failure criteria were reached. In-service fault records refer to on-site statistical information on LED chip failure events in actual application scenarios, such as abnormal brightness decay, excessive color deviation, intermittent lighting, or complete failure. This includes data such as operating conditions, usage duration, description of fault phenomena, and confirmation conclusions for returned samples.

[0031] Failure criteria refer to a set of performance boundary parameters used to determine whether an LED chip has entered a failure state, that is, the upper limit of each data in the photoelectric and thermal characteristic data.

[0032] Rated conditions refer to the normal operating state formed by the combination of rated operating current, rated environment or junction temperature, and corresponding driving voltage as specified in the LED chip product standard. The LED chip test data set includes photoelectric and thermal characteristic data of each LED chip in the batch under test collected under rated conditions, as well as additional information such as the corresponding chip number, test time, and test conditions.

[0033] By combining the LED chip test data set and failure criteria, the photoelectric and thermal characteristic data of each LED chip are verified and screened. LED chips that do not meet the failure criteria are eliminated, LED chips that meet the failure criteria are screened, and a calibration sample set of LED chips is generated.

[0034] Furthermore, combining the LED chip test data set and failure criteria, the corresponding photoelectric and thermal characteristic data of each LED chip in the LED chip test data set are read one by one. The photoelectric and thermal characteristic data are compared with each performance boundary parameter in the failure criteria. Based on the comparison results, a judgment mark is given for whether each LED chip meets the failure criteria. After obtaining the judgment mark, the LED chips marked as not meeting the failure criteria are removed from the LED chip test data set, and only the LED chips marked as meeting the failure criteria are retained. The LED chips that meet the failure criteria and their corresponding photoelectric and thermal characteristic data are re-summarized to form a calibration sample set of LED chips for subsequent current pulse stress loading and lifetime analysis.

[0035] S2. Based on photoelectric and thermal characteristic data, a current pulse stress sequence is designed for the calibration sample. Photoelectric and thermal response data are collected synchronously before and after the pulse, and then verified, completed, mapped, and stored to generate a photoelectric and thermal pulse response dataset.

[0036] Based on the difference between the photoelectric and thermal characteristic data of each LED chip in the calibration sample set and the failure criteria, the allowable space of each LED chip from the failure boundary is estimated. Furthermore, based on the calibration sample set of LED chips, the photoelectric and thermal characteristic data of each LED chip in the calibration sample set are read item by item, and the difference between the photoelectric and thermal characteristic data and the corresponding performance boundary parameters in the failure criterion is calculated to obtain the remaining margin of each photoelectric and thermal characteristic data from the failure criterion boundary. After obtaining the remaining margin of each photoelectric and thermal characteristic data, the photoelectric and thermal characteristic data of the same LED chip are comprehensively evaluated using the minimum remaining margin as a constraint. The corresponding minimum remaining margin is recorded as the allowable space of the LED chip from the failure boundary, and the allowable spaces of all LED chips from the failure boundary are organized into an allowable space set for subsequent determination of safety margin.

[0037] The allowable space is statistically analyzed, and the low quantile value is selected as the safety margin of the stress level that the calibration sample can withstand. The safety margin is then converted into the target degradation excitation intensity according to the degradation excitation conversion factor.

[0038] Furthermore, the allowable space set is sorted according to its numerical value, and the allowable space value at the selected percentile position is calculated in the sorted allowable space set. The allowable space value at the selected percentile position in the sorted allowable space set is determined as the low quantile value of the allowable space. After obtaining the low quantile value of the allowable space, it is used as the safety margin of the stress level that the calibration sample can withstand. According to the correspondence between the degradation excitation conversion factor and the safety margin of the stress level that the calibration sample can withstand, the safety margin of the stress level that the calibration sample can withstand is converted into the target degradation excitation strength. This provides a basis for selecting the current amplitude pulse width and pulse interval under the constraints of the safety margin and the target degradation excitation strength.

[0039] It should be noted that the low percentile value refers to the percentile position of each LED chip in terms of the allowable space from the failure boundary, sorted from smallest to largest. For example, the value in the smallest 10% of all allowable spaces. The degradation excitation reduction factor is determined by applying current pulses with different reduction ratios sequentially to a small number of calibration samples of LED chips and comparing whether the decrease in luminous flux at each reduction ratio falls within the degradation range (e.g., a decrease in luminous flux relative to the initial value of 5% to 15%) and does not trigger the failure criterion. The reduction ratio is determined by selecting a reduction ratio that simultaneously meets the degradation range requirement and the safety requirement.

[0040] Under the constraints of safety margin and target degradation excitation intensity, the pulse width and pulse interval of the current amplitude are selected so that the current pulse will not exceed the safety margin of the LED chip and can cause measurable photoelectric and thermal changes, thereby generating a current pulse stress sequence.

[0041] Furthermore, after determining the safety margin and the target degradation excitation intensity, the target degradation excitation intensity is used as the upper limit of the acceptable energy input of the LED chip under a single current pulse. Based on the safety margin, candidate combinations of current amplitude, pulse width, and pulse interval are calculated. For each combination of current amplitude, pulse width, and pulse interval, the corresponding current pulse energy is calculated, and it is verified that the current pulse energy does not exceed the safety margin and can cause measurable changes in photoelectric and thermal characteristics such as changes in luminous flux, reverse leakage current, and junction temperature based on empirical sensitivity thresholds (for example, the decrease in luminous flux relative to the initial value after a single current pulse exceeds the measurement noise level, the reverse leakage current shows a continuous upward trend compared to before the pulse and is higher than the instrument noise band, and the junction temperature change exceeds the minimum resolvable temperature difference limited by the temperature measurement resolution). The combinations of current amplitude, pulse width, and pulse interval that meet the safety margin constraint and the measurable change constraint are arranged in a preset order, and the current pulse stress sequence used for subsequent application of current pulses and acquisition of photoelectric and thermal response data is obtained by continuous arrangement.

[0042] It should be noted that the empirical sensitivity threshold refers to the minimum relative change that can be stably detected under the conditions of photoelectric and thermal characteristic measurement noise and instrument resolution. Generally, the range of empirical sensitivity threshold values ​​can be determined through preliminary experiments, with a change in luminous flux relative to the initial value of about 1% to 5% and a junction temperature change on the order of several degrees Celsius. The preset sequence is based on the principle of increasing current amplitude from low to high, increasing pulse width from short to long, ensuring that the pulse interval meets the heat dissipation recovery requirements, and gradually increasing stress to facilitate tracking degradation behavior.

[0043] Under controlled ambient temperature, current pulses are applied sequentially to each LED chip according to the current pulse stress sequence. Photoelectric and thermal response data are collected synchronously before and after each current pulse and collected according to the LED chip number and the order of current pulses to form a raw set of photoelectric and thermal response data.

[0044] Furthermore, under controlled ambient temperature, the calibration sample set of LED chips is fixed one by one on the test fixture. Current pulses are applied to each LED chip sequentially according to the pre-determined current amplitude, pulse width, and pulse interval in the current pulse stress sequence. Before the start of each current pulse, the photoelectric thermal response data of the LED chip is collected. Immediately after the end of the corresponding current pulse, the same type of photoelectric thermal response data as before the start of the current pulse is collected again. Using the photoelectric thermal response data before and after each current pulse, a single-pulse photoelectric thermal response record with LED chip number, current pulse sequence, and time stamp is generated. All single-pulse photoelectric thermal response records are collected and stored according to LED chip number and current pulse sequence to form a set of original photoelectric thermal response data for subsequent missing value completion, format unification, and standardization processing.

[0045] The original photoelectric and thermal response dataset is processed by filling in missing values, unifying the format, and standardizing the data to generate a photoelectric and thermal pulse response dataset that describes the degradation behavior of a set of calibration samples of LED chips under pulse conditions.

[0046] Furthermore, the original photoelectric thermal response data set was checked for completeness according to LED chip number, current pulse order, and time stamp. Missing photoelectric thermal response data in the original data set was filled in by interpolation using records of the same type under adjacent current pulses. Based on the completion, the field names and units of measurement of the photoelectric thermal response data were standardized, and the data was standardized according to the overall mean and standard deviation. The standardized photoelectric thermal response data was then collected according to LED chip number and current pulse order to generate a photoelectric thermal pulse response dataset for describing the degradation behavior of the calibration sample set of LED chips under pulse conditions.

[0047] S3. Based on the photoelectric thermal pulse response dataset, extract the photoelectric thermal change characteristics of the LED chip that are sensitive to its lifetime, and construct the photoelectric thermal pulse lifetime fingerprint of the LED chip.

[0048] Based on the photoelectric and thermal pulse response dataset, the photoelectric and thermal characteristics of each LED chip before and after each current pulse relative to the previous moment and the rate of change per unit time are calculated, resulting in a set of photoelectric and thermal change trajectories organized according to the LED chip number and the current pulse sequence.

[0049] Furthermore, based on the photoelectric thermal pulse response dataset, the photoelectric thermal characteristic data of each LED chip in the calibration sample set of LED chips are read in pairs at the start and end times of each current pulse. The change in photoelectric thermal characteristics at the end time of each current pulse relative to the previous time is calculated, and the change rate per unit time is obtained by dividing the change in photoelectric thermal characteristics by the corresponding time interval. The change in photoelectric thermal characteristics and the change rate per unit time of each LED chip under all current pulses are arranged in order of LED chip number and current pulse sequence to generate a set of photoelectric thermal change trajectories that continuously record the photoelectric thermal evolution process under each current pulse. This provides input for subsequent analysis of the ability of different changes in photoelectric thermal characteristics and change rates to distinguish the degree of degradation based on the set of photoelectric thermal change trajectories.

[0050] Based on the set of photoelectric and thermal change trajectories and the correspondence between photoelectric and thermal changes and lifetime performance in historical aging data, we analyze the ability of different photoelectric and thermal characteristic changes and change rates to distinguish the degree of degradation, and screen out the photoelectric and thermal change changes and change rates that can distinguish different lifetime levels to form a set of photoelectric and thermal change characteristics.

[0051] Furthermore, based on the set of photoelectric and thermal change trajectories and the correspondence between photoelectric and thermal changes and lifetime performance in historical aging data, the correlation index and group statistics of the change amount and rate of each type of photoelectric and thermal characteristic in the set of photoelectric and thermal change trajectories with lifetime performance are calculated. By comparing the differences in the numerical distribution of the change amount and rate of photoelectric and thermal characteristics of LED chips with different lifetime levels under the same number of current pulses, the stability and sensitivity of the change amount and rate of photoelectric and thermal characteristics in distinguishing the degree of degradation are evaluated. The change amount and rate of photoelectric and thermal characteristics with strong distinguishing ability are sorted and grouped into a set of photoelectric and thermal change characteristics of LED chips that are sensitive to lifetime, providing input for the subsequent construction of photoelectric and thermal pulse lifetime fingerprint of LED chips.

[0052] The ability to distinguish between changes in photoelectric and thermal characteristics with strong discriminative power refers to the ability of LED chips to differentiate between long-life and short-life characteristics when the average difference between the long-life and short-life groups is much greater than the fluctuation within each group and the correlation coefficient with lifespan performance is close to positive or negative one.

[0053] It should be noted that the correspondence between photoelectric and thermal changes and lifespan performance in historical aging data specifically refers to the correlation between photoelectric and thermal changes such as luminous flux changes, color coordinate shifts, and junction temperature changes within a certain period of time in historical aging data and the time required for LED chips to reach the failure criteria. For example, LED chips with a large decrease in luminous flux and a continuous increase in junction temperature in the first 100 hours tend to reach the failure criteria earlier, while LED chips with smaller changes in luminous flux and stable junction temperature have a relatively longer lifespan.

[0054] The ability to distinguish the amount and rate of change of photoelectric and thermal properties at different lifespan levels refers specifically to the photoelectric and thermal characteristics that show a stable difference between long-life LED chips and short-life LED chips under the same test time or the same number of current pulses. For example, LED chips with a smaller decrease in luminous flux and a slower increase in reverse leakage current in the first few current pulses often correspond to a longer lifespan, while LED chips with a rapid decrease in luminous flux and a faster increase in reverse leakage current correspond to a shorter lifespan.

[0055] Based on the set of photoelectric and thermal change characteristics, and in accordance with the preset feature order and fixed length, the photoelectric and thermal change amount and rate of change of each LED chip under each current pulse are spliced ​​together to construct the photoelectric and thermal pulse lifetime fingerprint of the LED chip for subsequent lifetime estimation.

[0056] Furthermore, based on the set of photoelectric and thermal change characteristics, the photoelectric and thermal change amounts and rates corresponding to each LED chip under each current pulse are arranged sequentially according to the order of photoelectric and thermal changes. First, within a single current pulse, the photoelectric and thermal change amounts and rates recorded in the set of photoelectric and thermal change characteristics are arranged sequentially. Then, in the current pulse sequence dimension, they are expanded sequentially according to the current pulse order to form a one-dimensional feature sequence. The one-dimensional feature sequence of each LED chip is truncated to a fixed length to ensure that the one-dimensional feature sequences of all LED chips are consistent in length. The one-dimensional feature sequences of uniform length are bound to the corresponding LED chip numbers to construct the photoelectric and thermal pulse lifetime fingerprint of the LED chips for subsequent lifetime estimation.

[0057] It should be noted that the photoelectric and thermal pulse lifetime fingerprint of LED chips condenses the complex photoelectric and thermal changes of LEDs under pulse stress into a set of features that can sensitively reflect the degree of degradation, enabling the lifetime rapid estimation model to more accurately determine the chip lifetime and remaining lifetime by relying on only a small amount of short-term test data.

[0058] S4. Construct a rapid lifetime estimation model, and predict the lifetime of the LED chip based on the photoelectric thermal pulse lifetime fingerprint and operating condition information, generating the predicted lifetime of the LED chip.

[0059] A fast lifetime estimation model is constructed based on an input fusion layer, a physical index extraction layer, a feature representation layer, and a lifetime regression layer.

[0060] The input fusion layer concatenates the photoelectric thermal pulse lifetime fingerprint of the LED chip with the operating condition information in a fixed order and performs standardization processing to generate a unified input feature vector that serves as the starting point for subsequent calculations.

[0061] Furthermore, the input fusion layer reads the photoelectric and thermal pulse lifetime fingerprint and corresponding operating condition information of each LED chip, and sequentially arranges and splices the photoelectric and thermal features of each dimension in the photoelectric and thermal pulse lifetime fingerprint with the operating condition information to obtain a dimensionally complete spliced ​​feature sequence. Based on the spliced ​​feature sequence, the mean and standard deviation of the spliced ​​feature sequences of all LED chips in the training phase are statistically analyzed according to the feature dimensions to generate a standardized parameter set for subsequent scale adjustment. The standardized parameter set is used to standardize the spliced ​​feature sequence of each LED chip dimension by dimension, so that the photoelectric and thermal feature data and operating condition information of different dimensions are on a unified numerical scale. The standardized spliced ​​feature sequence is output as a unified input feature vector, providing a unified input for the subsequent physical index extraction layer and feature representation layer.

[0062] It should be noted that the fixed order is "optical characteristics first, electrical characteristics second, thermal characteristics third, and finally operating condition information".

[0063] Operating condition information refers to the environmental temperature, operating current, and driving voltage parameters of the LED chip during current pulse testing or actual operation. These parameters are obtained by reading the driving power supply settings and simultaneously recording data from environmental monitoring sensors.

[0064] The unified input feature vector reflects the photoelectric and thermal change patterns and operating conditions of LED chips under current pulse stress, which are composed of the optical, electrical, and thermal properties that evolve with the pulse sequence, as well as the degradation state reflected therefrom.

[0065] The physical index extraction layer extracts photoelectric and thermal response rate information (the rate of change of luminous flux per unit time, the increase of reverse leakage current, and the rate of junction temperature rise) according to the degradation rate rule, and transforms them into degradation rate index and thermal stress index that can directly reflect the early degradation rate and thermal stress intensity through principal component analysis.

[0066] Furthermore, based on the degradation rate rule, the physical index extraction layer locates the feature dimensions of luminous flux change rate, reverse leakage current increment, and junction temperature rise rate within the unified input feature vector of the photoelectric thermal response rate information. It then arranges the photoelectric thermal response rate information of each LED chip within each current pulse interval into a photoelectric thermal response rate matrix according to the current pulse order. After obtaining the photoelectric thermal response rate matrix, principal component analysis is used to perform covariance analysis and eigenvector decomposition on the photoelectric thermal response rate information. Principal components that can centrally characterize the combined behavior of luminous flux change rate and reverse leakage current increment are extracted as degradation rate indices, and principal components that can centrally characterize the junction temperature rise rate are extracted as thermal stress indices. The degradation rate indices and thermal stress indices are output together for subsequent fusion and compression recombination by the feature characterization layer.

[0067] It should be noted that the degradation rate rule is based on the correspondence between different photoelectric and thermal response rates and lifetime length in historical aging data. It is obtained by combining thermoelectric physical mechanism analysis to select the rate of change of luminous flux, the increase of reverse leakage current, and the rate of increase of junction temperature, as well as their statistical calculation methods, which can stably distinguish lifetime levels. It is used to determine the feature dimension of photoelectric and thermal response rate information in a unified input feature vector and to specify the statistical window and average or extreme value calculation method in each current pulse interval, so as to provide a unified calculation standard for extracting degradation rate index and thermal stress index.

[0068] The feature representation layer employs feature concatenation operations to fuse the unified input feature vector with the degradation rate index and thermal stress index, and then compresses and recombines them through nonlinear transformation to obtain a lifetime representation vector that centrally represents the degree of degradation and the influence of usage conditions. The expression is as follows: ; in, It is a lifetime representation vector. It is the ReLU function. It is a trainable weight matrix. It is a unified input feature vector. It is an indicator of degradation rate. It is a thermal stress index. It is the bias vector; It should be noted that the trainable weight matrix refers to the parameter matrix used in the input fusion layer and feature representation layer to map the unified input feature vector to the lifetime representation vector. It is obtained by supervised learning using the measured lifetime of a portion of the calibrated samples and by gradient iterative optimization.

[0069] The lifetime regression layer estimates the predicted lifetime of the LED chip based on the lifetime representation vector and uses gradient boosting regression tree (GBDT). During the training phase, the error between the predicted lifetime and the measured lifetime is used to optimize the feature representation layer and the input fusion layer, thus forming a complete fast lifetime estimation model.

[0070] Furthermore, the lifetime regression layer takes the lifetime representation vector output by the feature representation layer as input and uses the gradient boosting regression tree (GBDT) to iteratively fit the lifetime representation vector in multiple rounds. In each round, a new regression subtree is generated based on the current residual and superimposed on the existing regression structure to obtain the predicted lifetime of the LED chip corresponding to the lifetime representation vector. During the training phase, the measured lifetime of some calibration samples is selected as training data. The error between the predicted lifetime of the LED chip output by the lifetime regression layer and the measured lifetime of some calibration samples is calculated. Based on the error, the tree structure parameters of the gradient boosting regression tree (GBDT) are updated. At the same time, the weights and biases in the feature representation layer and the input fusion layer are adjusted through error backpropagation to make the lifetime representation vector more closely match the measured lifetime response. Finally, a fast lifetime estimation model that can stably output the predicted lifetime of the LED chip is formed.

[0071] By inputting the photoelectric thermal pulse lifetime fingerprint and operating condition information into the lifetime fast estimation model, the predicted lifetime of the LED chip is obtained, expressed as: ; in, It is the predicted lifespan of the LED chip (in hours). It is the total number of trees in the gradient boosting regression tree. It is the first The learning rate of each tree. It is the first The number of leaf nodes in a tree. It is the first Tree No. The predicted value of each leaf node. It is the first Tree No. The feature space region corresponding to each leaf node. It is an indicator function used to determine the lifetime characterization vector. Whether it falls into the first Tree No. The feature space region corresponding to each leaf node If the value falls into the range, output 1; otherwise, output 0.

[0072] It should be noted that the learning rate is a coefficient used in gradient boosting regression trees to scale the output of the m-th tree to control the overall model update step size, and is determined through parameter tuning during training via cross-validation. The feature space region refers to the set of value ranges in the multidimensional space composed of the features of the lifetime representation vector, defined by the splitting conditions of the gradient boosting regression tree, that cause the lifetime representation vector to be grouped into the same leaf node.

[0073] S5. Based on the photoelectric thermal pulse lifetime fingerprint, perform cluster analysis on LED chips and extract representative samples of each type for accelerated aging verification, obtain the deviation between predicted lifetime and measured lifetime, and optimize the lifetime rapid estimation model.

[0074] Based on the photoelectric and thermal pulse lifetime fingerprint of LED chips, cluster analysis is performed on the LED chips. The photoelectric and thermal pulse lifetime fingerprints are divided into several fingerprint cluster categories according to the degree of similarity, and lifetime fingerprint clustering results containing the category information of each LED chip are generated.

[0075] Furthermore, the photoelectric and thermal pulse lifetime fingerprints of all LED chips are normalized based on a unified dimension and length. Euclidean distance or cosine distance is used as the similarity measure. Unsupervised clustering is performed on the photoelectric and thermal pulse lifetime fingerprints of LED chips using clustering analysis. LED chips with high similarity in photoelectric and thermal change patterns are grouped into the same fingerprint cluster category, resulting in an initial clustering result containing the central features and a list of members for each fingerprint cluster category. Based on the initial clustering result, the number of clusters and clustering stability are checked according to the number of LED chips and the dispersion of photoelectric and thermal changes within each fingerprint cluster category. If necessary, the number of clusters is adjusted and the clustering analysis is repeated. Finally, a lifetime fingerprint clustering result is formed, recording the fingerprint cluster category to which each LED chip belongs. This provides a basis for extracting accelerated aging verification samples from the lifetime fingerprint clustering result.

[0076] Based on the lifetime fingerprint clustering results, several LED chips are extracted from the LED chips corresponding to each type of photoelectric thermal pulse lifetime fingerprint as accelerated aging verification objects, generating an accelerated aging verification sample set.

[0077] Furthermore, combining the lifetime fingerprint clustering results, the LED chips corresponding to each fingerprint cluster category are traversed. Based on the distance between the photoelectric and thermal pulse lifetime fingerprint and the center of the fingerprint cluster category, as well as the predicted lifetime distribution of the LED chip, a representative score is calculated. LED chips that are close to the cluster center and cover different lifetime levels within the predicted lifetime range of this category are marked as candidate representative samples. After obtaining the candidate representative samples for each fingerprint cluster category, according to the preset extraction quantity for each fingerprint cluster category, LED chips with high representative scores and differential distribution in photoelectric and thermal pulse lifetime fingerprint features are selected from the candidate representative samples. These are then compiled into a list of accelerated aging verification objects labeled with fingerprint cluster categories. The accelerated aging verification object lists for each fingerprint cluster category are merged to generate an accelerated aging verification sample set containing LED chip number, fingerprint cluster category, and reserved accelerated aging condition configuration information.

[0078] The LED chips in the accelerated aging verification sample set were continuously operated under accelerated aging conditions until the failure criteria were met. The measured lifespan of each LED chip was recorded, and the measured lifespan was compared with the predicted lifespan of the corresponding LED chip to generate a lifespan deviation data set. Furthermore, based on the accelerated aging verification sample set, each LED chip is continuously powered on and operated under pre-set accelerated aging conditions, and its photoelectric and thermal characteristic data are periodically monitored. When the photoelectric and thermal characteristic data meet the failure criteria, the corresponding operating time is recorded as the measured lifetime of the LED chip. The predicted lifetime of each LED chip in the accelerated aging verification sample set is pre-calculated using a lifetime rapid estimation model. The measured lifetime and predicted lifetime of each LED chip are paired according to the LED chip number. The lifetime difference or relative error is calculated for each pair of measured lifetime and predicted lifetime, and is compiled and archived together with the LED chip number and corresponding accelerated aging conditions. Finally, a lifetime deviation data set is generated for subsequent optimization of the lifetime rapid estimation model.

[0079] Based on the lifetime deviation data set, the parameters of the lifetime regression layer and the photoelectric and thermal change characteristics in the feature characterization layer of the rapid lifetime estimation model are optimized and updated.

[0080] Furthermore, based on the lifetime deviation data set, the lifetime deviation data set is first associated with the corresponding lifetime characterization vector and the photoelectric thermal pulse lifetime fingerprint of the LED chip according to the LED chip number, generating a joint sample set containing the lifetime characterization vector, the photoelectric thermal pulse lifetime fingerprint of the LED chip, and the lifetime deviation. On the basis of the joint sample set, the lifetime deviation distribution on different lifetime characterization vector intervals and different photoelectric thermal change characteristics is statistically analyzed. By grouping the lifetime deviation data set according to the leaf node number of the lifetime regression layer and the value range of the photoelectric thermal change characteristics, the mean absolute lifetime deviation or variance of each group is calculated, and the leaf nodes and photoelectric thermal change characteristics with larger mean absolute lifetime deviation or variance are selected as objects that contribute more to the lifetime deviation. According to the lifetime deviation distribution, the tree structure parameters and leaf node output values ​​in the lifetime regression layer are updated by gradient, and the weight coefficients of the photoelectric thermal change characteristics in the feature characterization layer are adjusted to reduce the lifetime prediction error. The updated lifetime regression layer and feature characterization layer together constitute a fast lifetime estimation model corrected by the lifetime deviation data set.

[0081] This embodiment also provides a computer device applicable to the LED chip lifetime rapid estimation method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the LED chip lifetime rapid estimation method proposed in the above embodiment.

[0082] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0083] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the method for rapidly estimating the lifespan of LED chips as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0084] In summary, this invention achieves the following: by constructing a photoelectric and thermal pulse lifetime fingerprint of an LED chip, compressing the photoelectric and thermal evolution process under pulse stress into a unified lifetime characterization carrier, providing a directly callable degradation feature basis for lifetime modeling; by constructing a rapid lifetime estimation model and generating a predicted lifetime of the LED chip based on the photoelectric and thermal pulse lifetime fingerprint and operating condition information, it realizes a processing link that directly outputs lifetime characterization vectors and lifetime results from short-term photoelectric and thermal test results under given operating conditions, facilitating the rapid acquisition of LED chip lifetime information in production screening and reliability assessment.

[0085] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for rapid estimation of LED chip lifespan, characterized in that: include, Collect photoelectric and thermal characteristic data of LED chips under rated conditions, and perform verification and screening based on failure criteria to generate a set of calibration samples for LED chips; Based on photoelectric and thermal characteristic data, a current pulse stress sequence was designed for calibrating the sample. Photoelectric and thermal response data were collected synchronously before and after the pulse, and then verified, completed, mapped, and stored to generate a photoelectric and thermal pulse response dataset. Based on the photoelectric thermal pulse response dataset, the photoelectric thermal change characteristics of LED chips that are sensitive to lifetime are extracted, and the photoelectric thermal pulse lifetime fingerprint of LED chips is constructed. A rapid lifetime estimation model is constructed, and the lifetime of LED chips is predicted based on photoelectric thermal pulse lifetime fingerprint and operating condition information, generating the predicted lifetime of LED chips. Based on the photoelectric thermal pulse lifetime fingerprint, LED chips are clustered and representative samples of each type are extracted for accelerated aging verification. The deviation between the predicted lifetime and the measured lifetime is obtained, and the lifetime rapid estimation model is optimized.

2. The method for rapid estimation of LED chip lifetime as described in claim 1, characterized in that: The steps for combining failure criteria to perform verification and screening, and generating a set of calibration samples for LED chips, are as follows: Each LED chip in the batch to be tested is subjected to photoelectric and thermal joint testing under rated conditions. Photoelectric and thermal characteristic data of the LED chips are collected and standardized, and a set of LED chip test data is generated. Based on historical aging data and in-service fault records, multivariate logistic regression analysis was used to statistically fit the relationship between different luminous flux maintenance rates, color coordinate offsets, upper limits of electrical parameters, and actual failure probabilities, generating failure criteria for LED chips. By combining LED chip test data sets and failure criteria, the photoelectric and thermal characteristic data of each LED chip are verified and screened, and a set of calibration samples for LED chips is generated.

3. The method for rapid estimation of LED chip lifespan as described in claim 1, characterized in that: The steps for designing the current pulse stress sequence for calibrating the sample are as follows: Based on the difference between the photoelectric and thermal characteristic data of each LED chip in the calibration sample set and the failure criteria, the allowable space of each LED chip from the failure boundary is estimated. The allowable space is statistically analyzed, and the low quantile value is selected as the safety margin of the stress level that the calibration sample can withstand. The safety margin is then converted into the target degradation excitation intensity according to the degradation excitation conversion factor. Under the constraints of safety margin and target degradation excitation intensity, the pulse width and pulse interval of the current amplitude are selected to generate a current pulse stress sequence.

4. The method for rapid estimation of LED chip lifespan as described in claim 3, characterized in that: The steps for performing verification completion and mapping storage to generate a photoelectric thermal pulse response dataset are as follows. Under controlled ambient temperature, current pulses are applied sequentially to each LED chip according to the current pulse stress sequence. Photoelectric thermal response data are collected synchronously before and after each current pulse and collected according to the LED chip number and the order of current pulses to form a set of original photoelectric thermal response data. The original photoelectric thermal response dataset is processed by filling in missing values, unifying the format, and standardizing the data to generate a photoelectric thermal pulse response dataset.

5. The method for rapid estimation of LED chip lifetime as described in claim 1, characterized in that: The steps for extracting the photoelectric and thermal change characteristics of the LED chip that are sensitive to its lifetime, and constructing the photoelectric and thermal pulse lifetime fingerprint of the LED chip, are as follows: Based on the photoelectric and thermal pulse response dataset, the photoelectric and thermal characteristics of each LED chip before and after each current pulse relative to the previous moment and the rate of change per unit time are calculated to obtain a set of photoelectric and thermal change trajectories. Based on the set of photoelectric and thermal change trajectories and the correspondence between photoelectric and thermal changes and lifespan performance in historical aging data, we analyze the ability of different photoelectric and thermal characteristic change amounts and rates to distinguish the degree of degradation, screen out the photoelectric and thermal change amounts and rates that can distinguish different lifespan levels, and form a set of LED photoelectric and thermal change characteristics. Based on the set of photoelectric and thermal change characteristics, and in accordance with the preset characteristic order and fixed length, the photoelectric and thermal change amount and rate of change of each LED chip under each current pulse are spliced ​​together to construct the photoelectric and thermal pulse lifetime fingerprint of the LED chip.

6. The method for rapid estimation of LED chip lifetime as described in claim 1, characterized in that: The steps for constructing a rapid lifetime estimation model and predicting the lifetime of the LED chip based on the photoelectric thermal pulse lifetime fingerprint and operating condition information are as follows: A fast lifetime estimation model is constructed based on an input fusion layer, a physical index extraction layer, a feature representation layer, and a lifetime regression layer. The input fusion layer concatenates the photoelectric thermal pulse lifetime fingerprint of the LED chip with the operating condition information in a fixed order and performs standardization processing to generate a unified input feature vector. The physical index extraction layer extracts photoelectric and thermal response rate information according to the degradation rate rule, and transforms it into degradation rate index and thermal stress index through principal component analysis. The feature representation layer fuses the unified input feature vector with the degradation rate index and thermal stress index, and compresses and recombines them through nonlinear transformation to obtain the lifetime representation vector; The lifetime regression layer estimates the predicted lifetime of the LED chip based on the lifetime representation vector and uses gradient boosting regression tree. During the training phase, the error between the predicted lifetime and the measured lifetime is used to optimize the feature representation layer and the input fusion layer, thus forming a complete fast lifetime estimation model. By inputting photoelectric thermal pulse lifetime fingerprints and operating condition information into a rapid lifetime estimation model, the predicted lifetime of the LED chip can be obtained.

7. The method for rapid estimation of LED chip lifetime as described in claim 1, characterized in that: The steps are as follows: cluster analysis of LED chips based on photoelectric thermal pulse lifetime fingerprints, extraction of representative samples from each category for accelerated aging verification, obtaining the deviation between predicted and measured lifetimes, and optimizing the rapid lifetime estimation model. Based on the photoelectric and thermal pulse lifetime fingerprint of LED chips, cluster analysis is performed on LED chips, and the photoelectric and thermal pulse lifetime fingerprints are divided into several fingerprint cluster categories according to the degree of similarity, generating lifetime fingerprint clustering results. Based on the lifetime fingerprint clustering results, several LED chips are extracted from the LED chips corresponding to each type of photoelectric thermal pulse lifetime fingerprint as accelerated aging verification objects to generate an accelerated aging verification sample set. The LED chips in the accelerated aging verification sample set were continuously operated under accelerated aging conditions until the failure criteria were met. The measured lifespan of each LED chip was recorded, and the measured lifespan was compared with the predicted lifespan of the corresponding LED chip to generate a lifespan deviation data set. Based on the lifetime deviation data set, the parameters of the lifetime regression layer and the photoelectric and thermal change characteristics in the feature characterization layer of the lifetime rapid estimation model are optimized and updated.

8. The method for rapid estimation of LED chip lifetime as described in claim 1, characterized in that: The photoelectric and thermal characteristic data include luminous flux, color coordinates, peak wavelength, forward voltage, reverse leakage current, dynamic on-resistance, estimated junction temperature, junction-to-shell thermal resistance, and temperature rise under pulsed current.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the LED chip lifetime rapid estimation method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the LED chip lifetime rapid estimation method according to any one of claims 1 to 8.

Images