A detection method for a phosphor-free multi-primary color LED light source module

Abstract

The present invention relates to the field of LED detection technology, and proposes a detection method for a phosphor-free multi-primary color LED light source module, which specifically includes: preparing detection equipment and building a detection environment; module connection and initialization, connecting and initializing the LED light source module and the detection equipment; full performance testing and life tracking of the LED light source module, obtaining a diffraction spectrum through a high-resolution camera, and using a photoelectric test instrument to measure the key parameters of the module, including luminous flux, chromaticity, color coordinates and temperature, and recording the actual service life and failure data of the module; machine learning feature recognition, establishing a reliability analysis and performance prediction model for the module; stability and reliability evaluation, detecting module performance indicators according to color rendering index, reliability and model, judging whether the module meets the standard based on the evaluation results, and proposing optimization suggestions. The present invention can improve the quality and reliability of LED light source modules, and provides an effective detection method for the industry.

Description

Technical Field

[0001] The invention relates to the technical field of LED detection, and in particular to a detection method for a fluorescent powder-free multi-primary color LED light source module. Background Art

[0002] With the rapid development of LED technology, LED lighting products have been widely used in various fields, including home lighting, commercial lighting and special lighting. Among them, multi-primary color LED light source modules are favored by the market because they can provide richer and more accurate color performance. Traditional multi-primary color LED light source modules often use phosphors to convert the light emitted by LEDs to achieve the desired color effect. However, the use of phosphors not only increases the manufacturing cost, but also the performance degradation of phosphors after long-term use will affect the color stability and life of the light source.

[0003] In recent years, phosphor-free multi-primary color LED light source modules have gradually become a research hotspot. This type of module produces the required light color by directly combining LED chips of different colors, thus avoiding the problems caused by phosphors. The performance testing and evaluation of phosphor-free multi-primary color LED light source modules has become a new challenge. Traditional testing methods mainly focus on the performance testing of single-color LEDs, but are unable to evaluate the color mixing effect, primary color accuracy, and long-term operation stability of multi-primary color LED modules. Summary of the invention

[0004] In order to solve the above-mentioned technical problems, the present invention provides a method for detecting a phosphor-free multi-primary color LED light source module, comprising the following steps:

[0005] S1: Preparation work, ensuring that the testing equipment and environment meet the test requirements;

[0006] S2: Module connection and initialization, check and connect the LED light source module and the detection equipment, and test the connection stability;

[0007] S3: Full performance evaluation and life tracking of LED light source modules. A high-resolution camera is used to obtain the diffraction spectrum and the spectral distribution of the visible light band of the light source. Photoelectric testing instruments are used to measure the module's luminous flux, chromaticity, color coordinates, and temperature parameters. The life data and number of failed modules of the LED light source module in actual use are also recorded.

[0008] S4: Machine learning feature recognition, establishing reliability analysis and prediction model of LED light source module;

[0009] S5: Stability and reliability assessment, using the trained model and reliability curves for factor analysis and life prediction;

[0010] S6: Based on the evaluation results, the performance of the LED light source module is evaluated, and the evaluation results are compared with the preset performance indicators and life requirements to determine whether the module meets the standards.

[0011] Furthermore, the step S1 specifically includes:

[0012] The specific steps of the preparation work include equipment inspection and calibration, so that the indoor test brightness range is 800-1200cd / m², the outdoor test brightness is 5000-6000cd / m², the color temperature range is between 5000K and 6500K, and the photoelectric test instrument is calibrated;

[0013] The sensor equipment used for detection includes illuminance meter, spectrometer, colorimeter, temperature sensor, timer and infrared thermal imager.

[0014] Furthermore, step S2 specifically includes:

[0015] S21. The specific steps of module connection and initialization include preparing connection cables and interfaces, connecting modules and devices, and performing test connection and initialization to verify the connection effect and ensure that it is in good working condition;

[0016] S22, placing the sensor at a suitable location so as to capture various parameters of the light source;

[0017] S23. Use a transmission diffraction grating and a slit to make a light splitting attachment. By strictly controlling the width and spacing of the slits, the diffraction angle of the light is manipulated to achieve the dispersion of visible light radiated by the light source. After diffraction, the originally mixed visible light is accurately separated into monochromatic light of different wavelengths. A high-resolution camera is set behind the light splitting attachment. By adjusting the position and parameters of the camera, the spectrum image after diffraction is captured.

[0018] S24. Use professional photoelectric testing instruments to measure other key parameters of the module, including luminous flux, chromaticity, color coordinates and temperature;

[0019] S25. Check the initialization status by sending a simple test signal to the LED light source module and observing its response.

[0020] Furthermore, the step S3 specifically includes:

[0021] Use illuminance meters, spectrometers, colorimeters, temperature sensors and infrared thermal imagers to record the time series of the luminous flux, chromaticity and color coordinate parameters of the LED light source module, and check the data integrity and accuracy. The sensor is installed in a position that can accurately reflect the environmental conditions to be tested. After installation, the sensor is calibrated and measured multiple times at different spatial positions to ensure that the output data is accurate and reliable. A high-resolution camera is then used to accurately capture the diffraction spectrum to obtain the spectral distribution of the light source in the visible light band.

[0022] Furthermore, the step S4 specifically includes:

[0023] Machine learning technology is used to establish a factor analysis and prediction model. The machine learning method is used to take temperature, humidity and working hours as model inputs, and performance indicators luminous flux, chromaticity, brightness attenuation, and color shift as model outputs. Through the analysis of a large amount of test data, the key factors affecting the performance of LED light source modules are identified, and a prediction model is established based on these factors to predict module performance and the life of the LED light source module.

[0024] Furthermore, the step S5 specifically includes:

[0025] S51, using the trained machine learning model, by analyzing the weights of each input feature in the model, determining significant features that affect the performance of the LED light source module;

[0026] S52, sorting the collected life data, counting the number of modules that fail in different time periods, and calculating the survival rate of the modules at different time points based on the failure data;

[0027] S53, the Weibull distribution is used for reliability assessment, which mainly includes shape parameter k, scale parameter λ and location parameter θ. In reliability analysis, θ=0 is assumed, and the shape parameter k and scale parameter λ are mainly concerned;

[0028] S54, using the maximum likelihood estimation method, estimate the shape parameter k and scale parameter λ of the Weibull distribution;

[0029] S55. Draw a reliability curve of the light source module according to the estimated parameters k and λ. The horizontal axis of the curve represents time t, and the vertical axis represents reliability R(t). The reliability R(t) is calculated by the cumulative distribution function of the Weibull distribution, and the formula is:

[0030] R(t)=exp[-(t / λ) k ];

[0031] S56. Calculate the average life and median life according to Weibull distribution, compare and analyze the life prediction obtained by machine learning and reliability analysis with the actual service life, and evaluate the reliability and stability of the LED light source module;

[0032] S57, sorting the features according to weights to identify key factors affecting the performance of the LED light source module;

[0033] S58, comparing the predicted performance parameters with the actual measured values ​​to verify the accuracy of the model, and using the trained BP neural network model to predict the service life of the LED light source module based on the input key factors;

[0034] S59. Based on the analysis results of the above steps, the overall performance of the LED light source module is evaluated to determine whether it meets the preset performance indicators and industry standards.

[0035] Furthermore, the step S6 specifically includes:

[0036] S61. Calculate the color rendering index. According to the color rendering degree of the color sample specified by the International Commission on Illumination, compare the spectral distribution of the light source to be tested with the spectral distribution of the reference light source, and calculate the color difference value ΔE of the color sample under the two light sources. The specific calculation formula of the color difference value ΔE is:

[0037] ΔE=((ΔL) 2 +(Δa) 2 +(Δb) 2 ) 0.5 ;

[0038] Where ΔL, Δa and Δb represent the differences in the three dimensions of L, a and b respectively;

[0039] In the calculation of the color rendering index, the ΔE value in the CIE 1976 Lab* color space is used to express the color difference value. According to the color difference value between the standard sample and the sample under the reference light source, the average value of the relative color difference is obtained, which is the color rendering index Ra. Ri is the special color rendering index of each color sample, and Ra is the average value of these special color rendering indices.

[0040] S62. Based on the performance data obtained by model prediction and combined with the color rendering index, the performance of the LED light source module is divided into different levels according to industry standards or internal enterprise standards:

[0041] Excellent: Color rendering index ≥90, luminous flux ≥4000lm, X, Y chromaticity coordinate error within ±0.0008, module operating temperature is stable and the fluctuation range is within ±2℃;

[0042] Good: Color rendering index is between 80-90, luminous flux ≥3600lm, X, Y chromaticity coordinate error is within ±0.0014, and operating temperature fluctuates within ±4℃;

[0043] Generally: color rendering index is between 70-80, luminous flux ≥3200lm, X, Y chromaticity coordinate error is within ±0.002, working temperature fluctuation range is ±8℃;

[0044] Unqualified: Color rendering index <70, luminous flux ≤2800lm, but the X, Y chromaticity coordinate error exceeds ±0.002, and the operating temperature fluctuation reaches ±10℃;

[0045] S63, analyzing the performance of unqualified light source modules in the test results, and identifying quality problems of the LED light source modules caused by hardware, environment or human factors;

[0046] S64. Propose corresponding optimization suggestions based on the identified influencing factors.

[0047] Furthermore, it also includes conducting a long-term operation stability test on the phosphor-free multi-primary color LED light source module and setting relevant experimental parameters, specifically including:

[0048] The phosphor-free multi-primary color LED light source module needs to operate continuously for at least 100 hours under rated working conditions, and regularly record key performance indicators, including temperature, humidity, luminous flux, chromaticity and color coordinates, as well as abnormal conditions of color decay and color drift that occur during operation.

[0049] Furthermore, the training of the machine learning model specifically includes:

[0050] The machine learning model adopted is BP neural network;

[0051] Data preprocessing: normalize the acquired sensor data, remove outliers, noise, and missing data, ensure data quality and consistency, and facilitate the training of machine learning models;

[0052] Dataset division: divide the preprocessed dataset into training set, validation set and test set;

[0053] Construct a BP neural network model and determine the structure of the neural network, including the number of neurons in the input layer, hidden layer, and output layer;

[0054] Model training: Use the training set data to train the BP neural network, calculate the network output through forward propagation, and then use the back propagation algorithm to adjust the network weights and bias values ​​according to the output error. Repeat this process until the network performance on the validation set reaches the optimal level or meets the stopping condition.

[0055] Model evaluation and adjustment, using validation set data to evaluate model performance;

[0056] Model testing and verification: Use the test set data to test and verify the trained model.

[0057] The beneficial effects of the present invention are:

[0058] The present invention provides a method for detecting a phosphor-free multi-primary color LED light source module, which can solve the shortcomings of existing detection technologies in detecting phosphor-free multi-primary color LED light source modules, increase detection efficiency, and reduce labor costs during the detection process. It can achieve comprehensive performance evaluation, precise testing and data analysis, standardization and repeatability and reliability assurance, provide support for quality control and technological progress of LED lighting products, and is a more economical and comprehensive detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of a method for detecting a phosphor-free multi-primary color LED light source module in an embodiment of the present invention;

[0060] Figure 2 This is a flow chart of the photoelectric performance and temperature stability test in the embodiment of the present invention;

[0061] Figure 3 It is an overview diagram of the BP neural network in the embodiment of the present invention;

[0062] Figure 4 Flow chart of reliability calculation in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, some embodiments of the present application are further described in detail below in conjunction with the accompanying drawings and embodiments.

[0064] Embodiment 1, a method for detecting a phosphor-free multi-primary color LED light source module, comprising the following steps:

[0065] Step S1: Preparation: Before the test begins, ensure that all test equipment and environment meet the test requirements. Specifically include:

[0066] S11: Equipment inspection and calibration, check the status of all testing equipment to ensure they are in good working condition; calibrate the photoelectric testing instruments to ensure the accuracy of the measurement results.

[0067] S12: The detection sensor equipment specifically includes: illuminance meter, spectrometer, colorimeter, temperature sensor, infrared thermal imager and LED junction temperature tester.

[0068] S12: Test environment temperature and humidity control, ensure that the temperature and humidity of the test environment are controlled within the optimal working temperature range of -40℃~85℃, and the humidity is controlled between 5%~60% to reduce the impact of environmental factors on the test results.

[0069] S13: Test environment brightness and color temperature control, indoor test brightness range is 800-1200cd / m², outdoor test brightness is 5000-6000cd / m², color temperature range is between 5000K and 6500K.

[0070] Step S2: Module connection and initialization, check the interface type and specifications of the LED light source module and the detection equipment to ensure that they can fully match; during the connection process, strictly follow the equipment manual or operating guide to avoid connection errors or damage to the equipment due to improper operation, correctly connect the LED light source module to be tested to the detection equipment, and ensure that the connection is stable and reliable. The specific steps are:

[0071] S21, module connection and initialization: prepare the appropriate connection cables and interfaces, securely connect the module and the testing equipment, perform test connections and initialization operations to strictly verify the stability and effect of the connection, thereby ensuring that all components are in good working condition;

[0072] S22, Sensor positioning: Accurately select the placement of the sensor to ensure that it can accurately capture and record the key parameters of the light source and provide reliable data for subsequent analysis;

[0073] S23. Spectral separation and capture: The spectroscopic attachment is carefully made using a transmission diffraction grating and a slit. The diffraction angle of the light can be effectively controlled by precisely adjusting the width and spacing of the slit. This design enables the visible light radiated by the light source to achieve fine dispersion. After diffraction, the originally mixed visible light is accurately separated into monochromatic light of different wavelengths. A high-resolution camera is configured behind the spectroscopic attachment, and its position and parameters are precisely adjusted to ensure that the spectral image after diffraction can be clearly captured.

[0074] S24, key parameter measurement: With the help of professional optoelectronic testing instruments, the core parameters of the module are comprehensively and accurately measured. These parameters include luminous flux, chromaticity, color coordinates and temperature. Accurate acquisition of these data is crucial for evaluating the performance of the module.

[0075] S25, initialization status verification: Check the initialization status of the module by sending a concise test signal to the LED light source module and carefully observing its response. This step helps ensure that the module can operate normally in subsequent tests.

[0076] Step S3: Obtain sensor data, use a illuminance meter, spectrometer, colorimeter, temperature sensor and infrared thermal imager to record the time series of the luminous flux, chromaticity and color coordinate parameters of the LED light source module, and check the data integrity and accuracy; the sensor is installed in a position that can accurately reflect the environmental conditions to be measured. After installation, the sensor is calibrated and multiple measurements are performed at different spatial positions to ensure that the output data is accurate and reliable; then use a high-resolution camera to accurately capture the diffraction spectrum, so as to obtain the spectral distribution of the light source in the visible light band in detail. The specific steps are:

[0077] S31: Prepare test equipment, including illuminometer, spectrometer, colorimeter and infrared thermal imager;

[0078] S32: Luminous flux measurement: use an illuminance meter to measure the brightness of the LED light source module. Place the illuminance meter at a certain distance from the LED light source module, record the measurement results, and pay attention to the unit.

[0079] S33: Chromaticity measurement, using a spectrometer and a colorimeter to measure the LED light source module, and using the colorimeter to measure the color difference value of the LED light source module. The spectrometer and the colorimeter are used to analyze the chromaticity parameters of the light source;

[0080] S34: temperature measurement, when the LED light source module is not working, measure and record its initial temperature, make the LED light source module operate under rated working conditions, and continuously monitor and record the temperature change of the LED chip during the operation of the module, observe and record the time required for the module to reach thermal equilibrium and the temperature value at equilibrium;

[0081] S35: Long-term operation test: let the LED light source module run continuously for 100-1000 hours under rated working conditions. The test is determined according to the usage time of different LED light sources. The key performance indicators of the module, such as temperature, luminous flux, chromaticity, and color coordinate time series, are regularly recorded. Any abnormal conditions of the module during long-term operation, such as light decay and color drift, are monitored and recorded.

[0082] S36: Data processing, smoothing and filtering the data to remove outliers and noise. When any abnormal data or unexpected results appear, check the test environment and instrument status. After troubleshooting, retest the LED light source module to ensure data accuracy.

[0083] Step S4: Machine learning feature recognition, using machine learning technology to establish factor analysis and prediction models, identifying key factors that affect the performance of LED light source modules through analysis of a large amount of test data, and establishing a prediction model based on these factors to predict module performance and the life of the LED light source module. The specific steps are:

[0084] S41: Data preprocessing: Check the data collected by the sensor again, remove outliers, missing values ​​or duplicate values, ensure the quality and consistency of the data, normalize the data, including the input variables temperature and humidity, except for the working time, to ensure the stability and convergence speed of the network training, ensure that the data set is large enough and diverse, and output the luminous flux, chromaticity, brightness attenuation, and color shift to train an accurate model;

[0085] S43: Dataset division: the preprocessed data set is divided into a training set, a validation set and a test set. The training set is used to train the BP neural network model, the validation set is used to adjust the model parameters and prevent overfitting during the training process, and the test set is used to evaluate the final performance of the model.

[0086] S42: Construct a BP neural network model, design the number of neurons in the input layer, hidden layer, and output layer. The number of neurons in the input layer should match the number of features, and the number of neurons in the output layer should be determined based on the prediction target, including performance parameters and lifespan. Calculate the output of the network through forward propagation, and then use the backpropagation algorithm to adjust the weights and bias values ​​of the network according to the output error until the performance of the BP neural network on the validation set reaches the optimal level or the stopping condition is met; use the Sigmoid function as the activation function, randomly initialize the weights and bias values ​​in the network, adjust these values ​​during the training process, and determine the learning rate and number of iterations;

[0087] S43: Prediction model evaluation: Use validation set or test set data to evaluate the prediction performance of the model, calculate the correlation and error between the predicted value and the actual value, and adjust the network structure and training parameters according to the evaluation results to optimize the model performance.

[0088] Step S5: Stability and reliability assessment, using the feature importance assessment function of the machine learning model to analyze the impact of the corresponding features on the stability of the LED module; using the trained machine learning model to predict the life of the LED module, by inputting the key parameters of the module, the model can output the expected life of the module, thereby evaluating its reliability; based on the model prediction results and analysis, identify the key factors affecting the stability and reliability of the LED module, and take corresponding improvement measures, such as optimizing the heat dissipation design and improving the material quality, to improve the overall performance of the module. The specific steps are:

[0089] S51. Using the trained machine learning model, deeply analyze the weight distribution of each input feature in the model, so as to accurately identify the characteristic factors that have a significant impact on the performance of the LED light source module;

[0090] S52. The system organizes the collected life data and makes detailed statistics on the specific number of module failures in different time intervals. Based on these data, the survival rate of the modules at each time point is accurately calculated, providing a solid foundation for subsequent reliability analysis.

[0091] S53. Use Weibull distribution to conduct an in-depth reliability assessment. In this assessment, the shape parameter k and scale parameter λ are mainly focused on (assuming the location parameter θ is 0) to comprehensively analyze the reliability characteristics of the module.

[0092] S54. Use the maximum likelihood estimation method to estimate the shape parameter k and scale parameter λ in the Weibull distribution to ensure the accuracy and effectiveness of the evaluation;

[0093] S55. Based on the estimated parameters k and λ, a reliability curve of the light source module is drawn. The abscissa of the curve represents time t, and the ordinate reflects the reliability R(t). The reliability R(t) can be accurately calculated by the cumulative distribution function of the Weibull distribution, and the formula is expressed as follows:

[0094] R(t)=exp[-(t / λ)k];

[0095] S56. Based on the Weibull distribution, the average life and median life of the module are further calculated. By carefully comparing the performance data predicted by machine learning with the life prediction obtained by reliability analysis, the reliability and stability performance of the LED light source module are comprehensively evaluated.

[0096] S57, sorting according to the size of the feature weight, so as to accurately identify the factors that have a key impact on the performance of the LED light source module, and provide a clear direction for subsequent optimization and improvement;

[0097] S58. The performance parameters predicted by the model are compared with the actual measured values ​​in detail to verify the accuracy of the model. At the same time, the mature BP neural network model is used to accurately predict the service life of the LED light source module based on the identified key factors, providing strong support for product design and use;

[0098] S59. Combine the results of machine learning and reliability analysis to form a comprehensive test of the overall performance of the LED light source module and determine whether it meets the preset performance indicators and industry standards.

[0099] Step S6: Based on the test and prediction results, comprehensively evaluate the performance of the LED module, check whether key indicators such as brightness attenuation and color shift are within the acceptable range, and comprehensively compare the evaluation results with the preset performance indicators and life requirements to determine whether the module meets the standards. If it does, the module passes the evaluation; if it does not, further analyze the reasons and take improvement measures. The specific steps include:

[0100] S61. Calculate the color rendering index. According to the color rendering degree of the color sample specified by the International Commission on Illumination, compare the spectral distribution of the light source to be tested with the spectral distribution of the reference light source, and calculate the color difference value ΔE of the color sample under the two light sources. The specific calculation formula of the color difference value ΔE is:

[0101] ΔE=((ΔL) 2 +(Δa) 2 +(Δb) 2 ) 0.5 ;

[0102] Among them, ΔL, Δa and Δb represent the differences in the three dimensions of L, a and b respectively;

[0103] The ΔE value in the CIE 1976 Lab* color space is used to quantify this color difference. The color rendering index Ra is obtained by calculating the average value of the relative color difference between the standard sample under the measured light source and the reference light source, while Ri represents the specific color rendering index of each color sample;

[0104] S62. Combining the performance data and color rendering index predicted by the model, the performance of the LED light source module is divided into the following levels:

[0105] Excellent: Color rendering index ≥90, luminous flux ≥4000lm, X, Y chromaticity coordinate error within ±0.0008, module operating temperature is stable and the fluctuation range is within ±2℃;

[0106] Good: Color rendering index is between 80-90, luminous flux ≥3600lm, X, Y chromaticity coordinate error is within ±0.0014, and operating temperature fluctuates within ±4℃;

[0107] Generally: color rendering index is between 70-80, luminous flux ≥3200lm, X, Y chromaticity coordinate error is within ±0.002, working temperature fluctuation range is ±8℃;

[0108] Unqualified: Color rendering index <70, luminous flux ≤2800lm, but the error of X, Y chromaticity coordinates exceeds ±0.002, and the operating temperature fluctuation may reach ±10℃;

[0109] S63. In-depth analysis of the performance parameters that do not meet the standards in the evaluation results, accurately identify the specific performance shortcomings of the LED light source module, and point out the direction for subsequent improvement work;

[0110] S64. Based on the identified performance influencing factors, match them with the specific hardware design or operating environment of the LED light source module, and propose targeted optimization measures. These suggestions are intended to help improve the overall performance of the module and ensure that it is more stable and efficient in actual applications.

[0111] See also Figure 1 , is a flow chart of an exemplary method for detecting a phosphor-free multi-primary color LED light source module in an embodiment of the present invention. The method for detecting a phosphor-free multi-primary color LED light source module specifically comprises the following steps:

[0112] S101. Before testing the phosphor-free multi-primary color LED light source module, select test instruments that meet international standards and have high precision and high stability, such as flux meters, color temperature meters and spectrum analyzers, to accurately measure the key performance indicators of the module. In order to comprehensively evaluate the performance of the module, other advanced auxiliary equipment such as illuminance meters and colorimeter are also required. The test is carried out in a constant temperature and humidity test room to accurately control environmental parameters such as temperature, humidity and light to ensure the consistency of test conditions. The specific steps of the preparation work include equipment inspection and calibration, so that the indoor test brightness range is 800-1200cd / m², the outdoor test brightness is 5000-6000cd / m², the color temperature range is between 5000K and 6500K, and the optoelectronic test instruments are calibrated. At the same time, the test room is equipped with high-precision temperature control equipment, light control equipment, and data acquisition system.

[0113] S102. Use the appropriate connection cables and interfaces to securely connect the module and the testing equipment, and perform test connections and initialization operations to ensure that all components are in good working condition and verify the stability and effectiveness of the connection. Select and place sensors to ensure that they can accurately capture and record the key parameters of the light source to provide reliable data support for subsequent analysis. Use transmission diffraction gratings and slits to make spectroscopic accessories, adjust the slits to achieve fine dispersion of visible light radiated by the light source, and configure a high-resolution camera to capture the spectral image after diffraction. Use optoelectronic test instruments to comprehensively and accurately measure the core parameters of the module, including luminous flux, chromaticity, color coordinates and temperature, to ensure the accuracy of the data to evaluate the module performance. Send test signals and observe the module response to check its initialization status to ensure that the module can operate normally in subsequent tests.

[0114] S103. Obtain sensor data. Photoelectric performance testing is a key link in evaluating the performance of LED light source modules. By using professional testing instruments, the core parameters of the module, such as luminous flux, chromaticity, and color rendering index, can be accurately measured. These parameters reflect the optical performance of the module and help to understand its characteristics in terms of luminous efficiency and color reproduction ability. The long-term operation test allows the LED light source module to run continuously for 100-1000 hours under rated working conditions, which is determined according to the use time of different LED light sources. The key performance indicators of the module, such as temperature, luminous flux, chromaticity, and color coordinate time series, are regularly recorded. The abnormal phenomena that occur during long-term operation of the module are monitored and recorded. All data during the test process, including test conditions, test parameters, and test results, are recorded in detail for subsequent analysis and processing. During measurement, ensure the stability of the test environment to avoid interference with the test results caused by external factors. At the same time, in order to improve the accuracy and reliability of the test, the test data needs to be collected and averaged multiple times.

[0115] S104. To clean, organize and standardize the collected sensor data to improve data quality and availability, use BP neural network to extract key features from the preprocessed data, such as brightness change rate, chromaticity stability, etc. Based on the extracted features, establish factor analysis and prediction models to reveal the relationship between module performance and environmental conditions, predict the performance of the module under different conditions, and output the weights of related influencing factors and life prediction. By observing the changing trends of these outputs, the performance of the module at different temperatures can be deeply evaluated, and then its thermal stability and working reliability can be understood, which will help to more accurately predict and control the performance of the module in practical applications.

[0116] S105. Using the BP neural network machine learning model, the weights of each input feature in the model are deeply analyzed, the features that have a significant impact on the performance of the LED light source module are accurately identified, the module life data is systematically sorted, the number of failures in different time intervals is counted, and the survival rate at each time point is accurately calculated based on this. The Weibull distribution is used for reliability assessment, and the shape parameter k and scale parameter λ are accurately estimated by the maximum likelihood estimation method. The horizontal axis of the curve represents time t, and the vertical axis represents reliability R(t). The reliability R(t) is calculated by the cumulative distribution function of the Weibull distribution, and the formula is:

[0117] R(t)=exp[-(t / λ) k ];

[0118] Then, the reliability curve of the light source module is drawn, and the performance data predicted by machine learning and the life prediction obtained by reliability analysis are combined to comprehensively evaluate the reliability and stability of the LED light source module. The accuracy of the model is verified by comparing the performance parameters predicted by the model with the actual measured values, and the BP neural network model is used to accurately predict the service life of the module. Based on the results of machine learning and reliability analysis, the overall performance of the LED light source module is comprehensively tested to ensure that it meets the preset performance indicators and industry standards.

[0119] S106. According to the standards of the International Commission on Illumination, the color rendering index is calculated by comparing the spectral distribution of the light source to be tested with the reference light source. The index reflects the color rendering ability of the light source for the color sample. The specific calculation formula of the color difference value ΔE is:

[0120] ΔE=((ΔL) 2 +(Δa) 2 +(Δb) 2 ) 0.5 ;

[0121] Where ΔL, Δa and Δb represent the differences in the three dimensions of L, a and b respectively;

[0122] In the calculation of the color rendering index, the ΔE value in the CIE 1976 Lab* color space is used to represent the color difference value. According to the color difference value between the standard sample and the tested sample under the reference light source, the average value of the relative color difference is obtained, which is the color rendering index Ra. Ri is the special color rendering index of each color sample, and Ra is the average value of these special color rendering indexes. Combining the performance data predicted by the model and the color rendering index, the performance of the LED light source module is carefully divided into four levels: excellent, good, general and unqualified:

[0123] Excellent: Color rendering index ≥90, luminous flux ≥4000lm, X, Y chromaticity coordinate error within ±0.0008, module operating temperature is stable and the fluctuation range is within ±2℃;

[0124] Good: Color rendering index is between 80-90, luminous flux ≥3600lm, X, Y chromaticity coordinate error is within ±0.0014, and operating temperature fluctuates within ±4℃;

[0125] Generally: color rendering index is between 70-80, luminous flux ≥3200lm, X, Y chromaticity coordinate error is within ±0.002, working temperature fluctuation range is ±8℃;

[0126] Unqualified: Color rendering index <70, luminous flux ≤2800lm, but the X, Y chromaticity coordinate error exceeds ±0.002, and the operating temperature fluctuation reaches ±10℃;

[0127] This assessment can help to fully understand the performance of the module, conduct in-depth analysis of the performance parameters that do not meet the standards in the evaluation results, accurately identify the specific performance shortcomings of the LED light source module, and match them with the module's hardware design or operating environment based on these shortcomings in order to propose targeted optimization measures.

[0128] See also Figure 2 , is a flow chart of the photoelectric performance and temperature stability test in an embodiment of the present invention, and the test process specifically includes:

[0129] Step S201, preparations before testing. First, optoelectronic equipment that meets the test requirements needs to be selected, including illuminance meters, spectrometers, colorimeters, etc. They will be used to accurately measure the performance indicators of the phosphor-free multi-primary LED light source module. At the same time, in order to ensure the stability of the test environment, temperature sensors, infrared thermal imagers and other equipment need to be prepared to monitor and record temperature changes during the test in real time. In addition, it is also necessary to build a stable test environment, including installing temperature control equipment and lighting control equipment to simulate different environmental conditions, and configuring an efficient data acquisition system for real-time collection and analysis of test data.

[0130] During the preparation work, it is also necessary to prepare corresponding test software and hardware equipment according to the test requirements. The test software should have data analysis functions with flexible configuration options so that it can be customized according to different test scenarios and conduct in-depth mining and analysis of the collected data. The selection of hardware equipment should focus on its accuracy, stability and reliability to ensure the accuracy and repeatability of the test results.

[0131] Step S202, initial performance test is to evaluate the basic performance of the phosphor-free multi-primary color LED light source module under standard conditions. In this step, it is necessary to measure and record the module's key parameters such as chromaticity, light intensity (luminous flux), temperature, etc. These parameters not only reflect the module's luminous efficiency and lighting capacity, but also are important bases for evaluating performance changes in subsequent tests. In order to ensure the accuracy of the test results, high-precision measuring equipment should be used, and multiple measurements should be performed under constant environmental conditions to take the average value.

[0132] Step S203: When connecting and initializing the phosphor-free multi-primary color LED light source module, it is necessary to strictly follow the standard operating procedures to ensure that the power connection of the module is stable and reliable, the signal transmission is correct, and the necessary software configuration is performed. During the initialization process, it is also necessary to carefully check the various parameter settings of the module, such as brightness, chromaticity, etc., to ensure that they meet the test requirements. The purpose of this step is to lay a solid foundation for subsequent tests and ensure that the module can work normally during the test and output accurate data.

[0133] Step S211, the initial temperature measurement is performed to establish a baseline for temperature changes. In this step, it is necessary to record the initial temperature of the test environment and the temperature of the module before operation. These data will be used for subsequent comparison and analysis of the performance of the module at different temperatures, so as to more accurately evaluate its thermal stability.

[0134] Step S212, thermal stability test is an important step in evaluating the working stability and reliability of the module under different temperature conditions. In this step, by gradually changing the temperature of the test environment, observe and record the performance changes of the module under various temperature conditions, including the changing trends of key parameters such as luminous flux and chromaticity. By in-depth analysis of these data, we can fully understand the working characteristics of the module at different temperatures and provide strong support for its optimization in practical applications.

[0135] Step S204, brightness and chromaticity testing is an important means to evaluate the optical performance of the phosphor-free multi-primary color LED light source module. By using professional brightness testing equipment and chromaticity measuring equipment, the brightness and chromaticity performance of the module under different temperature conditions can be accurately measured. The data is used to analyze the impact of temperature on the optical performance of the module and provide feedback for the optimal design of the module.

[0136] Step S213, long-term operation stability test is a key step in evaluating the reliability and durability of the module in actual applications. In this step, the module is allowed to work continuously for a period of time under set temperature conditions, and its performance is continuously monitored. By paying attention to whether there is performance degradation or other abnormal conditions, the stability and reliability of the module in actual operation can be fully understood.

[0137] Step S205, perform a chromaticity test, using a spectrometer, a colorimeter and other equipment to measure the chromaticity parameters of the light emitted by the phosphor-free multi-primary color LED light source module, such as chromaticity, color coordinates, etc. This step is used to evaluate the color performance stability of the module at different temperatures.

[0138] Step S206: Check and verify the rationality and accuracy of all test data, check and eliminate abnormal data to ensure the reliability of the test results. At the same time, organize and analyze the data to extract useful performance change trends and rules to ensure the rationality and accuracy of the test data and provide a reliable basis for subsequent data analysis and evaluation.

[0139] See also Figure 3 , is an overview diagram of the BP neural network in an embodiment of the present invention, specifically including:

[0140] The BP neural network in the embodiment of the present invention is a multi-layer feedforward neural network applied to factor analysis and life prediction of LED light source modules. The network is trained by the back propagation algorithm to achieve accurate prediction of LED module performance and life. Specifically, it includes:

[0141] Input layer: Receives multi-dimensional feature data related to the LED light source module, such as environmental factors such as temperature, humidity, current, voltage, and the module's own photoelectric parameters. These data are the basis for the machine learning model to perform factor analysis and life prediction.

[0142] Hidden layer: The complex relationship between input features is captured through nonlinear transformation of multiple hidden layers. Each layer increases the nonlinear expression ability of the network through the activation function, so that the model can learn more complex feature representations. In this embodiment, the Sigmoid function is used as the activation function.

[0143] Output layer: The output layer generates performance evaluation indicators and life expectancy prediction values ​​of LED light source modules. These outputs can be used as the basis for module quality assessment and maintenance plan formulation.

[0144] Using the back propagation algorithm, according to the difference between the output layer and the actual label, the error is back propagated layer by layer, and the weights in the network are adjusted. Through multiple iterative training, the network gradually approaches the real input-output mapping relationship.

[0145] During the training process, combined with the characteristics of LED light source modules, appropriate optimization algorithms and learning rate adjustment strategies are used to improve the convergence speed and prediction accuracy of the model. The BP neural network can automatically learn the key factors that affect the performance and life of LED light source modules, providing strong support for improving module design and manufacturing processes; through the trained model, the expected life of LED light source modules in different environments and working conditions can be accurately predicted, providing important reference for corporate product planning and user maintenance plans.

[0146] See also Figure 4 , Figure 4 This is a flow chart of reliability calculation in an embodiment of the present invention, which specifically includes:

[0147] Step 401, determine the analysis model: select Weibull distribution as the reliability assessment model because it has good adaptability in describing product life and failure mode, and verify whether the data conforms to the characteristics of Weibull distribution through historical data and professional judgment.

[0148] Step 402, data collection: collect failure data or service life data related to the LED light source module, obtain failure data from the actual use environment, or obtain service life data through laboratory testing to ensure the accuracy and completeness of the data.

[0149] Step 403: Establish a likelihood function. Based on the probability density function of the Weibull distribution, a likelihood function is constructed to describe the possibility of model parameters under given data.

[0150] Step 404, parameter solution: by maximizing the likelihood function, the best estimated values ​​of the shape parameter (k) and the scale parameter (λ) of the Weibull distribution are solved, and the parameters are estimated by programming means to ensure the accuracy and efficiency of the solution.

[0151] Step 405, verifying the model: Substitute the solved parameters into the probability density function of the Weibull distribution to verify the fitting effect of the model on the actual data. Evaluate the fitting effect of the model by drawing a fitting curve, calculating statistical indicators (such as AIC, BIC), etc., to ensure the accuracy and reliability of the model.

[0152] Step 406, apply the model: Based on the verification results, if the model fitting effect is good, the model can be applied to the reliability evaluation of the LED light source module; if the fitting effect is not good, it is necessary to reconsider the validity of the data or try other distribution models. According to the model evaluation results, formulate targeted reliability improvement measures, such as optimizing product design, improving production processes, etc., to improve the reliability of the LED light source module.

[0153] The Weibull distribution is used as the reliability evaluation model in the embodiment, which has the following advantages over the traditional reliability analysis method:

[0154] Wide adaptability: Weibull distribution can well characterize different failure stages in the product life cycle, such as early failure, random failure, and aging failure (bathtub curve), which makes it more adaptable in analyzing the reliability of complex products.

[0155] Flexible parameter system: shape parameter (β or k), the shape parameter of the Weibull distribution determines the distribution form of the failure rate. When the shape parameter is less than 1, the failure rate decreases with time. When it is equal to 1, the failure rate is a constant. When it is greater than 1, the failure rate increases with time. This flexibility makes the Weibull distribution applicable to products with different failure characteristics. The scale parameter (η) plays the role of enlarging or reducing the coordinate scale, further enhancing the adaptability of the Weibull distribution in reliability assessment.

[0156] Clear analytical expressions: convenient for mathematical processing. The basic reliability functions of Weibull distribution (such as reliability function, failure rate function, etc.) have clear analytical expressions, which enables accurate mathematical processing and analysis in the reliability assessment process. It is easy to fit and verify. The parameters of Weibull distribution can be easily estimated through methods such as maximum likelihood estimation, and the fitting effect of the model can be verified by drawing fitting curves and calculating statistical indicators.

[0157] Predictive Capabilities Future Reliability Prediction: The reliability assessment model based on Weibull distribution can not only analyze current and historical data, but also predict future reliability performance, providing important reference for product design, production and maintenance.

[0158] As a reliability assessment model, Weibull distribution has significant advantages over traditional reliability analysis methods in terms of adaptability, parameter system, analytical expression, practical application background and predictive ability. These advantages make Weibull distribution more effective and reliable in dealing with reliability assessment problems of complex products.

[0159] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for detecting a phosphor-free multi-primary color LED light source module, characterized in that: The following steps are involved: S1: Preparation work, ensuring that the testing equipment and environment meet the test requirements. The specific steps of step S1 include: the specific steps of the preparation work include equipment inspection and calibration, so that the indoor test brightness range is 800-1200cd / m 2 , outdoor test brightness is 5000-6000cd / m 2 , the color temperature range is between 5000K and 6500K, and the optoelectronic test instruments are calibrated; the sensor equipment used for detection includes illuminometer, spectrometer, colorimeter, temperature sensor, timer and infrared thermal imager; S2: Module connection and initialization, checking and connecting the LED light source module and the detection equipment, and testing the connection stability. The step S2 specifically includes: S21. The specific steps of module connection and initialization include preparing connection cables and interfaces, connecting modules and devices, and performing test connection and initialization to verify the connection effect and ensure that it is in good working condition; S22, placing the sensor at a suitable location so as to be able to capture various parameters of the light source; S23. Use a transmission diffraction grating and a slit to make a light splitting attachment. By strictly controlling the width and spacing of the slits, the diffraction angle of the light is manipulated to achieve the dispersion of visible light radiated by the light source. After diffraction, the originally mixed visible light is accurately separated into monochromatic light of different wavelengths. A high-resolution camera is set behind the light splitting attachment. By adjusting the position and parameters of the camera, the spectrum image after diffraction is captured. S24. Use professional photoelectric testing instruments to measure other key parameters of the module, including luminous flux, chromaticity, color coordinates and temperature; S25, checking the initialization state by sending a simple test signal to the LED light source module and observing its response; S3: Full performance evaluation and life tracking of LED light source module. A high-resolution camera is used to obtain a diffraction spectrum diagram, obtain the spectral distribution of the visible light band of the light source, and use a photoelectric testing instrument to measure the module's luminous flux, chromaticity, color coordinates, and temperature parameters, and record the life data and number of failed modules of the LED light source module in actual use. The step S3 specifically includes: using an illuminance meter, a spectrometer, a colorimeter, a temperature sensor, and an infrared thermal imager to record the time series of the luminous flux, chromaticity, and color coordinate parameters of the LED light source module, and check the data integrity and accuracy; wherein, the sensor is installed at a position that can accurately reflect the environmental conditions to be tested. After installation, the sensor is calibrated and multiple measurements are performed at different spatial positions to ensure that the output data is accurate and reliable; then a high-resolution camera is used to accurately capture the diffraction spectrum diagram to obtain the spectral distribution of the light source in the visible light band; S4: machine learning feature recognition, establishing a reliability analysis and prediction model for the LED light source module, the step S4 specifically includes: using machine learning technology to establish a factor analysis and prediction model, using a machine learning method to take temperature, humidity and working time as model inputs, and performance indicators luminous flux, chromaticity, brightness attenuation, and color shift as model outputs, identifying key factors affecting the performance of the LED light source module through analysis of a large amount of test data, and establishing a prediction model based on these factors to predict module performance and the life of the LED light source module; S5: Stability and reliability assessment, using the trained model and reliability curve to perform factor analysis and life prediction; the step S5 specifically includes: S51, using the trained machine learning model, by analyzing the weights of each input feature in the model, determining significant features that affect the performance of the LED light source module; S52, sorting the collected life data, counting the number of modules that fail in different time periods, and calculating the survival rate of the modules at different time points based on the failure data; S53, using Weibull distribution for reliability assessment, which mainly includes shape parameter k, scale parameter λ and location parameter θ. In reliability analysis, θ = 0 is assumed, and the shape parameter k and scale parameter λ are mainly concerned; S54, using the maximum likelihood estimation method, estimate the shape parameter k and scale parameter λ of the Weibull distribution; S55. Draw a reliability curve of the light source module according to the estimated parameters k and λ. The horizontal axis of the curve represents time t, and the vertical axis represents reliability R(t). The reliability R(t) is calculated by the cumulative distribution function of the Weibull distribution, and the formula is: R(t)=exp[-(t / λ) k ]; S56. Calculate the average life and median life according to Weibull distribution, compare and analyze the life prediction obtained by machine learning and reliability analysis with the actual service life, and evaluate the reliability and stability of the LED light source module; S57, sorting the features according to weights to identify key factors affecting the performance of the LED light source module; S58, comparing the predicted performance parameters with the actual measured values ​​to verify the accuracy of the model, and using the trained BP neural network model to predict the service life of the LED light source module based on the input key factors; S59, based on the analysis results of the above steps, the overall performance of the LED light source module is evaluated to determine whether it meets the preset performance indicators and industry standards; In step S5, the training of the machine learning model specifically includes: The machine learning model adopted is BP neural network; Data preprocessing: normalize the acquired sensor data, remove outliers, noise, and missing data, ensure data quality and consistency, and facilitate the training of machine learning models; Dataset division: divide the preprocessed dataset into training set, validation set and test set; Construct a BP neural network model and determine the structure of the neural network, including the number of neurons in the input layer, hidden layer, and output layer; Model training: Use the training set data to train the BP neural network, calculate the output of the network through forward propagation, and then use the back propagation algorithm to adjust the weight and bias value of the network according to the output error, and repeat this process until the network's performance on the validation set reaches the optimal level or meets the stopping condition; Model evaluation and adjustment, using validation set data to evaluate model performance; Model testing and verification: Use test set data to test and verify the trained model; S6: According to the evaluation results, the performance of the LED light source module is evaluated, and the evaluation results are compared with the preset performance indicators and life requirements to determine whether the module meets the standards; the step S6 specifically includes: S61. Calculate the color rendering index. According to the color rendering degree of the color sample specified by the International Commission on Illumination, compare the spectral distribution of the light source to be tested with the spectral distribution of the reference light source, and calculate the color difference value ΔE of the color sample under the two light sources. The specific calculation formula of the color difference value ΔE is: ΔE=((ΔL) 2 +(Δa) 2 +(Δb) 2 ) 0.5 ; Where ΔL, Δa and Δb represent the differences in the three dimensions of L, a and b respectively; In the calculation of the color rendering index, the ΔE value in the CIE 1976Lab* color space is used to represent the color difference value. According to the color difference value between the standard sample and the sample under the reference light source, the average value of the relative color difference is obtained, which is the color rendering index Ra. S62. Based on the performance data obtained by model prediction and combined with the color rendering index, the performance of the LED light source module is divided into different levels according to industry standards or internal enterprise standards: Excellent: Color rendering index ≥90, luminous flux ≥4000lm, X, Y chromaticity coordinate error within ±0.0008, module operating temperature is stable and the fluctuation range is within ±2℃; Good: Color rendering index is between 80-90, luminous flux ≥3600lm, X, Y chromaticity coordinate error is within ±0.0014, and operating temperature fluctuates within ±4℃; Generally: color rendering index is between 70-80, luminous flux ≥3200lm, X, Y chromaticity coordinate error is within ±0.002, working temperature fluctuation range is ±8℃; Unqualified: Color rendering index <70, luminous flux ≤2800lm, but the X, Y chromaticity coordinate error exceeds ±0.002, and the operating temperature fluctuation reaches ±10℃; S63, analyzing the performance of unqualified light source modules in the test results, and identifying quality problems of the LED light source modules caused by hardware, environment or human factors; S64. Propose corresponding optimization suggestions based on the identified influencing factors.

2. The detection method of a phosphor-free multi-primary color LED light source module according to claim 1, characterized in that: It also includes long-term operation stability testing of the phosphor-free multi-primary color LED light source module and setting relevant experimental parameters, including: The phosphor-free multi-primary color LED light source module needs to operate continuously for at least 100 hours under rated working conditions, and regularly record key performance indicators, including temperature, humidity, luminous flux, chromaticity and color coordinates, as well as abnormal conditions of color decay and color drift that occur during operation.

Images