Defect detection method and system for LED flexible lamp strip

Abstract

The invention relates to the technical field of LED lamp strip testing, solves the problem that in the prior art, multi-dimensional defect detection of an LED flexible lamp strip cannot be achieved under a complex working condition, and provides a defect detection method and system for the LED flexible lamp strip, and the method comprises the steps: obtaining preset identification information of a to-be-detected LED flexible lamp strip; obtaining a power-on initial electrical parameter according to a power supply safety parameter corresponding to the preset identification information; when the power-on initial electrical parameter meets a preset stable condition, driving detection is performed on the LED flexible lamp strip, and a driving protocol parameter of a target driving integrated circuit is acquired; according to a preset performance test condition corresponding to the driving protocol parameter, performing a multi-dimensional performance test on the LED flexible lamp strip to obtain a performance test result; and inputting the performance test result into a pre-trained multi-dimensional defect detection model to obtain a defect detection result. According to the invention, defect detection of the LED flexible lamp strip under multi-model driving and complex working conditions can be realized.

Description

Technical Field

[0001] This invention relates to the field of LED light strip testing technology, and in particular to a defect detection method and system for flexible LED light strips. Background Technology

[0002] LED flexible light strips are widely used in architectural decoration, stage lighting, displays, and automotive interiors due to their flexibility, bendability, rollability, installation flexibility, and uniform light emission. However, during production and use, flexible light strips are prone to uneven brightness, color shift, flickering, and even partial LED failure due to differences in driver integrated circuits, insufficient process consistency, or device performance issues caused by prolonged operation. Therefore, systematic defect detection of LED flexible light strips is not only a necessary step to ensure product quality and reliability but also an important means to improve user experience and reduce after-sales maintenance costs.

[0003] Existing LED strip defect detection technologies mainly include static electrical performance testing and optical inspection methods. For example, sampling the operating current and voltage of the strip can determine if the circuit is abnormal; or optical sensors can be used to detect brightness and chromaticity parameters to determine if there are dark spots, color drift, or uneven light emission. Some solutions also incorporate high-temperature performance testing to assess the reliability of the strip under long-term high-load operation. These detection methods can detect obvious failures to some extent, but they still have significant limitations. However, existing defect detection technologies generally suffer from the following problems: First, most methods are only designed for a single type of driver circuit, lacking adaptive identification and matching for different driver integrated circuits, and thus failing to cover the testing needs of multi-model compatibility; second, traditional tests are mostly based on single operating conditions, making it difficult to accurately reflect the potential defects of the strip under complex operating scenarios such as multiple power levels, different color combinations, and different refresh rates; third, for the unique bending and winding applications of flexible strips, existing detection technologies typically cannot simultaneously acquire electrical, optical, thermal, and mechanical response data, making it difficult to effectively assess the failure risk caused by mechanical stress.

[0004] Chinese patent CN118351422A discloses a training method, device, computer-readable storage medium, and LED strip defect detection method for an LED strip defect detection model. The method includes: selecting typical positive samples from an LED strip training set containing only positive samples using K-means++ clustering to construct a memory pool; expanding the training set by generating pseudo-anomaly samples using Berlin noise and texture data; and training a detection model that can output a defect prediction map for the image under test by combining a pre-trained ResNet18 feature extraction network, a multi-scale feature fusion network, and a spatial attention mechanism. This improves the robustness of the model and facilitates its transfer and use across different image datasets. The aforementioned patented solutions always use the luminous images captured by the camera as the sole input, and the detection dimensions are mainly focused on brightness distribution and luminous appearance. They do not model or monitor the multi-dimensional characteristic defects of flexible light strips under complex working conditions such as different driving currents, different PWM dimming modes, bending and stretching states, thermal drift caused by long-term lighting, and voltage drops, including electrical, optical, thermal, and mechanical stresses. They also do not involve the integration with data from online electrical performance testing, temperature sensing, strain monitoring, etc., making it difficult to reflect the comprehensive performance failure risk of flexible light strips in real application scenarios in a timely manner.

[0005] Therefore, how to achieve multi-dimensional defect detection of flexible LED light strips under complex working conditions is an urgent technical problem to be solved. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a defect detection method and system for flexible LED light strips, in order to solve the problem that the prior art cannot achieve multi-dimensional defect detection of flexible LED light strips under complex working conditions.

[0007] In a first aspect, embodiments of the present invention provide a defect detection method for flexible LED light strips, the method comprising: Obtain the preset identification information of the LED flexible light strip to be tested; Based on the power supply safety parameters corresponding to the preset identification information, the LED flexible light strip is powered on in a controlled manner to obtain the initial power-on electrical parameters. When the initial electrical parameters meet the preset stability conditions, drive detection is performed on the LED flexible light strip to obtain the drive protocol parameters of the target driver integrated circuit; Based on the preset performance test conditions corresponding to the driving protocol parameters, the LED flexible light strip is subjected to multi-dimensional performance tests to obtain the performance test results. The performance test results are input into the pre-trained multi-dimensional defect detection model to obtain the defect detection results.

[0008] In an optional embodiment, the controlled power-on of the LED flexible light strip based on the power supply safety parameters corresponding to the preset identification information, and the acquisition of initial power-on electrical parameters, includes: Based on the preset identification information, determine the rated power supply voltage and rated operating current corresponding to the preset identification information to obtain the power supply reference parameters; Based on the power supply reference parameters, the output voltage of the power supply and the current threshold of the current limiting circuit are configured to obtain the power supply control parameters; According to the power supply control parameters, the control power supply gradually applies voltage to the LED flexible light strip to execute a controlled power-on process. During the controlled power-on process, the starting current and the voltage across the two ends of the LED flexible light strip are used as electrical measurement data for power-on. Based on the electrical measurement data obtained upon power-on, the peak starting current and end-to-end voltage drop are calculated to obtain the initial electrical parameters upon power-on.

[0009] In an optional embodiment, when the initial power-on electrical parameters meet a preset stability condition, performing drive detection on the LED flexible light strip to obtain the drive protocol parameters of the target driver integrated circuit includes: When the initial electrical parameters of the power-on meet the preset stability conditions, the candidate driver integrated circuit corresponding to the preset identification information is determined according to the preset identification information; Based on the drive signal template set corresponding to the candidate driver integrated circuit, test drive signals are applied to the LED flexible light strip one by one, and the optical response data and electrical response data of the LED flexible light strip during the test drive process are obtained to obtain the test response dataset. Based on the trial response dataset, the response consistency of different driving signal templates in the driving signal template set is compared to determine the target driving signal template that matches the LED flexible light strip. Based on the target drive signal template, the protocol type, timing parameters, level parameters, and data format parameters are extracted to obtain the drive protocol parameters.

[0010] In an optional embodiment, when the initial power-on electrical parameters meet a preset stability condition, obtaining the drive signal template set corresponding to the preset candidate driver integrated circuit according to the preset identification information includes: Based on the initial electrical parameters after power-on, the rate of change of starting current and the end-to-end voltage fluctuation rate are calculated to obtain stability judgment indicators; The stability determination index is compared with a preset stability threshold. If the stability determination index is less than the stability threshold, then the initial electrical parameters of the power-on meet the preset stability condition. When the initial electrical parameters meet the preset stability conditions, the product model, power supply level and interface level of the LED flexible light strip are analyzed according to the preset identification information to obtain a set of candidate driver integrated circuits. According to the preset drive signal template library, the drive signal template corresponding to each candidate driver integrated circuit in the candidate driver integrated circuit set is obtained as the drive signal template set.

[0011] In an optional embodiment, the step of performing multi-dimensional performance testing on the LED flexible light strip according to preset performance test conditions corresponding to the driving protocol parameters, and obtaining the performance test results, includes: The protocol type, timing parameters, and level parameters are analyzed to determine the driving method of the flexible LED light strip; Based on the driving method and preset identification information, a set of candidate driver integrated circuit models is determined; Based on the driving method and the set of candidate driver integrated circuit models, performance test conditions are retrieved from a preset performance test condition library to obtain an initial performance test condition set. Based on the timing parameters and level parameters, the initial performance test condition set is verified and filtered to obtain the preset performance test conditions; Based on the preset performance test conditions, the LED flexible light strip is subjected to multi-dimensional performance tests to obtain performance test results.

[0012] In an optional embodiment, the step of performing multi-dimensional performance tests on the flexible LED light strip according to the preset performance test conditions to obtain performance test results includes: Based on the preset performance test conditions, the output parameters used for drive control are configured to obtain drive configuration parameters; Based on the driver configuration parameters, performance test conditions corresponding to different driver integrated circuit models are invoked to generate an initial performance test sequence covering the characteristics of multiple driver models. According to the initial performance test sequence, drive signals are sequentially applied to the flexible LED light strip to form tests at different power levels, and performance execution results are obtained. Based on the performance execution results, obtain a multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performances; Based on the multidimensional performance response dataset, the response characteristics of different driver integrated circuit models are compared and analyzed, and the test results under different performance conditions are classified and processed to obtain the performance test results.

[0013] In an optional embodiment, obtaining the multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performance conditions based on the performance execution results includes: Based on the performance results, the operating times of the LED flexible light strip under different driver integrated circuit models and different performance conditions are marked to obtain operating condition index information; Based on the operating condition index information, obtain the electrical operating data, optical operating data, thermal operating data and mechanical response data of the LED flexible light strip under various operating conditions; The electrical operating data, optical operating data, thermal operating data, and mechanical response data are time-aligned and spatially calibrated to obtain the multidimensional performance response dataset.

[0014] In an optional embodiment, inputting the performance test results into a pre-trained multi-dimensional defect detection model to obtain defect detection results includes: Based on the performance test results, the electrical, optical, and mechanical parameters of different driver integrated circuit models are analyzed to obtain the target detection dataset corresponding to each model. Based on the target detection datasets described above, statistical calculations are performed on current fluctuations, voltage drop drift, and power consumption changes to obtain electrical anomaly indicators; Based on the target detection datasets described above, statistical calculations are performed on brightness attenuation, chromaticity shift, and flicker amplitude to obtain optical anomaly indicators. Based on the target detection datasets described above, the temperature rise gradient, hot spot location, and bending response are analyzed to obtain mechanical anomaly indicators; The electrical anomaly index, optical anomaly index, and mechanical anomaly index are input into the multi-dimensional defect detection model to obtain the corresponding initial defect detection results; Based on the initial defect detection results, the defect types of different driver integrated circuit models and under different performance conditions are classified and summarized to obtain the defect detection results.

[0015] In an optional embodiment, the step of analyzing the temperature rise gradient, hotspot locations, and bending response based on each of the target detection datasets to obtain mechanical anomaly indicators includes: Based on the target detection datasets described above, the temperature distribution data at different sampling times are extracted and aligned to obtain a temperature distribution sequence; Based on the temperature distribution sequence, the rate of temperature change in each region is calculated to obtain the temperature rise gradient index. Based on the temperature distribution sequence, local extreme points in the temperature distribution are located and tracked to obtain hotspot location indicators. Based on the target detection datasets described above, the bending curvature, the number of bending cycles, and the corresponding current change rate and brightness change rate are correlated and calculated to obtain the bending response index; Based on the temperature rise gradient index, hot spot location index, and bending response index, the detection results are comprehensively compared and correlated to obtain the mechanical anomaly index.

[0016] In a second aspect, embodiments of the present invention provide a defect detection system for flexible LED light strips, comprising: at least one processor, at least one memory, and computer program instructions stored in the memory, wherein when the computer program instructions are executed by the processor, the method described in the first aspect is implemented.

[0017] In summary, the beneficial effects of the present invention are as follows: The present invention provides a defect detection method and system for flexible LED light strips. The method includes: acquiring preset identification information of the flexible LED light strip to be detected; controlling the power supply safety parameters of the flexible LED light strip according to the preset identification information to acquire initial power-on electrical parameters; when the initial power-on electrical parameters meet preset stability conditions, performing drive detection on the flexible LED light strip to acquire drive protocol parameters of the target drive integrated circuit; performing multi-dimensional performance testing on the flexible LED light strip according to preset performance test conditions corresponding to the drive protocol parameters to obtain performance test results; and inputting the performance test results into a pre-trained multi-dimensional defect detection model to obtain defect detection results. This invention first acquires the preset identification information of the LED flexible light strip and matches it with the corresponding power supply safety parameters. It then implements controlled power-on of the light strip and collects the initial power-on electrical parameters. This step pre-constrains uncertainties such as power-on impact and undervoltage jitter under different batches, different driver ICs, and different line losses and load conditions within a controllable range, avoiding false triggering, misjudgment, or secondary damage to devices due to unstable power-on. Secondly, once the initial power-on electrical parameters meet the preset stability conditions, the system proceeds to drive detection, extracting the drive protocol parameters of the target driver IC. This enables the system to complete protocol identification and drive alignment under different conditions, thereby mitigating the most easily changing control layer differences under complex operating conditions. The variants are transformed into resolvable protocol features; then, based on the driving protocol parameters, preset performance test conditions are selected to perform multi-dimensional performance tests on the light strip and obtain performance test results with a unified caliber. This ensures that multi-dimensional indicators are repeatable and comparable under the corresponding protocol and operating condition constraints, avoiding test condition mismatch, indicator distortion, and dimensional fragmentation caused by operating condition drift in existing technologies. Finally, the performance test results are input into a pre-trained multi-dimensional defect detection model, which comprehensively judges the coupling relationship between multi-dimensional indicators and can distinguish defect types that are similar in a single dimension. Thus, stable and generalizable multi-dimensional defect detection and location output can still be achieved under complex operating conditions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0019] Figure 1 This is a schematic flowchart of the defect detection method for LED flexible light strips in Embodiment 1 of the present invention; Figure 2 This is a flowchart illustrating the process of performing multi-condition performance tests on the LED flexible light strip and obtaining the performance test results in Embodiment 1 of the present invention. Figure 3 This is a schematic diagram of the defect detection method for LED flexible light strips in Embodiment 1 of the present invention. Figure 4 This is a schematic diagram of the defect detection system for the flexible LED light strip in Embodiment 2 of the present invention; The labels in the diagram are as follows: 1-Performance testing device; 2-Flexible LED light strip. Detailed Implementation

[0020] The features and exemplary embodiments of various aspects of the present invention will now be described in detail. To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present invention and are not configured to limit the present invention. For those skilled in the art, the present invention can be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present invention by illustrating examples of the invention.

[0021] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0022] It should be noted that all actions involving the acquisition of signals, information, or data in this invention are carried out in compliance with the relevant data protection laws and regulations of the locality and with authorization from the owner of the relevant device.

[0023] Example 1 Please see Figure 1 This invention provides a defect detection method for flexible LED light strips, the method comprising: Obtain the preset identification information of the LED flexible light strip to be tested; Specifically, before testing LED flexible light strips, their preset identification information is first obtained. This information typically includes product model, production batch, electrical parameters, and driver chip type, which is the basis for clarifying product attributes and testing conditions. This process provides a basis for subsequent power supply parameter configuration and test scheme selection, ensuring the relevance and accuracy of the test. This is accomplished by reading tags, database records, or system preset information, ensuring a one-to-one correspondence between the tested object and the test conditions, and avoiding test failures or equipment damage due to information discrepancies.

[0024] Based on the power supply safety parameters corresponding to the preset identification information, the LED flexible light strip is powered on in a controlled manner to obtain the initial power-on electrical parameters. Specifically, after obtaining the basic identification information of the LED strip, it is necessary to perform controlled power-on according to its corresponding power supply safety parameters. These parameters include rated voltage, current range, and power consumption limits. Controlled power-on prevents damage to the LED strip due to excessive voltage or current. By recording the voltage and current during the power-on process, electrical parameter data reflecting the initial operating state of the LED strip can be obtained. This data is not only used to determine whether the circuit is normal, but also provides a reliable basis for subsequent stability testing.

[0025] When the initial electrical parameters meet the preset stability conditions, drive detection is performed on the LED flexible light strip to obtain the drive protocol parameters of the target driver integrated circuit; Specifically, after confirming a smooth and reliable power-on process, further testing of the LED strip's driver section is necessary to obtain its driver protocol parameters. These parameters include data transmission methods, control signal timing, and color arrangement, which determine the LED strip's response to control signals. Identifying the driver protocol ensures that subsequent control signals can be correctly parsed and executed, enabling the LED strip to illuminate and change according to the set requirements during testing, thus creating conditions for performance testing.

[0026] Based on the preset performance test conditions corresponding to the driving protocol parameters, the LED flexible light strip is subjected to multi-dimensional performance tests to obtain the performance test results. Specifically, after mastering the driving protocol parameters, the LED strip undergoes performance testing under multiple operating conditions based on its corresponding test conditions. Preset performance test conditions typically include different electrical loads, operating brightness, temperature environments, and humidity variations. Based on these preset performance test conditions, the flexible LED strip is subjected to multi-dimensional performance testing. By running the strip under different conditions for extended periods, various scenarios in actual use can be simulated. The performance test results cover electrical performance, optical performance, and structural stability, providing a comprehensive and reliable basis for subsequent defect identification.

[0027] The performance test results are input into the pre-trained multi-dimensional defect detection model to obtain the defect detection results.

[0028] Specifically, the pre-trained multi-dimensional defect detection model is a discrimination model with pre-trained and fixed parameters. It is used to comprehensively analyze the performance test results of LED strips under multiple operating conditions and multiple indicators, and output defect detection results. These results are typically presented in the form of whether a defect exists, the type of defect, and the corresponding confidence level or risk level. For example, the output might indicate abnormal power supply risk, abnormal drive response, abnormal light color consistency, or abnormal thermal stability. This transforms the dispersed electrical and optical performance data into detection conclusions that can be directly used for quality inspection or graded handling. By addressing the problem that single-threshold or single-dimensional judgments under complex operating conditions are easily distorted by noise, line loss, temperature rise, and load fluctuations, the model jointly discriminates the correlations between multi-dimensional data. This distinguishes normal fluctuations introduced by changes in operating conditions from abnormal patterns truly caused by device defects, ensuring consistency and transferability of detection results across different batches, different types of driver chips, and different test conditions. It also provides clear defect indications for subsequent traceability and rework.

[0029] In an optional embodiment, before obtaining the defect detection result by inputting the performance test result into the pre-trained multi-dimensional defect detection model, the method further includes: Obtain the driver integrated circuit model and performance information corresponding to the pre-collected LED flexible light strip samples to be tested; Specifically, before training begins, the attribute boundaries of each LED flexible light strip sample are clearly defined, ensuring a one-to-one correspondence between the sample and the driver IC model and performance conditions in the subsequent training process. The driver IC model is used to characterize the differences in the driving control foundation and protocol of the sample, while the performance condition information is used to characterize the sample's operating background under different loads, operating brightness, temperature, and humidity variations. By obtaining the above information, a basic data record consistent with the sample number can be formed, ensuring that subsequent sample selection and association, as well as multi-dimensional performance testing, are conducted under traceable conditions, avoiding training bias caused by confusion in training data sources.

[0030] Based on the model and performance information of the driving integrated circuit, the LED flexible light strip samples are selected and associated to obtain a training sample set.

[0031] Specifically, the samples are organized along the dimensions of driver IC model and performance condition to ensure the training sample set is comprehensive and structured. The selection process includes samples with different driver IC models and performance conditions within the training scope, while the association process binds the same LED flexible light strip sample to its corresponding driver IC model and performance condition information, forming data units that can be directly used for subsequent testing and annotation. By obtaining the training sample set, the data input during training has a clear grouping basis, and the training performance test results generated by subsequent multi-dimensional performance tests can naturally correspond to specific driver IC models and specific performance conditions, ensuring continuous and non-skipping training logic.

[0032] Based on the training sample set, multi-dimensional performance tests were performed on the LED flexible light strip under different driver integrated circuit models and performance conditions. Electrical parameters, optical parameters and mechanical parameters were collected to obtain the performance test results for training. Specifically, multi-dimensional performance tests are performed under different driver IC models and performance conditions, ensuring complete coverage of the same type of defect under varying conditions. The collected electrical, optical, and mechanical parameters form the original basis for subsequent unified analysis and feature construction. The performance test results used for training maintain a correspondence with the training sample set, recording the multi-dimensional performance of each LED flexible light strip sample under the corresponding driver IC model and performance conditions, providing a stable data source for subsequent defect type labeling and model parameter training.

[0033] Based on the performance test results for training, electrical parameters, optical parameters and mechanical parameters are uniformly analyzed and feature constructed to obtain a multi-dimensional training feature dataset. Specifically, the performance test results used for training are transformed from raw observations into structured inputs suitable for model training, ensuring a consistent representation of data from different acquisition channels, with different dimensions, and under different driver integrated circuit models and performance conditions. Waveform and statistical analyses are performed on electrical parameters, stability and drift analyses are performed on optical parameters, and response feature analyses are performed on mechanical parameters. These analytical results are then aligned, normalized, and combined according to preset dimensions to form a multi-dimensional training feature dataset that corresponds one-to-one with each sample number. This provides a unified data foundation for subsequent defect type labeling and model training, reducing training noise introduced by sampling differences.

[0034] Based on the multi-dimensional training feature dataset and the known defect states of the samples, each training sample is labeled with a defect type to obtain a training dataset with defect labels. Specifically, the defect-labeled training dataset is used to establish supervised information for training samples, enabling the model to learn the correspondence between multi-dimensional training feature datasets and defect types. In practice, based on the known defect states of the samples, the multi-dimensional training feature datasets corresponding to the sample numbers are matched to the corresponding defect types, and can be categorized according to the manifestations of electrical anomalies, optical anomalies, and mechanical anomalies, forming a defect-labeled training dataset. This labeling process ensures that the defect manifestations of different driver integrated circuit models and under different performance conditions are uniformly incorporated into the same labeling system, facilitating the formation of transferable discrimination boundaries in subsequent training stages and reducing the risk of misjudgment caused by relying solely on a single-dimensional threshold.

[0035] Based on the training dataset with defect labels, the parameters of the preset deep learning model are trained and optimized to obtain the multi-dimensional defect detection model.

[0036] Specifically, a multi-dimensional training feature dataset is used as input, and defect labels are used as supervised targets. A pre-defined deep learning model is iteratively trained, and the model parameters are updated and optimized based on the deviation between the model output and the defect labels during training, until the preset convergence or performance requirements are met. Through this training and optimization process, the resulting multi-dimensional defect detection model can comprehensively utilize the correlation features between electrical, optical, and mechanical parameters, outputting stable defect detection results under different driver integrated circuit models and different performance conditions, thereby improving the consistency and reliability of multi-dimensional defect detection under complex operating conditions.

[0037] In an optional embodiment, the controlled power-on of the LED flexible light strip based on the power supply safety parameters corresponding to the preset identification information, and the acquisition of initial power-on electrical parameters, includes: Based on the preset identification information, determine the rated power supply voltage and rated operating current corresponding to the preset identification information to obtain the power supply reference parameters; Specifically, the rated supply voltage and rated operating current are determined based on preset identification information to obtain the power supply baseline parameters. The preset identification information refers to the model and electrical specifications recorded on the manufacturer's label, production records, or onboard storage. The rated supply voltage is, for example, 24 volts, and the rated operating current can be the typical consumption of the entire LED strip or the unit current calculated per meter. The power supply baseline parameters are the power supply baseline formed by combining these rated values ​​with engineering tolerances. The purpose of this is to clarify the safe operating range and provide a unified reference for subsequent power supply configuration, reducing misjudgments caused by inconsistent information. Typically, the consistency between the label and the actual product is confirmed by reading product files or labels, querying the production management system, and performing low-voltage short-time measurements on several samples. The measurement results are written into the test formula and archived for traceability, thus establishing a reliable baseline for subsequent power-on and testing.

[0038] Based on the power supply reference parameters, the output voltage of the power supply and the current threshold of the current limiting circuit are configured to obtain the power supply control parameters; Specifically, power supply control parameters are obtained by configuring the power supply output voltage and the current threshold of the current limiting circuit based on the power supply reference parameters. These parameters include the target output voltage, the startup curve, and the current limiting threshold. The current threshold can be set on the programmable DC power supply or implemented through external current limiting components or protection modules. The configuration aims to match the power supply behavior to the power supply reference and automatically control it in the event of abnormal transients. Therefore, soft-start or constant current mode is enabled on the power supply, a current limiter or thermal current limiting component is connected in parallel on the fixture side, and all settings are recorded in the test task and equipment log for verification. After configuration, the validity of the parameters is verified by a short-term preset power-on to ensure consistent and reproducible subsequent power-on conditions. According to the power supply control parameters, the control power supply gradually applies voltage to the LED flexible light strip to execute a controlled power-on process. Specifically, based on power supply control parameters, the power supply is controlled to gradually apply voltage to the LED strip, executing a controlled power-on process. The controlled power-on process refers to increasing the voltage to the target value according to a preset slope or stepped increments, while simultaneously applying current limits and sampling electrical signals in real time. The controller issues a boost curve, and the power supply outputs according to the curve, immediately stopping or reverting if the sampling exceeds the limit. During power-on, pre-charging, capacitor pre-discharging, or parallel current limiting can be used to weaken inrush current, and different power-on regions can be sequentially segmented to avoid transient impacts at the end. The entire process requires synchronous recording of timestamps and sampling data for playback analysis. This not only suppresses the impact of power-on transients on the devices but also captures abnormal signals in the early stages of power-on for analysis.

[0039] During the controlled power-on process, the starting current and the voltage across the two ends of the LED flexible light strip are used as electrical measurement data for power-on. Specifically, during controlled power-on, the starting current and the voltage across the terminals are collected and stored as electrical measurement data. The starting current is the current response during the power-on transient, and the voltage across the terminals are the voltage readings at the power supply inlet and outlet. Data acquisition can be achieved using sampling resistors in conjunction with differential amplifiers and high-speed analog-to-digital converters, or by using Hall effect sensors and oscilloscopes for transient capture. Voltage measurements are taken at sampling points at both the input and outlet terminals, and sampled through isolation amplifiers. Sampling requires time synchronization and timestamps; transient and steady-state data are stored separately. Anti-aliasing filtering and noise reduction processing are performed on the data when necessary. The collected data provides primary evidence for subsequent indicator extraction and anomaly detection.

[0040] Based on the electrical measurement data obtained upon power-on, the peak starting current and end-to-end voltage drop are calculated to obtain the initial electrical parameters upon power-on.

[0041] Specifically, the peak starting current and end-to-end voltage drop are calculated from the collected electrical measurement data to obtain the initial electrical parameters for power-on. Data processing first involves synchronization correction and filtering, then the peak starting current is extracted using a peak detection algorithm, and the end-to-end voltage drop is obtained by subtracting the end voltage from the input voltage. To improve robustness, the results of short-time maximum values ​​and steady-state average values ​​can be combined, and these initial electrical parameters are compared with preset thresholds or historical baselines to determine if any anomalies exist. The obtained initial electrical parameters for power-on serve as a quantitative benchmark for subsequent stability assessment and drive identification, and are archived together with the original sampling records to support quality traceability and subsequent analysis.

[0042] In an optional embodiment, when the initial power-on electrical parameters meet a preset stability condition, performing drive detection on the LED flexible light strip to obtain the drive protocol parameters of the target driver integrated circuit includes: When the initial electrical parameters of the power-on meet the preset stability conditions, the candidate driver integrated circuit corresponding to the preset identification information is determined according to the preset identification information; Specifically, when the initial electrical parameters meet the preset stability conditions, the corresponding candidate driver integrated circuit is determined based on the preset identification information. Candidate driver integrated circuits refer to several possible control chip types inferred from product identification and physical information, such as single-wire serial drive type, parallel or SPI type, multi-channel PWM type, etc. The purpose of this step is to narrow down the possible driver range based on the known basic attributes of the LED strip for subsequent targeted testing. A common practice is to first search the database or bill of materials for the driver list corresponding to the identification, while also considering the board silkscreen, package features, and the number of visible pins for identification. If necessary, non-invasive electrical measurements are used to determine the number of data lines and the presence of pull-up resistors to assist in screening. Finally, a candidate driver set list is formed and written into the test task. This reduces blind testing, lowers the risk of incorrect driving, and improves subsequent identification efficiency.

[0043] Based on the drive signal template set corresponding to the candidate driver integrated circuit, test drive signals are applied to the LED flexible light strip one by one, and the optical response data and electrical response data of the LED flexible light strip during the test drive process are obtained to obtain the test response dataset. Specifically, based on the drive signal template set corresponding to the candidate drive set, trial drive signals are applied to each LED strip one by one, and optical and electrical response data are acquired during the trial drive process to obtain a trial response dataset. The drive signal template set is a library of signal sequences with typical timing, level, and data format prepared for each candidate drive. The trial drive signal refers to a low-risk verification sequence emitted one by one, the purpose of which is to verify the matching degree between the template and the device under test by observing the actual response. In actual operation, the template sequence is usually sent first in a low brightness or short pulse width manner, while a high-speed camera or photoelectric sensor is used to record the pixel brightness and color changes, and a current sensor and oscilloscope are used to record the electrical waveform. All optical and electrical samples are timestamped and saved synchronously as a trial response dataset. If abnormal current or overvoltage is encountered, the test is immediately stopped and the event is recorded, thereby accumulating original response samples for comparison under the premise of ensuring safety.

[0044] Based on the trial response dataset, the response consistency of different driving signal templates in the driving signal template set is compared to determine the target driving signal template that matches the LED flexible light strip. Specifically, the response consistency of different templates in the driving signal template set is compared based on the trial response dataset to determine the target driving signal template that matches the LED strip. Response consistency refers to the degree of agreement between the optical and electrical responses and the expected characteristics of the template in terms of timing, amplitude, and mapping relationship. The purpose is to select the most likely correct template from the candidate templates using quantitative indicators. The comparison method may include correlation calculation of pixel brightness curves and color coordinates, shape similarity measurement of current and signal timing, and channel-by-channel verification of color sequence mapping. The templates are sorted by weighted scoring or threshold filtering, and the one with the best score is selected. When the scores are close, the top few can be sent a confirmation sequence to eliminate ambiguity, so that the most matching template can still be reliably identified in a confusing environment.

[0045] Based on the target drive signal template, the protocol type, timing parameters, level parameters, and data format parameters are extracted to obtain the drive protocol parameters.

[0046] Specifically, based on the determined target drive signal template, the protocol type, timing parameters, level parameters, and data format parameters are extracted to obtain the drive protocol parameters. The protocol type refers to the bus and communication method used by the driver; timing parameters refer to bit width, pulse width, and reset timing, etc.; level parameters refer to logic high and low levels and threshold values; and data format parameters refer to the number of channel bits, color sequence, and frame encapsulation rules. The purpose of this step is to transform the empirical description of the template into a reusable and precise definition for formal control. A common practice is to perform fine edge detection and amplitude measurement on trial sampling of the target template to calculate specific timing and level values. Then, the color sequence and channel bit width are confirmed by writing a known reference pattern and observing the optical output. Finally, a playback verification step is used to confirm that the extracted parameters can stably drive the light strip. The obtained drive protocol parameters can be directly used for subsequent normal driving and performance testing.

[0047] In an optional embodiment, when the initial power-on electrical parameters meet a preset stability condition, obtaining the drive signal template set corresponding to the preset candidate driver integrated circuit according to the preset identification information includes: Based on the initial electrical parameters after power-on, the rate of change of starting current and the end-to-end voltage fluctuation rate are calculated to obtain stability judgment indicators; Specifically, the initiation current change rate refers to the relative rate of change of current over time during the power-on transient, and the end-to-end voltage fluctuation rate refers to the amplitude fluctuation of the input and output voltages during the power-on phase. These two are combined to form a stability assessment index. The purpose of this index is to describe the power-on transient stability using quantitative dimensions, so as to use a unified standard to determine whether the system can proceed to the subsequent identification process. The specific operation involves high-speed sampling of the current and voltage at both ends during the power-on phase, extracting data during the initiation period according to a time window, performing noise reduction and synchronization, and then extracting the difference between the current peak value and the baseline, as well as the peak-to-peak value or standard deviation of the voltage. These features are then combined into a single index according to predetermined weights and recorded. This index transforms transient behavior into comparable numerical data, thereby enabling rapid identification of abnormal power-on responses and providing a basis for subsequent processing.

[0048] The stability determination index is compared with a preset stability threshold. If the stability determination index is less than the stability threshold, then the initial electrical parameters of the power-on meet the preset stability condition. Specifically, the stability threshold is a numerical limit set based on product specifications or historical samples. The preset stability condition means that when the stability assessment index falls within this limit, the power-on performance is considered normal. The purpose of this step is to determine whether to allow entry into the drive identification and performance testing process with clear admission criteria, preventing unstable samples from continuing to be tested. The specific method involves calculating the threshold from product design parameters or batch test data. The threshold can be a percentage tolerance or a statistical limit, and automatic comparison is implemented in the testing system. If the index is below the threshold, it is recorded as passed and the process continues; if it is above the threshold, a retest is triggered, the power-on level is reduced, or it is marked as abnormal and detailed sampling information is recorded. This comparison mechanism can eliminate unstable samples early, reduce false negatives, and improve subsequent identification efficiency and data reliability.

[0049] When the initial electrical parameters meet the preset stability conditions, the product model, power supply level and interface level of the LED flexible light strip are analyzed according to the preset identification information to obtain a set of candidate driver integrated circuits. Specifically, based on preset identification information, the product model, power supply level, and interface level are analyzed to generate a set of candidate driver integrated circuits. The power supply level refers to the supported voltage levels, and the interface level refers to the static voltage characteristics of the drive data lines. The purpose of this is to limit possible driver types to a manageable candidate set based on existing material information and electrical characteristics, thus providing direction for subsequent template testing. The specific method involves searching for possible driver models through product files and bills of materials, supplemented by non-invasive methods such as board identification, silkscreen inspection, pin count, and static level measurement. If necessary, simple low-voltage probe tests are performed on several candidates to confirm the number of interfaces and pull-up resistance. Finally, a candidate driver list sorted by priority is generated and written into the test task. This approach reduces the number of blind tests, lowers the risk of driver mismatch, and improves identification speed and security.

[0050] According to the preset drive signal template library, the drive signal template corresponding to each candidate driver integrated circuit in the candidate driver integrated circuit set is obtained as the drive signal template set.

[0051] Specifically, a drive signal template set is formed by selecting templates corresponding to the candidate drive set from a pre-set drive signal template library. This library consists of pre-saved timing, level, and frame format samples from different drivers. This step aims to provide a targeted signal set for exploratory drives, enabling rapid identification of the correct template through response observation. Operationally, matching templates are retrieved from the candidate list in the signal library, and each template is sequentially deployed under low brightness and short pulse constraints. Optical sensing and electrical sampling are used to synchronously record the response, and the data is archived according to the template. If abnormal electrical parameters are detected, the process stops and the abnormality details are recorded. This orderly template probing method efficiently confirms matching templates while ensuring the safety of the device under test, laying the foundation for accurate extraction of drive protocol parameters.

[0052] In an optional embodiment, the step of performing multi-condition performance testing on the LED flexible light strip according to preset performance test conditions corresponding to the driving protocol parameters, and obtaining the performance test results, includes: The protocol type, timing parameters, and level parameters are analyzed to determine the driving method of the flexible LED light strip; Specifically, the protocol type refers to the transmission method used for the control signal, such as single-wire serial, parallel transmission, or pulse width modulation controlled by duty cycle. Timing parameters refer to time characteristics such as bit width, pulse interval, and reset interval. Level parameters refer to logic high and low voltages, thresholds, and drive voltage levels. The purpose of analyzing this information is to confirm the basic response form of the LED strip to the control signal, thereby determining which driving method should be used to ensure correct driving. Under safe low-brightness conditions, an oscilloscope or logic analyzer is used to acquire the data line waveform, performing edge detection, bit rate calculation, and period statistics. Simultaneously, static high and low levels are measured, and resets or frame boundaries are detected. By analyzing the bit width, gap duration, and voltage threshold, the driving method can be determined. This clarifies the timing and level requirements of subsequent signals, avoiding abnormal responses or device damage caused by incorrect excitation.

[0053] Based on the driving method and preset identification information, a set of candidate driver integrated circuit models is determined; Specifically, the driving method is combined with preset identification information to determine the set of candidate driver IC models. The driving method reflects the control mechanism, while the preset identification information provides clues such as product model, power supply level, and interface type. A verifiable candidate set is established from possible driver types to reduce blind testing and improve identification efficiency. The process involves searching the bill of materials and device database for IC models that match known driving methods and interface levels, and cross-validating them by observing the board package, pin count, silkscreen and pull-up resistor configuration, as well as non-invasive measurements of the number of data lines and static level. Models meeting the criteria are then prioritized to form the candidate set. This screening method limits subsequent template testing to a reasonable set, thereby reducing the risk of mis-driving and accelerating the matching process.

[0054] Based on the driving method and the set of candidate driver integrated circuit models, performance test conditions are retrieved from a preset performance test condition library to obtain an initial performance test condition set. Specifically, the initial performance test condition set is obtained by retrieving performance test conditions corresponding to the driving mode and candidate drive set from a preset performance test condition library. The performance test condition library is a rule base that maps drive type, electrical level, and test matrix. Entries typically include power supply voltage level, duty cycle or brightness level, temperature and humidity cycling, and accelerated stress configuration. The purpose of the retrieval is to provide test templates that match the actual driving characteristics for subsequent performance tests, avoiding the use of irrelevant or overly aggressive stress schemes. In implementation, the driving mode and candidate device model are used as search keys to extract all matching or partially matching entries from the library. Conflicting parts between entries are parameterized to form an initial condition set covering different operating conditions. At the same time, the applicable prerequisites and reference sources for each entry are recorded for traceability.

[0055] Based on the timing parameters and level parameters, the initial performance test condition set is verified and filtered to obtain the preset performance test conditions; Specifically, the initial performance test condition set is verified and screened item by item using timing and level parameters to obtain preset performance test conditions. Then, multi-condition performance tests are performed according to these conditions, and the results are collected. The purpose of verification is to ensure that each test condition does not conflict with the actual drive parameters in terms of timing and level, and can apply sufficient stress within a safe range. Verification methods include comparing the voltage level, PWM frequency, and pulse width of each test item with the extracted timing and level limits one by one, correcting or eliminating items that do not match or exceed the tolerance range, and, if necessary, using staged or progressive loading for certain stress items. Finally, with the cooperation of an environmental chamber, programmable power supply, and data acquisition system, long-term cyclic tests are run in parallel or sequentially according to the screened conditions, recording multimodal data such as electrical curves, optical output, temperature distribution, and pixel changes in real time. The collected performance test results are used for subsequent defect identification, failure mode analysis, and lifetime estimation.

[0056] The LED flexible light strip is subjected to multi-condition performance tests according to the preset performance test conditions, and the performance test results are obtained.

[0057] Specifically, based on the preset performance test conditions, various operating conditions such as different voltages, currents, temperatures, humidity, operating durations, and drive modes are simulated. The purpose of performing tests according to the preset performance test conditions is to accelerate the exposure of potential electrical or optical defects in the LED strip under a controlled and repeatable environment, thereby completing reliability screening before the product enters actual application. During the process, the testing device automatically adjusts the power supply voltage and drive waveform according to the preset conditions, switches different loads and environmental parameters, and simultaneously records brightness changes, current fluctuations, heat distribution, and signal integrity in real time through multi-channel measurement devices such as photoelectric sensors, thermocouples, and sampling circuits. The entire process can last from several hours to hundreds of hours to ensure coverage of both short-term and long-term operating scenarios. Through this step, not only can performance test results containing optical, electrical, and thermal information be obtained, but also solid data support can be provided for subsequent defect detection and lifespan prediction, making the testing process more systematic and scientific.

[0058] In an alternative embodiment, please refer to Figure 2 The step of conducting multi-condition performance tests on the LED flexible light strip according to the preset performance test conditions, and obtaining the performance test results, includes: Based on the preset performance test conditions, the output parameters used for drive control are configured to obtain drive configuration parameters; Specifically, preset performance test conditions refer to items such as voltage levels, duty cycles, and temperature and humidity curves recorded in the library, as well as the output parameters of the performance test device, such as the power supply output voltage and current, the signal characteristics of the waveform generator, and the temperature and humidity control range of the environmental chamber. The drive configuration parameters obtained by mapping these two are the specific values ​​and control scripts that the device can directly execute. The abstract test intent is transformed into executable device instructions, which must meet stress coverage requirements while taking into account the safety constraints of the device under test. First, the performance condition items are analyzed and broken down into electrical sub-items and environmental sub-items. Then, the parameter correspondence is established between the programmable power supply, signal generator, environmental chamber, and test management system. The output voltage, current threshold, PWM frequency, pulse width, period, and temperature and humidity step sequence are set. Subsequently, calibration operation is performed to confirm that the output is consistent with the expectation and the calibration coefficient and device capability boundary are recorded. Through this configuration, the test device can implement multi-level drive in a precise and repeatable manner, ensuring the comparability between different test batches and reducing the result deviation caused by equipment errors.

[0059] Based on the driver configuration parameters, performance test conditions corresponding to different driver integrated circuit models are invoked to generate an initial performance test sequence covering the characteristics of multiple driver models. Specifically, the driver IC model set refers to the candidate IC models listed based on the bill of materials and driver method determination, and the initial performance test sequence refers to the execution script that arranges and parameterizes multiple performance conditions along a time axis. Multiple driver characteristics are considered simultaneously in a single test process, avoiding the need to design test sequences from scratch for each model, thereby improving test efficiency and ensuring comparability. First, entries compatible with each IC model are retrieved from the performance condition library using the driver configuration parameters as the key. Then, parameters are replaced and safety constraints are adjusted for each entry based on the power supply level and interface level of each IC. Different entries are then concatenated into a sequence according to functional and temporal requirements. If necessary, a segmented escalation strategy is adopted for high-stress segments, and observation points and rollback strategies are defined for each segment. Finally, the sequence is sent to the test scheduler and pre-run on virtual or equivalent loads for verification. By generating such a comprehensive initial sequence, the responses of different ICs under the same and different operating conditions can be systematically examined, facilitating subsequent attribution analysis and process standardization.

[0060] According to the initial performance test sequence, drive signals are sequentially applied to the flexible LED light strip to form performance conditions with different power levels, and the performance execution results are obtained. Specifically, the drive signals are applied to the LED strip segment by segment according to the initial performance test sequence to form performance conditions at different power levels, and the performance execution results are recorded. Key terms include performance conditions, which refer to the operating status under different power, duty cycles, or environmental conditions; and performance execution results, which are the multi-channel raw data such as current, voltage, brightness, spectrum, temperature, and images collected in real time during the execution process. The goal of this step is to promote the manifestation of different failure mechanisms within an observable timescale through controllable power steps and environmental cycles, and to fully collect raw evidence for analysis. The implementation method is to execute the test sequence in segments, setting the target voltage or brightness level, duration, and sampling frequency for each segment. A programmable power supply and signal generator are used to output the corresponding drive, while an integrating sphere or photoelectric sensor is used for photometry, a spectrometer is used to record color shift, a thermal imager or temperature sensor is used to record the thermal field, a high-speed camera is used to record pixel-level changes, and a high-speed sampling channel is used to record transient electrical waveforms. All channels are synchronized with a unified trigger signal timestamp. When an electrical abnormality or safety threshold is encountered, the preset rollback or interruption process is followed and the event is recorded. By using this segmented loading and synchronous acquisition, detailed execution data across power levels can be obtained, providing high-quality time-series evidence for subsequent multidimensional analysis.

[0061] Based on the performance execution results, a multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performance conditions is obtained. Specifically, a multi-dimensional performance response dataset is constructed from performance execution results, covering the performance of different driver IC models under various performance conditions. Key components of this dataset include time-series electrical data, optical output curves, color coordinates and spectral variations, temperature field evolution, and pixel-level failure maps. These dimensions require labeling with the driver IC model and corresponding operating conditions for cross-sectional comparison. The purpose of this dataset is to transform scattered raw samples into a structured and analyzable information body to identify patterns, calculate degradation rates, and perform modeling. Implementation details include preprocessing and aligning the raw channels, denoising and calibrating optical measurements, calculating derived metrics such as lumen maintenance, color difference, flicker index, and current drift rate, and merging these metrics with sample metadata into a unified data warehouse. When necessary, time-series segmentation or windowing statistics are used to extract short-term and long-term features, while generating visualizations such as failure heatmaps, degradation curves, and spectral evolution diagrams. This dataset provides input for comparative analysis and machine learning, clarifying the impact of different driver ICs and operating conditions on lifetime and reliability, thereby supporting targeted improvements and risk assessments.

[0062] Based on the multidimensional performance response dataset, the response characteristics of different driver integrated circuit models are compared and analyzed, and the test results under different performance conditions are classified and processed to obtain the performance test results.

[0063] Specifically, the constructed multidimensional performance response dataset undergoes comparative analysis and classification processing, summarizing and outputting the final performance test results. The comparative analysis includes horizontal comparison of key performance indicators among different driver IC models and clustering of sample behavior under various performance conditions. The classification process categorizes the response results into qualified, boundary, or failure types and labels the failure modes. By extracting judgment conclusions from a large number of quantitative indicators and revealing driver-related or condition-related weaknesses, decision-making and improvement are facilitated. Basic indicators are first filtered using rule-based thresholds, then statistical tests and unsupervised algorithms are combined to cluster and identify typical failure modes. Significantly abnormal samples undergo manual review and fault location analysis. Finally, a standardized report containing various indicators, fault classifications, timestamps, and recommended actions is generated, and the results are written back to the quality management system for traceability and supply chain feedback. This systematic comparative and classification processing transforms massive amounts of performance data into actionable quality conclusions, supporting product verification, production adjustments, and subsequent reliability improvements.

[0064] In an optional embodiment, obtaining the multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performance conditions based on the performance execution results includes: Based on the performance results, the operating times of the LED flexible light strip under different driver integrated circuit models and different performance conditions are marked to obtain operating condition index information; Specifically, performance execution results refer to the raw time-series records collected by each channel during the test, including voltage and current waveforms, optical measurements, temperature readings, and camera images. The operating condition index information is an index mapping established on the timeline between each running segment and its corresponding driver integrated circuit model and performance condition. Marking the running time is done to segment the dispersed sampling stream into segments with clear experimental semantics, facilitating subsequent retrieval and comparison by operating condition. Operationally, by parsing the test scheduler logs and device timestamps, the start and end times of each issued test segment are bound to the corresponding device model, power supply level, and environmental conditions, generating uniquely identified operating condition index records and writing them to the metadata database. Simultaneously, verification information is saved for each index to support playback. After completing this work, the raw data segment under any operating condition can be quickly located, facilitating traceability, sampling verification, and batch export by conditions, providing a reliable retrieval foundation for subsequent multimodal analysis.

[0065] Based on the operating condition index information, obtain the electrical operating data, optical operating data, thermal operating data and mechanical response data of the LED flexible light strip under various operating conditions; Specifically, electrical operating data typically includes transient and steady-state current, voltage, and power curves; optical operating data includes luminous flux, chromaticity coordinates, and spectral distribution; thermal operating data consists of time series of surface and local temperature fields; mechanical stress conditions refer to specifications such as bending radius, bending speed, vibration frequency, and amplitude; brightness change rate and current change rate refer to the relative rates of change of brightness and current under mechanical cycles or transients, respectively; and mechanical response data is a structured record of these change rates and their statistical characteristics. The starting point for characterizing mechanical effects with these quantities is to capture the differences in electro-optical behavior of flexible materials under stress using quantitative indicators. The purpose of extracting various types of data by index is to aggregate multimodal observations under the same operating condition into a single investigation unit; data corresponding to the time window are retrieved from each acquisition channel according to the operating condition index; if necessary, channels with different sampling rates are resampled and units are converted; and the original readings are corrected using sensor calibration coefficients. After correction, the electrical, optical, and thermal data are packaged into an operating condition-level data package in a unified format and stored in the database. This ensures that the comparative analysis is based on a consistent and verified measurement standard, making cross-modal indices under the same working conditions comparable and facilitating subsequent statistics and modeling.

[0066] The electrical operating data, optical operating data, thermal operating data, and mechanical response data are time-aligned and spatially calibrated to obtain the multidimensional performance response dataset.

[0067] Specifically, time alignment refers to correcting and synchronizing the time axes from different acquisition devices according to a unified benchmark. Spatial calibration refers to mapping camera pixel or sensor coordinates to the physical location or pixel index of the LED strip. The multidimensional performance response dataset is a fused data table organized by time and spatial coordinates. The starting point for fusing these data is to establish a multimodal view that can accurately correspond in time and space. Implementation details include using a unified trigger signal or network time protocol as a time benchmark, performing delay compensation and resampling for each channel, using a calibration board, reference light point, or reference marks on the strip to perform homography transformation from camera to physical coordinates and mapping pixels to specific LED indices, while performing spatial interpolation calibration for thermal images and contact temperature points. Finally, the sampled values ​​of all channels are merged into a tabular record according to timestamps and LED indices and metadata is saved. The resulting multidimensional performance response dataset can be used for pixel-level fault location, temporal degradation analysis, and cross-modal anomaly detection, thereby significantly improving the accuracy and interpretability of defect judgment.

[0068] In an optional embodiment, inputting the performance test results into a pre-trained multi-dimensional defect detection model to obtain defect detection results includes: Based on the performance test results, the electrical, optical, and mechanical parameters of different driver integrated circuit models are analyzed to obtain the target detection dataset corresponding to each model. Specifically, using the driver integrated circuit model as a group index, the previously obtained performance test results are analyzed and structured to form target detection datasets corresponding to each model. The electrical parameters mainly come from the steady-state and transient characteristics of the electrical operation data, such as driving current, voltage waveform, power factor and power loss. Optical parameters include brightness, color temperature, chromaticity coordinates and luminous flux. The mechanical related parts are combined with the quantified results of bending response reflected by the mechanical response data and merged, so that the multi-dimensional observations of the same model under different performance conditions can be unified into the same data caliber. The purpose of this approach is to address the incomparability issues arising from differences in the driving characteristics and acquisition links of different models. Since the test sequence covers the driving characteristics of multiple models and acquires multi-dimensional responses under different performance conditions, directly using the original observations for anomaly judgment can easily lead to misinterpreting differences in sampling scale or units as electro-optical anomalies. Therefore, it is necessary to map the original multimodal observations into standardized quantitative indicators at the model dimension, and then form a structured dataset. This provides a unified input basis for subsequent statistical analysis of indicators such as current fluctuations, voltage drop drift, power consumption changes, brightness attenuation, chromaticity shift, flicker amplitude, and bending response. Electrical, optical, and mechanical operation data of the same model under various operating conditions are extracted and aligned according to the operating condition time window. Signal processing is used to extract features such as steady-state mean, ripple amplitude, and transient fluctuations from voltage waveforms and current curves. The original readings are corrected by calibration coefficients. If necessary, channels with different sampling rates are resampled and units are converted and normalized. Finally, the parsed fields are packaged into a model-level data table in a unified format and stored in the database to obtain the target detection dataset. This can significantly improve the horizontal comparability and consistency between different models, reduce the interference of original sampling differences on defect judgment, make the input of subsequent multi-dimensional defect detection models clearer and more stable, and make the calculation of abnormal indicators more reliable.

[0069] Based on the target detection datasets described above, statistical calculations are performed on current fluctuations, voltage drop drift, and power consumption changes to obtain electrical anomaly indicators; Specifically, current fluctuations reflect differences in the stability of the drive circuit, voltage drop drift reflects the degradation of the conduction path or material properties, and power consumption changes reveal the degradation of power utilization efficiency. After extracting these indicators from the target detection dataset, the long-term stability of electrical performance can be evaluated by calculating the mean, variance, offset, and trend curves. This process can reveal potential electrical hazards, such as localized overcurrent, increased conduction losses, or poor power supply adaptation, and provide quantitative evidence for distinguishing between normal performance and abnormal degradation.

[0070] Based on the target detection datasets described above, statistical calculations are performed on brightness attenuation, chromaticity shift, and flicker amplitude to obtain optical anomaly indicators. Specifically, brightness decay reflects a decrease in luminous efficacy, chromaticity shift indicates a decrease in luminous consistency and color reproduction, while flicker amplitude is related to the driving waveform and human eye comfort. By statistically analyzing target detection datasets, attenuation curve fitting, chromaticity drift vector calculation, and frequency domain analysis methods can be used to quantify optical anomalies. This multi-faceted characterization approach helps to quickly identify optical performance degradation caused by packaging performance issues, phosphor decay, or driving circuit instability, thus providing early warning in product manufacturing inspections or lifespan assessments.

[0071] Based on the target detection datasets described above, the temperature rise gradient, hot spot location, and bending response are analyzed to obtain mechanical anomaly indicators; Specifically, the temperature gradient reflects the difference in heat dissipation path efficiency, hotspot locations reveal local heat dissipation bottlenecks, and the bending response reflects the stability of materials and solder joints under mechanical stress. The analysis process typically includes spatial interpolation calculations of infrared thermography or contact temperature sampling data, as well as joint modeling of electrical and optical parameters under bending or vibration tests to quantify the combined effects of thermo-mechanical coupling and mechanical stress. These data can effectively reveal potential structural failure risks, such as solder joint cracking, material fatigue, or localized overheating.

[0072] The electrical anomaly index, optical anomaly index, and mechanical anomaly index are input into the multi-dimensional defect detection model to obtain the corresponding initial defect detection results; Specifically, the three types of quantified anomaly representations are used as joint inputs within the same discrimination space, allowing the model to utilize cross-dimensional coupling relationships to complete a more reliable initial defect judgment. The initial defect detection result is the defect candidate conclusion and its confidence characterization output by the pre-trained multi-dimensional defect detection model under the current driver integrated circuit model and current performance conditions. For example, it provides the defect tendency dominated by electrical anomalies, optical anomalies, or mechanical anomalies, and can correspond to candidate types such as overcurrent risk, flicker anomalies, local overheating-related failures, or connection reliability risks caused by bending. First, the three types of abnormal indicators are aligned and structured according to the driver integrated circuit model, so that indicators with different sampling frequencies and different dimensions are mapped to a unified input format. For example, current fluctuations, voltage drop drift, and power consumption changes are concatenated with brightness attenuation, chromaticity shift, flicker amplitude, temperature rise gradient, hot spot location, and bending response into feature vectors according to preset dimensions, or organized into sequence features according to time slices. Then, the necessary normalization and scaling are completed in conjunction with the calibration database to avoid one type of indicator dominating the judgment due to its larger dimension. Subsequently, the joint features are input into the pre-trained model to perform inference operations. Based on the correspondence between the abnormal combination patterns and defect types learned in the training phase, the model outputs the initial defect detection results and simultaneously provides the confidence or risk score of each defect candidate. Compared to judgment methods based on single index thresholds, the model can utilize cross-dimensional features such as the co-occurrence of current fluctuations and flicker amplitude, the linkage between voltage drop drift and temperature rise gradient, and the correlation between bending response and brightness unevenness to suppress random fluctuation interference caused by complex operating conditions, thereby reducing misjudgments and omissions, and providing a more stable and comparable initial judgment basis for subsequent defect classification and summarization by model and operating condition.

[0073] Based on the initial defect detection results, the defect types of different driver integrated circuit models and under different performance conditions are classified and summarized to obtain the defect detection results.

[0074] Specifically, after defect identification, it is necessary to further classify and summarize the defect types of different driver integrated circuit models and under different performance conditions. During the classification process, the results can be categorized into electrical, optical, thermal, or mechanical types based on defect characteristics, and further refined into specific types such as overcurrent failure, uneven brightness, localized overheating, or solder joint fatigue. Summarizing the results not only helps to present the overall defect distribution but also reveals the vulnerability characteristics of different driver models under specific operating conditions, providing a systematic reference for subsequent quality improvement, design optimization, and lifespan prediction.

[0075] In an optional embodiment, the step of analyzing the temperature rise gradient, hotspot locations, and bending response based on each of the target detection datasets to obtain mechanical anomaly indicators includes: Based on the target detection datasets described above, the temperature distribution data at different sampling times are extracted and aligned to obtain a temperature distribution sequence; Specifically, when extracting and aligning temperature distribution data from different sampling times to obtain a temperature distribution sequence, the temperature distribution data is understood as a two-dimensional temperature field snapshot recorded at each sampling time by infrared thermal imaging, arrayed thermocouples, or contact temperature point arrays. The sampling time refers to the timestamp of each recording. By organizing these snapshots into a temperature distribution sequence in chronological order, the thermal evolution process can be observed, providing a basis for subsequent feature extraction. First, the raw thermal image is exported from the acquisition device, and emissivity and sensor calibration are performed. A unified time reference or trigger signal is used to align each frame in time. If the sampling rates are inconsistent, the data is resampled and interpolated. Each frame is mapped to the physical coordinate system of the light strip and saved as a serialized data structure. Simultaneously, acquisition metadata needs to be recorded for backtracking. The temperature distribution sequence obtained through this processing facilitates the identification of transient thermal events, quantifies the evolution of thermal response over time, and provides comparable and traceable input for gradient and hotspot analysis.

[0076] Based on the temperature distribution sequence, the rate of temperature change in each region is calculated to obtain the temperature rise gradient index. Specifically, when calculating the temperature change rate of each region based on the temperature distribution sequence to obtain the temperature rise gradient index, each region should be defined as several thermal units divided according to the LED index or a preset grid. The temperature change rate can be represented by the inter-frame difference or the time-domain derivative. Transforming the change of local temperature over time into a gradient index can characterize the heat conduction efficiency and the strength of local heat accumulation. The specific approach involves first dividing the temperature distribution sequence into regions and performing spatial filtering to reduce noise. Then, for each region, the instantaneous change rate and average rise rate are calculated according to a selected time window. Further calculation of the temperature difference between neighboring regions yields the spatial gradient, which is then normalized to the temperature rise gradient index. If necessary, multiple time-scale windows are used to distinguish between short-term abrupt changes and long-term accumulation. The obtained temperature rise gradient index can be used to identify abnormal heat dissipation paths or areas of power concentration, providing a quantitative basis for locating design defects or insufficient thermal management.

[0077] Based on the temperature distribution sequence, local extreme points in the temperature distribution are located and tracked to obtain hotspot location indicators. Specifically, when locating and tracking local extrema in a temperature distribution sequence to obtain hotspot location indices, local extrema are pixels or thermal units with significantly higher temperatures than their surroundings within a spatial neighborhood, described by their location coordinates and temperature peaks. Tracking these extrema over time reveals the generation, migration, and persistence characteristics of hotspots. Local extrema detection is performed on each frame, and noise spurious peaks are removed using a salience threshold. For each candidate peak, the centroid, peak temperature, and salience are calculated. Then, methods such as nearest neighbor or Kalman filtering are used to establish the peak trajectory over time, recording the frequency of occurrence, duration, and location offset. The hotspots are then mapped back to the physical LED index. The resulting hotspot location indices not only pinpoint the specific location of high-temperature anomalies but also reflect the stability and evolution trend of the hotspots, facilitating the correlation analysis between hotspots and electrical or structural problems.

[0078] Based on the target detection datasets described above, the bending curvature, the number of bending cycles, and the corresponding current change rate and brightness change rate are correlated and calculated to obtain the bending response index; Specifically, based on the target detection dataset, the bending curvature, the number of bending cycles, and the corresponding rates of change in current and brightness are correlated to calculate the bending response index. The bending curvature is the reciprocal of the radius of curvature, the number of bending cycles refers to the number of repeated bends, and the rates of change in current and brightness represent the relative changes in electrical and optical quantities within each bending event or cycle, respectively. Combining these quantities allows for the quantification of the impact of mechanical fatigue on electro-optical performance. The output of the curvature sensor or encoder is synchronously recorded on a bending table or winding platform, and each bending event is aligned with the current and optical samples using a unified trigger and timestamp. The instantaneous rate of change is calculated for each cycle, and trend analysis and fracture point detection are performed on the cycle sequence. Then, statistical quantities or degradation models are used to obtain the bending response index. This index can reveal the degradation rate and sudden failure mode under different curvatures and cycle intensities, providing direct quantitative support for evaluating the durability of flexible structures and optimizing materials or welding processes.

[0079] Based on the temperature rise gradient index, hot spot location index, and bending response index, the detection results are comprehensively compared and correlated to obtain the mechanical anomaly index.

[0080] Specifically, when comprehensively comparing and analyzing temperature rise gradient indicators, hotspot location indicators, and bending response indicators to obtain mechanical anomaly indicators, this comprehensive comparison and correlation analysis involves jointly examining different categories of indicators in time and space to identify correlations and causal clues. By analyzing thermal and mechanical indicators side-by-side, it is possible to distinguish between simple thermal management problems and heat dissipation deterioration induced by mechanical stress. In practice, various indicators are first standardized and spliced ​​according to time windows or operating conditions. Correlation coefficients, cross-lags, or causal inference methods are used to assess the sequential relationships between indicators on the indicator matrix. The samples are then classified into typical anomaly patterns, and manual verification is performed using original images and event logs to confirm the judgment. The obtained mechanical anomaly indicators can be used to clarify the type of anomaly and locate its possible root causes, thereby supporting subsequent repair decisions or design process improvements.

[0081] In an alternative embodiment, please refer to Figure 3 The performance testing device 1, as the core execution unit of the entire testing process, is used to implement controlled power-on and multi-condition loading of the LED flexible light strip 2. Specifically, by controlling the power-on of the LED flexible light strip 2 and applying a corresponding load, the performance testing device 1 makes the LED flexible light strip 2 operate under conditions close to actual use, thereby simultaneously collecting its electrical operation data, optical operation data, thermal operation data, and mechanical response data during operation. Specifically, the electrical operation data reflects the current, voltage, and power consumption characteristics of the driver integrated circuit under different operating states; the optical operation data reflects the brightness, color temperature, chromaticity, and luminous uniformity of the light strip; and the thermal operation data... The data reflects the temperature rise and heat dissipation characteristics of the LED strip under continuous power supply and load conditions, while the mechanical response data reflects the structural and performance changes of the flexible LED strip 2 under mechanical actions such as bending and stretching. The above multi-dimensional performance test results are uniformly sent to a multi-dimensional defect detection model for analysis and processing, and then output the corresponding electrical anomaly indicators, optical anomaly indicators, and mechanical anomaly indicators. This enables comprehensive detection and defect judgment of the flexible LED strip 2 in terms of electrical, optical, and mechanical multi-dimensional performance. This intuitively demonstrates the overall technical approach of this solution, which uses the performance testing device 1 and the flexible LED strip 2 to work together to complete multi-source data acquisition and support subsequent multi-dimensional defect identification.

[0082] Example 2 In addition, combined Figure 1 The defect detection method for LED flexible light strips described in this embodiment of the invention can be implemented by an LED flexible light strip defect detection system. Figure 4 A schematic diagram of the hardware structure of the defect detection system for flexible LED light strips provided in an embodiment of the present invention is shown.

[0083] A defect detection system for flexible LED light strips may include a processor and a memory storing computer program instructions.

[0084] Specifically, the processor may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement embodiments of the present invention.

[0085] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0086] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient media, such as modulated communication signals and carrier waves.

[0087] The processor reads and executes computer program instructions stored in the memory to implement any of the defect detection methods for flexible LED light strips in the above embodiments.

[0088] In one example, the defect detection system for flexible LED light strips may also include a communication interface and a bus. For example, Figure 4 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 410 and complete communication with each other.

[0089] The communication interface is mainly used to enable communication between various modules, devices, units and / or equipment in the embodiments of the present invention.

[0090] A bus, including hardware, software, or both, couples the components of a defect detection system for a flexible LED light strip together. For example, and not limitingly, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, a bus may include one or more buses. While specific buses are described and illustrated in embodiments of the invention, the invention contemplates any suitable bus or interconnect.

[0091] In summary, the embodiments of the present invention provide a defect detection method and system for flexible LED light strips.

[0092] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0093] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0095] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0096] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0097] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0098] The above description is merely a specific embodiment of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A defect detection method for flexible LED light strips, characterized in that, The method includes: Obtain the preset identification information of the LED flexible light strip to be tested; Based on the power supply safety parameters corresponding to the preset identification information, the LED flexible light strip is powered on in a controlled manner to obtain the initial power-on electrical parameters. When the initial electrical parameters meet the preset stability conditions, drive detection is performed on the LED flexible light strip to obtain the drive protocol parameters of the target driver integrated circuit; Based on the preset performance test conditions corresponding to the driving protocol parameters, the LED flexible light strip is subjected to multi-dimensional performance tests to obtain the performance test results. The performance test results are input into the pre-trained multi-dimensional defect detection model to obtain the defect detection results.

2. The defect detection method for LED flexible light strips according to claim 1, characterized in that, The controlled power-on of the LED flexible light strip based on the power supply safety parameters corresponding to the preset identification information, and the acquisition of initial power-on electrical parameters, include: Based on the preset identification information, determine the rated power supply voltage and rated operating current corresponding to the preset identification information to obtain the power supply reference parameters; Based on the power supply reference parameters, the output voltage of the power supply and the current threshold of the current limiting circuit are configured to obtain the power supply control parameters; According to the power supply control parameters, the control power supply gradually applies voltage to the LED flexible light strip to execute a controlled power-on process. During the controlled power-on process, the starting current and the voltage across the two ends of the LED flexible light strip are used as electrical measurement data for power-on. Based on the electrical measurement data obtained upon power-on, the peak starting current and end-to-end voltage drop are calculated to obtain the initial electrical parameters upon power-on.

3. The defect detection method for LED flexible light strips according to claim 1, characterized in that, When the initial power-on electrical parameters meet the preset stability conditions, the LED flexible light strip is driven and detected to obtain the driving protocol parameters of the target driving integrated circuit, including: When the initial electrical parameters of the power-on meet the preset stability conditions, the candidate driver integrated circuit corresponding to the preset identification information is determined according to the preset identification information; Based on the drive signal template set corresponding to the candidate driver integrated circuit, test drive signals are applied to the LED flexible light strip one by one, and the optical response data and electrical response data of the LED flexible light strip during the test drive process are obtained to obtain the test response dataset. Based on the trial response dataset, the response consistency of different driving signal templates in the driving signal template set is compared to determine the target driving signal template that matches the LED flexible light strip. Based on the target drive signal template, the protocol type, timing parameters, level parameters, and data format parameters are extracted to obtain the drive protocol parameters.

4. The defect detection method for LED flexible light strips according to claim 3, characterized in that, When the initial power-on electrical parameters meet the preset stability conditions, obtaining the drive signal template set corresponding to the preset candidate driver integrated circuit according to the preset identification information includes: Based on the initial electrical parameters after power-on, the rate of change of starting current and the end-to-end voltage fluctuation rate are calculated to obtain stability judgment indicators; The stability determination index is compared with a preset stability threshold. If the stability determination index is less than the stability threshold, then the initial electrical parameters of the power-on meet the preset stability condition. When the initial electrical parameters meet the preset stability conditions, the product model, power supply level and interface level of the LED flexible light strip are analyzed according to the preset identification information to obtain a set of candidate driver integrated circuits. According to the preset drive signal template library, the drive signal template corresponding to each candidate driver integrated circuit in the candidate driver integrated circuit set is obtained as the drive signal template set.

5. The defect detection method for LED flexible light strips according to claim 3, characterized in that, The multi-dimensional performance test of the LED flexible light strip is performed according to the preset performance test conditions corresponding to the driving protocol parameters, and the performance test results are as follows: The protocol type, timing parameters, and level parameters are analyzed to determine the driving method of the flexible LED light strip; Based on the driving method and preset identification information, a set of candidate driver integrated circuit models is determined; Based on the driving method and the set of candidate driver integrated circuit models, performance test conditions are retrieved from a preset performance test condition library to obtain an initial performance test condition set. Based on the timing parameters and level parameters, the initial performance test condition set is verified and filtered to obtain the preset performance test conditions; Based on the preset performance test conditions, the LED flexible light strip is subjected to multi-dimensional performance tests to obtain performance test results.

6. The defect detection method for LED flexible light strips according to claim 5, characterized in that, The multi-dimensional performance test of the LED flexible light strip according to the preset performance test conditions, and the resulting performance test results include: Based on the preset performance test conditions, the output parameters used for drive control are configured to obtain drive configuration parameters; Based on the driver configuration parameters, performance test conditions corresponding to different driver integrated circuit models are invoked to generate an initial performance test sequence covering the characteristics of multiple driver models. According to the initial performance test sequence, drive signals are sequentially applied to the flexible LED light strip to form tests at different power levels, and performance execution results are obtained. Based on the performance execution results, obtain a multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performances; Based on the multidimensional performance response dataset, the response characteristics of different driver integrated circuit models are compared and analyzed, and the test results under different performance conditions are classified and processed to obtain the performance test results.

7. The defect detection method for LED flexible light strips according to claim 6, characterized in that, The step of obtaining the multi-dimensional performance response dataset of the LED flexible light strip under different driver integrated circuit models and different performance conditions based on the performance execution results includes: Based on the performance results, the operating times of the LED flexible light strip under different driver integrated circuit models and different performance conditions are marked to obtain operating condition index information; Based on the operating condition index information, obtain the electrical operating data, optical operating data, thermal operating data and mechanical response data of the LED flexible light strip under various operating conditions; The electrical operating data, optical operating data, thermal operating data, and mechanical response data are time-aligned and spatially calibrated to obtain the multidimensional performance response dataset.

8. The defect detection method for LED flexible light strips according to any one of claims 1-7, characterized in that, The step of inputting the performance test results into the pre-trained multi-dimensional defect detection model to obtain defect detection results includes: Based on the performance test results, the electrical, optical, and mechanical parameters of different driver integrated circuit models are analyzed to obtain the target detection dataset corresponding to each model. Based on the target detection datasets described above, statistical calculations are performed on current fluctuations, voltage drop drift, and power consumption changes to obtain electrical anomaly indicators; Based on the target detection datasets described above, statistical calculations are performed on brightness attenuation, chromaticity shift, and flicker amplitude to obtain optical anomaly indicators. Based on the target detection datasets described above, the temperature rise gradient, hot spot location, and bending response are analyzed to obtain mechanical anomaly indicators; The electrical anomaly index, optical anomaly index, and mechanical anomaly index are input into the multi-dimensional defect detection model to obtain the corresponding initial defect detection results; Based on the initial defect detection results, the defect types of different driver integrated circuit models and under different performance conditions are classified and summarized to obtain the defect detection results.

9. The defect detection method for LED flexible light strips according to claim 8, characterized in that, Based on the target detection datasets, the temperature rise gradient, hotspot locations, and bending response are analyzed to obtain mechanical anomaly indicators, including: Based on the target detection datasets described above, the temperature distribution data at different sampling times are extracted and aligned to obtain a temperature distribution sequence; Based on the temperature distribution sequence, the rate of temperature change in each region is calculated to obtain the temperature rise gradient index. Based on the temperature distribution sequence, local extreme points in the temperature distribution are located and tracked to obtain hotspot location indicators. Based on the target detection datasets described above, the bending curvature, the number of bending cycles, and the corresponding current change rate and brightness change rate are correlated and calculated to obtain the bending response index; Based on the temperature rise gradient index, hot spot location index, and bending response index, the detection results are comprehensively compared and correlated to obtain the mechanical anomaly index.

10. A defect detection system for flexible LED light strips, characterized in that, include: At least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method as described in any one of claims 1-9.

Images