Cloud-based LED tube quality management system and method

Abstract

The application relates to the technical field of quality management, in particular to a cloud-based LED lamp tube quality management system and method, which comprises the following steps: obtaining chip brightness value, voltage stability value and service time value parameters based on cloud data flow, and standardizing and processing parameter ranges.The application realizes the accuracy and systematization of data processing by standardizing and classifying and hierarchically grouping analyzing the brightness value, the voltage stability value and the service time value, reveals the internal relation between key performance parameters by correlation calculation combined with cumulative change rate, improves the depth of data analysis, enhances the data adaptability by dynamic adjustment of environmental temperature fluctuation and performance parameters, guarantees the real-time and accuracy of analysis results, makes performance classification more flexible by dynamic weight optimization and cluster structure generation, realizes the synchronous optimization of key performance indicators in light source, drive and life categories by combining with the correction of classification boundaries, and comprehensively improves the fine level of quality management and the problem feedback efficiency.

Description

Technical Field

[0001] This invention relates to the field of quality management technology, and in particular to a cloud-based LED tube quality management system and method. Background Technology

[0002] The field of quality management technology encompasses the techniques and methods for monitoring, evaluating, analyzing, and controlling the quality of products, processes, and services. Its core content includes ensuring that products or services meet predetermined standards and requirements through scientific quality management methods and techniques, covering multiple aspects from production process optimization and the implementation of quality testing technologies to the collection and analysis of quality data. The overall quality management technology field systematically integrates hardware equipment, software platforms, and management methods to achieve effective quality control through end-to-end monitoring and data management.

[0003] The LED tube quality management system refers to a technical system specifically designed for monitoring and managing the quality of LED tubes. This patent addresses quality inspection, performance evaluation, and data analysis management during the production and operation of LED tubes. It covers the detection and data collection of key performance indicators such as brightness, color temperature, and energy consumption. Based on a cloud computing platform, the system stores, classifies, and analyzes the collected data. Combined with IoT devices, it provides real-time feedback and decision support for quality issues. Ultimately, it completes the quality management of the entire LED tube lifecycle through specific hardware monitoring devices and a cloud processing platform.

[0004] Existing technologies lack dynamic relationship analysis between parameters when processing brightness and voltage stability values, making it difficult to fully reflect the intrinsic connections between performance parameters and reducing the effectiveness of monitoring. The long-term impact of environmental factors on performance is not fully considered, causing assessment results to easily deviate from actual operating conditions. The grouping and classification mechanisms lack flexibility, resulting in low efficiency in performance differentiation management and affecting the speed of refined monitoring and problem feedback. Insufficient ability to adjust classification boundaries makes quality assessment difficult to adapt to complex operating environments and limits the ability of data management to support the entire product lifecycle. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a cloud-based LED tube quality management system and method.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The cloud-based LED tube quality management system includes:

[0007] The data standardization processing module acquires chip brightness, voltage stability, and usage duration parameters based on cloud data streams, standardizes the parameter range, classifies chip brightness and voltage stability values ​​into light source and drive parameters, classifies usage duration values ​​into lifespan parameters, and hierarchically and group the classified data to generate a multi-dimensional performance standardization set.

[0008] The parameter correlation and fuzzy matrix generation module, based on the multidimensional performance standardization set, extracts the chip brightness value and voltage stability value and performs difference calculation to obtain parameter correlation. It calculates the cumulative change rate of the usage duration value and chip brightness value, merges the correlation and change rate to form a fuzzy correlation matrix, normalizes the matrix data, and generates an initial weight parameter set.

[0009] The dynamic weight generation module, based on the initial weight parameter set, performs difference analysis by combining the external ambient temperature change with the chip brightness value and voltage stability value, extracts the offset adjustment value between the environmental fluctuation value and the usage time value, normalizes all offset values ​​and matches the weight category, and generates a dynamic weight parameter set.

[0010] The multi-level performance clustering module, based on the dynamic weight parameter set, groups and sorts the chip brightness values ​​and voltage stability values ​​of the light source and driver categories, calculates the group differences by combining the usage time values ​​of the lifespan category, extracts the group offset and weight matching parameters to verify the clustering, and generates a multi-level performance clustering structure.

[0011] The classification boundary optimization and dynamic adjustment module, based on the multi-level performance clustering structure, extracts the classification boundary parameters of chip brightness value, voltage stability value, and usage time value from the clustering results, and corrects the classification boundary by combining the dynamic weight adjustment value, thereby completing the synchronous optimization and dynamic adjustment of light source, driver, and lifespan categories, and generating a set of dynamic adjustment values ​​for classification management.

[0012] The multidimensional performance standardization set includes chip brightness value, voltage stability value, and usage time value. The initial weight parameter set specifically includes parameter correlation, cumulative change rate, and fuzzy correlation matrix. The dynamic weight parameter set includes environmental fluctuation value, offset adjustment value, and weight category. The multi-level performance clustering structure includes group offset, weight matching parameter, and clustering verification result. The classification management dynamic adjustment value set specifically includes classification boundary parameter, dynamic weight adjustment value, and synchronous optimization and dynamic adjustment result.

[0013] As a further aspect of the present invention, the steps for obtaining the multidimensional performance standardization set are specifically as follows:

[0014] Based on cloud data streams, the chip brightness value, voltage stability value, and usage time value are standardized. The chip brightness value and voltage stability value are combined after being normalized by the range of their respective distribution ranges to generate light source and driving parameters. The usage time value is normalized by the ratio of cumulative usage time to normal usage to generate lifespan parameters.

[0015] Using the light source and driving parameters and the lifetime parameters, the layering process is performed based on their distribution characteristics in multidimensional space. The distribution center point of each layer is calculated by calling the mean and standard deviation of the parameter value distribution. The layering range is determined by comparing the distance between the center points, and an initial set of multi-layer performance is generated.

[0016] The parameters within the initial set of multi-layer performance are grouped, and the correlation index of the parameters within each group is judged to meet the minimum threshold condition using the following formula:

[0017]

[0018] Calculate the comprehensive score for each group, filter the groups based on the goodness-of-fit threshold of the group index values, and generate a multidimensional performance standardized set;

[0019] Where G represents the combined score of the parameters after grouping, P i w represents the normalized value of the i-th parameter in the group. i This represents the weight coefficient of the corresponding parameter, where i is the index of the parameter in the current group and n is the total number of parameters in the current group.

[0020] As a further aspect of the present invention, the step of obtaining the initial weight parameter set specifically includes:

[0021] Based on the chip brightness value and voltage stability value in the multidimensional performance standardization set, the correlation parameter is calculated by the difference operation of the corresponding positions of the two, and it is determined whether the absolute value of the difference of each pair of parameters meets the specified threshold range. The parameter combinations that meet the conditions are included in the chip brightness and voltage stability difference correlation parameter set.

[0022] The cumulative change rate of the usage duration value and the chip brightness value is calculated. The change rate parameter is generated by standardizing the ratio of the cumulative change value to the initial parameter value. The linear correlation coefficient between the change rate and the cumulative parameter is calculated. It is determined whether the correlation coefficient meets the set range. Parameters that meet the conditions are screened into the usage duration and brightness change rate parameter set.

[0023] The parameter set relating chip brightness and voltage stability difference and the parameter set relating usage duration and brightness change rate are merged using the following formula:

[0024]

[0025] Calculate the comprehensive correlation value of the positions of multiple matrices and generate a fuzzy correlation matrix;

[0026] Among them, M ij P represents the comprehensive correlation value between the i-th row and the j-th column in the fuzzy correlation matrix. i Q represents the chip's brightness value. j Represents the voltage stability value, R ij The cumulative rate of change correlation weights are defined as follows: i is the brightness value index, j is the voltage stability value index, n is the total number of parameters, and P is the cumulative rate of change correlation weights. k Q represents the k-th parameter value in the chip's brightness values. k Represents the k-th parameter value in the voltage stability value, |P i -Q j | represents the absolute difference between the chip's brightness value and its voltage stability value;

[0027] Normalization is performed on the comprehensive correlation values ​​in the fuzzy correlation matrix. Matrix parameters whose comprehensive correlation values ​​meet the preset threshold range are selected. The weight normalization values ​​of the selected matrix parameters are calculated. The normalization values ​​and the initial parameter values ​​of the corresponding matrix positions are called to generate an initial weight parameter set.

[0028] As a further aspect of the present invention, the step of obtaining the dynamic weight parameter set specifically includes:

[0029] Based on the chip brightness value, voltage stability value and external ambient temperature value in the initial weight parameter set, the difference between the chip brightness value and the external ambient temperature value and the difference between the voltage stability value and the external ambient temperature value are calculated respectively. Parameter combinations that satisfy the absolute value of the difference within a set threshold range are selected to generate an environmental fluctuation difference parameter set.

[0030] Extract usage duration values ​​and environmental fluctuation differences from the set of environmental fluctuation difference parameters, calculate the adjustment offset value of the change trend of the usage duration values ​​to the environmental fluctuation difference, and generate an offset adjustment parameter set by accumulating the change trend and normalization ratio.

[0031] The offset adjustment parameter set is used to match weight categories. This is achieved by comprehensively calculating the offset adjustment value, weight parameter value, and environmental fluctuation parameter using the following formula:

[0032]

[0033] The combined value of multiple matching categories is called to calculate and generate a dynamic weight parameter set;

[0034] Among them, W dyn P represents the dynamic weight parameter. i T represents the offset adjustment parameter. j For the corresponding weight category, The sum of squares of environmental fluctuation parameters. This is a normalized calculation of the sum of squares of the weight values ​​and the offset values, where i, j, k, and m are parameter indices, and n is the total number of parameters.

[0035] As a further aspect of the present invention, the step of obtaining the multi-level performance clustering structure specifically includes:

[0036] Based on the light source and driver category parameters in the dynamic weight parameter set, the chip brightness value and voltage stability value are extracted. The chip brightness value and voltage stability value are sorted according to the light source and driver category. The sorting difference value within the group is calculated and the grouping of the difference value that meets the target threshold range is selected to generate the sorting result of the light source and driver category.

[0037] Extract the usage duration value of the lifespan category from the grouping and sorting results of the light source and drive categories, calculate the grouping difference of the usage duration value within the group with respect to the light source and drive categories, and through weighted cumulative analysis of the grouping difference and lifespan category offset parameter, filter the groups whose difference offset values ​​meet the threshold range, and generate grouping offset and weight matching parameters.

[0038] Using the group offset and weight matching parameters, a weighted normalized index of the intra-group offset value and dynamic weight parameters is calculated, using the formula:

[0039]

[0040] Calculate multi-level performance clustering metrics, integrate the clustering metrics with dynamic weight parameters, and generate a multi-level performance clustering structure.

[0041] Among them, C p A represents a multi-level performance clustering metric. i B represents the offset parameter within the group. i D represents the dynamic weight parameter. i R represents the group difference value. i The association weights for the offset parameters, |A i ·B i +D i | represents the absolute value of the offset parameter, weight parameter, and difference value. The summation of the sum of squares of differences between groups and the sum of squares of weights is calculated, where k is the total number of cluster groups, i and j are parameter indices, and n and m are the total number of parameters within and between groups.

[0042] As a further aspect of the present invention, the step of obtaining the dynamically adjusted value set for classification management specifically includes:

[0043] The clustering boundary parameters of chip brightness value, voltage stability value and usage time value are extracted from the multi-level performance clustering structure. By comparing the distribution range of the multi-classification boundary parameters, the initial parameter range of the classification boundary is determined, and the clustering classification boundary parameters are generated.

[0044] The dynamic weight adjustment value is interactively analyzed with the clustering classification boundary parameters. By calculating the offset correction amount of the dynamic weight for each classification boundary value, the corrected classification boundary value set is called to generate the corrected classification boundary parameters.

[0045] By combining the corrected classification boundary parameters and dynamic weight adjustment values, synchronous optimization of light source, drive, and lifetime categories is achieved, using the formula:

[0046]

[0047] Calculate the dynamic adjustment index for the classification boundary, integrate the adjustment index into the corrected classification boundary value, and generate a set of dynamic adjustment values ​​for classification management.

[0048] Among them, T opt A represents the dynamic adjustment index of the classification boundary. i B represents the corrected classification boundary parameter. i C represents the dynamic weight adjustment value. i For boundary offset correction, |A i -B i To correct the absolute deviation between the boundary and the weight adjustment value, F k is the optimized clustering index for the classification groups, p is the total number of classification groups, i, j, k are parameter indices, and n, m, p are the total number of parameters within, between, and within the classification groups, respectively.

[0049] The cloud-based LED tube quality management method, which is executed based on the aforementioned cloud-based LED tube quality management system, includes the following steps:

[0050] S1: Based on cloud data stream, collect chip brightness value, voltage stability value, and usage duration parameters. Classify chip brightness value and voltage stability value as light source and driving parameters, and use duration value as lifespan parameter. Perform standardization operations through data classification, and perform data layering and grouping based on parameter characteristics to generate a multi-dimensional performance standardization set.

[0051] S2: Based on the multidimensional performance standardization set, calculate the difference between the chip brightness value and voltage stability value to obtain the correlation value. At the same time, calculate the cumulative change rate of usage time and brightness value, merge the two into a fuzzy correlation matrix, normalize the matrix data, and generate an initial weight parameter set.

[0052] S3: Based on the initial weight parameter set, analyze the deviation between the external ambient temperature and the chip brightness and voltage value, extract the offset adjustment value, normalize all offset values, classify weight categories according to the results, and generate a dynamic weight parameter set.

[0053] S4: Based on the dynamic weight parameter set, perform grouping and sorting of chip brightness value and voltage stability value, calculate the difference between usage duration value groups, verify data offset according to weight matching parameter, and generate multi-level performance clustering structure through clustering verification process;

[0054] S5: Based on the multi-level performance clustering structure, analyze the classification boundaries of chip brightness value, voltage stability value and usage time value in the cluster, perform dynamic weight adjustment, simultaneously optimize the classification boundaries of light source, driver and lifespan categories, and generate a set of dynamic adjustment values ​​for classification management.

[0055] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0056] This invention achieves precise and systematic data processing by standardizing and classifying brightness values, voltage stability values, and usage duration values ​​through hierarchical grouping analysis. Correlation calculations combined with cumulative change rates reveal the intrinsic relationships between key performance parameters, enhancing the depth of data analysis. Dynamic adjustments to performance parameters in response to environmental temperature fluctuations enhance data adaptability, ensuring the real-time nature and accuracy of the analysis results. Dynamic weight optimization and cluster structure generation make performance classification more flexible. Combined with the correction of classification boundaries, this enables simultaneous optimization of key performance indicators across light source, drive, and lifespan categories, comprehensively improving the refinement of quality management and the efficiency of problem feedback. Attached Figure Description

[0057] Figure 1 This is a system flowchart of the present invention;

[0058] Figure 2 This is a flowchart illustrating the steps for obtaining the multidimensional performance standardization set of this invention.

[0059] Figure 3 This is a flowchart of the steps for obtaining the initial weight parameter set in this invention;

[0060] Figure 4 This is a flowchart illustrating the steps for obtaining the dynamic weight parameter set of the present invention.

[0061] Figure 5 This is a flowchart illustrating the steps for obtaining the multi-level performance clustering structure of this invention.

[0062] Figure 6 This is a flowchart illustrating the steps for obtaining the dynamically adjusted value set for classification management in this invention. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 merely illustrative and not intended to limit the invention.

[0064] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0065] Example 1

[0066] Please see Figure 1 This invention provides a technical solution: a cloud-based LED tube quality management system comprising:

[0067] The data standardization processing module acquires chip brightness, voltage stability, and usage duration parameters based on cloud data streams, standardizes the parameter range, classifies chip brightness and voltage stability values ​​into light source and drive parameters, classifies usage duration values ​​into lifespan parameters, and hierarchically and group the classified data to generate a multi-dimensional performance standardization set.

[0068] The parameter correlation and fuzzy matrix generation module is based on a multi-dimensional performance standardization set. It extracts the chip brightness value and voltage stability value and performs difference calculation to obtain the parameter correlation. It calculates the cumulative change rate of the usage duration value and the chip brightness value, merges the correlation and change rate to form a fuzzy correlation matrix, and normalizes the matrix data to generate an initial weight parameter set.

[0069] The dynamic weight generation module is based on the initial weight parameter set. It combines the difference analysis between the external environment temperature change and the chip brightness value and voltage stability value to extract the offset adjustment value between the environmental fluctuation value and the usage time value. It normalizes all offset values ​​and matches them with the weight category to generate a dynamic weight parameter set.

[0070] The multi-level performance clustering module is based on a dynamic weight parameter set. It groups and sorts the chip brightness and voltage stability values ​​of light source and driver categories, calculates the group differences by combining the usage time values ​​of lifespan categories, extracts the group offset and weight matching parameters to verify the clustering, and generates a multi-level performance clustering structure.

[0071] The classification boundary optimization and dynamic adjustment module is based on a multi-level performance clustering structure. It extracts the classification boundary parameters of chip brightness value, voltage stability value, and usage time value from the clustering results, and corrects the classification boundary by combining dynamic weight adjustment values. This completes the synchronous optimization and dynamic adjustment of light source, driver, and lifespan categories, and generates a set of dynamic adjustment values ​​for classification management.

[0072] The multidimensional performance standardization set includes chip brightness value, voltage stability value, and usage time value. The initial weight parameter set specifically includes parameter correlation, cumulative change rate, and fuzzy correlation matrix. The dynamic weight parameter set includes environmental fluctuation value, offset adjustment value, and weight category. The multi-level performance clustering structure includes group offset, weight matching parameter, and clustering verification result. The classification management dynamic adjustment value set specifically includes classification boundary parameter, dynamic weight adjustment value, and synchronous optimization and dynamic adjustment result.

[0073] Please see Figure 2 The specific steps for obtaining the multidimensional performance standardization set are as follows:

[0074] Based on cloud data streams, the chip brightness value, voltage stability value, and usage time value are standardized. The chip brightness value and voltage stability value are combined after being normalized by the range of their respective distribution ranges to generate light source and driving parameters. The usage time value is normalized by the ratio of cumulative usage time to normal usage to generate lifespan parameters.

[0075] Firstly, the normalization process for the chip's brightness value requires collecting brightness range data from the chip's light source. The normalization ratio is then calculated using the difference between the maximum and minimum brightness values. The calculation formula is as follows: Where L i L represents the current brightness value. max L represents the maximum brightness value. min To determine the minimum brightness value, the same operation was performed on the stable voltage value, collecting the voltage fluctuation range under stable conditions. The voltage upper and lower limit differences were then normalized using the formula: Where V i The current voltage value, V max V is the maximum voltage. min The minimum voltage value is used as the basis for normalizing both values ​​and combining them to generate the light source and driving parameters. Next, the chip usage time value is normalized by collecting actual usage time and calculating the ratio based on the chip's design life standard. The formula is as follows: Where T i T represents the current usage duration. norm To design the standard lifespan of the chip, the duration is normalized to generate lifespan parameters, and then standardized results are generated by combining the light source and driving parameters with the lifespan parameters.

[0076] Using light source and driving parameters as well as lifetime parameters, the system performs layered processing based on their distribution characteristics in multidimensional space. The distribution center point of each layer is calculated by calling the mean and standard deviation of the parameter value distribution. The layering range is determined by comparing the distance between the center points, and an initial set of multi-layer performance is generated.

[0077] By collecting the distribution range of each parameter in multidimensional space, calculating the distribution center point, and performing layered processing, firstly, for the light source and driving parameters, the mean values ​​of the chip's normalized brightness and voltage are used to calculate the parameter distribution center point, using the following formula: Where P i Here, n represents the normalized parameter value, and n is the number of parameters. The distribution range of the lifetime parameter is calculated similarly, using the mean of the normalized parameter values ​​to calculate the distribution center point. The calculated distribution center points of each parameter are compared, and the stratification boundary in the multidimensional space is calculated by defining the standard deviation of the distribution range. The formula is as follows: Where R is the distribution range and C is the center point. The layer range is determined based on the distribution range and the position of the center point, and the initial set of multi-layer performance is generated.

[0078] The parameters within the initial set of multi-level performance are grouped, and the correlation index of the parameters within each group is judged to meet the minimum threshold condition using the following formula:

[0079]

[0080] Calculate the comprehensive score for each group, filter the groups based on the goodness-of-fit threshold of the group index values, and generate a multidimensional performance standardized set;

[0081] Where G represents the combined score of the parameters after grouping, P i w represents the normalized value of the i-th parameter in the group. i This represents the weight coefficient of the corresponding parameter, where i is the index of the parameter in the current group and n is the total number of parameters in the current group.

[0082] formula:

[0083]

[0084] The advantage of the formula is that by combining the parameter values ​​with their weights and taking into account the normalization of the weights, it can effectively calculate the comprehensive performance score of multidimensional data, making the comparison between parameters more accurate.

[0085] Detailed explanation of the formula and its calculation derivation:

[0086] Set the grouping parameter normalization value to P. i The normalized parameter values ​​collected from a certain group are: P1 = 0.75, P2 = 0.65, P3 = 0.85; the weight parameter w is set.i The values ​​are w1 = 0.5, w2 = 0.3, and w3 = 0.2, respectively.

[0087] First, calculate the sum of the weighted values ​​for the numerator, using the following formula:

[0088]

[0089] Next, calculate the square root of the sum of squares of the weighted parts in the denominator, using the following formula:

[0090]

[0091] Finally, divide the numerator by the denominator to calculate the overall score:

[0092]

[0093] The results show that the overall score G = 1.2 reflects the overall performance level of the grouped parameters. The higher overall score indicates that the normalized values ​​of the parameters in the group match the weight values ​​well. In the generated multidimensional performance standardized set, the performance parameters of this group are superior.

[0094] Please see Figure 3 The specific steps for obtaining the initial weight parameter set are as follows:

[0095] Based on the chip brightness value and voltage stability value in the multidimensional performance standardization set, the correlation parameter is calculated by the difference between the corresponding positions of the two. It is then determined whether the absolute value of the difference between each pair of parameters meets the specified threshold range, and the parameter combinations that meet the conditions are included in the chip brightness and voltage stability difference correlation parameter set.

[0096] The original recorded data of the chip's brightness and voltage stability values ​​are retrieved, and the difference between corresponding positions is calculated point by point. The absolute value is then calculated using the formula |P|. i -Q j | Obtain the absolute value of the difference to determine its magnitude. Through a threshold-based filtering mechanism, retain parameter pairs whose absolute values ​​fall within a preset range. Perform difference calculations and judgments on all data in sequence, and combine the data that meet the conditions to form a set of parameters related to the difference between chip brightness and voltage stability.

[0097] The cumulative change rate of the usage duration value and the chip brightness value is calculated. The change rate parameter is generated by standardizing the ratio of the cumulative change value to the initial parameter value. The linear correlation coefficient between the change rate and the cumulative parameter is calculated. It is determined whether the correlation coefficient meets the set range. Parameters that meet the conditions are screened into the usage duration and brightness change rate parameter set.

[0098] By calling the hourly records of cumulative usage time and the corresponding change records of chip brightness value, the difference between the changes in usage time and brightness is calculated respectively. The cumulative change rate formula (current value - initial value) / initial value is applied to obtain the change rate for each time period. The change rate is standardized to the [0,1] interval. The relationship coefficient between the change rate and usage time is calculated through linear correlation analysis, and it is determined whether the set correlation threshold is met. Parameters that meet the threshold are screened into the correlation set to generate a parameter set of usage time and brightness change rate.

[0099] The parameter sets relating chip brightness to voltage stability difference and usage time to brightness change rate are merged using the following formula:

[0100]

[0101] Calculate the comprehensive correlation value of the positions of multiple matrices and generate a fuzzy correlation matrix;

[0102] Among them, M ij P represents the comprehensive correlation value between the i-th row and the j-th column in the fuzzy correlation matrix. i Q represents the chip's brightness value. j Represents the voltage stability value, R ij The cumulative rate of change correlation weights are defined as follows: i is the brightness value index, j is the voltage stability value index, n is the total number of parameters, and P is the cumulative rate of change correlation weights. k Q represents the k-th parameter value in the chip's brightness values. k Represents the k-th parameter value in the voltage stability value, |P i -Q j | represents the absolute difference between the chip's brightness value and its voltage stability value;

[0103] formula:

[0104]

[0105] The advantage of the formula is that by combining the absolute value of the difference between the chip brightness value and the voltage stability value with the correlation weight of the rate of change, the accuracy of the comprehensive correlation in the fuzzy correlation matrix is ​​improved, and the denominator normalization process ensures the fusion of data with different dimensions.

[0106] Detailed explanation of the formula and its calculation derivation:

[0107] Let the chip brightness value P i =0.8, 0.85, 0.9, 1.0, 1.05, voltage stability values

[0108] Q j =1.0, 1.1, 1.2, 0.95, 1.0, cumulative rate of change weight R ij=0.8,0.9,0.95,1.0,0.85, calculate the positional correlation of the first position in the matrix:

[0109]

[0110] Calculate the numerator and denominator step by step:

[0111] molecular:

[0112] P1-Q1|=|0.8-1.0|=0.2,0.2·0.8=0.16;

[0113] Denominator:

[0114]

[0115] Calculate the square value for each item:

[0116] 0.64 + 0.7225 + 0.81 + 1.0 + 1.1025 + 1.0 + 1.21 + 1.44 + 0.9025 + 1.0 = 9.8275;

[0117] Denominator value:

[0118]

[0119] Complete calculation:

[0120]

[0121] Repeat the above calculation process sequentially to complete the calculation of the comprehensive correlation value for each position in the matrix. The integrated result is as follows:

[0122]

[0123] The results show that the comprehensive correlation value in the fuzzy correlation matrix, through normalization calculation and weighting, reflects the comprehensive coupling relationship between chip brightness value and voltage stability value in terms of rate of change and absolute difference, providing basic data support for subsequent matrix normalization processing and result screening.

[0124] Normalize the comprehensive correlation values ​​in the fuzzy correlation matrix, filter the matrix parameters whose comprehensive correlation values ​​meet the preset threshold range, calculate the weight normalization values ​​of the filtered matrix parameters, call the normalization values ​​and the initial parameter values ​​of the corresponding matrix positions, and generate the initial weight parameter set.

[0125] By extracting the maximum value of all correlation values ​​in the matrix, the ratio of each correlation value to the maximum value is calculated to complete the normalization process. All parameter values ​​in the normalization matrix are traversed sequentially. By setting a threshold range and judging whether the normalized value falls within the threshold range, parameter combinations that meet the range are filtered out. At the same time, the chip brightness value, voltage stability value, and cumulative change rate value at the corresponding matrix position are called. The brightness value, voltage value, and change rate of the filtered results are integrated and grouped. The grouping results and normalized values ​​are integrated into an initial weight parameter set.

[0126] Please see Figure 4 The specific steps for obtaining the dynamic weight parameter set are as follows:

[0127] Based on the chip brightness value, voltage stability value and external ambient temperature value in the initial weight parameter set, the difference between the chip brightness value and the external ambient temperature value and the difference between the voltage stability value and the external ambient temperature value are calculated respectively. Parameter combinations that satisfy the absolute value of the difference within the set threshold range are selected to generate an environmental fluctuation difference parameter set.

[0128] The chip brightness value is processed by segmenting and normalizing the light source brightness information extracted from the multidimensional performance set. The voltage stability value is calculated by calculating its variation range through the fluctuation interval and extracting the stable value distribution. The external ambient temperature value is aligned with the average temperature during the actual monitoring period through the daily temperature change curve. The difference between the chip brightness value and the external ambient temperature value, as well as the difference between the voltage stability value and the external ambient temperature value, are calculated respectively. The difference is used to determine whether the condition of the interval threshold range is met. The difference combination that meets the condition is screened and used to construct the environmental fluctuation parameter group, thus constructing the environmental fluctuation difference parameter set.

[0129] Extract usage duration values ​​and environmental fluctuation differences from the environmental fluctuation difference parameter set, calculate the adjustment offset value of the environmental fluctuation difference based on the changing trend of usage duration values, and generate an offset adjustment parameter set by using the cumulative changing trend and normalized ratio.

[0130] The usage duration value is extracted and normalized from the cumulative usage duration in the equipment's work records. It is then subjected to time series fitting analysis with the environmental fluctuation difference to calculate the adjustment impact of the environmental fluctuation difference on the usage duration change at each time point. The adjustment offset value is obtained by multiplying the fluctuation difference by the rate of change of usage duration. The distribution of the offset value is normalized to generate an offset adjustment parameter set.

[0131] The offset adjustment parameter set is used to match the weight categories. This is achieved by comprehensively calculating the offset adjustment value, weight parameter value, and environmental fluctuation parameter using the following formula:

[0132]

[0133] The combined value of multiple matching categories is called to calculate and generate a dynamic weight parameter set;

[0134] Among them, W dyn P represents the dynamic weight parameter. i T represents the offset adjustment parameter. j For the corresponding weight category, The sum of squares of environmental fluctuation parameters. This is a normalized calculation of the sum of squares of the weight values ​​and the offset values, where i, j, k, and m are parameter indices, and n is the total number of parameters.

[0135] formula:

[0136]

[0137] The advantage of the formula is that by combining the sum of the squares of the offset adjustment value, the weight category parameter, and the environmental fluctuation parameter, it fully reflects the comprehensive influence of multiple parameters in the dynamic weight calculation, making the weight parameters more accurate under fluctuating conditions.

[0138] Detailed explanation of the formula and its calculation derivation:

[0139] Assume P i The offset adjustment parameter is generated by normalizing the offset adjustment value, specifically 0.12, T. j For the weighted category parameters, the standard normalization value for category matching is 0.85, R0 k The environmental fluctuation parameter, calculated using the sum of squares, has a value of 3.5. The sum of the squares of the offset adjustment parameters and the sum of the squares of the weights is 8.1. Substitute this value into the formula:

[0140] first step:

[0141] (P i ·T j ) = 0.12 * 0.85 = 0.102;

[0142] Step Two:

[0143]

[0144] Step 3:

[0145]

[0146] Comprehensive calculation:

[0147]

[0148] The results show that the dynamic weight parameter is 1.265, which reflects the comprehensive weight value of multiple parameters under the current environmental fluctuation conditions. The parameter result is used to match the corresponding dynamic weight category to achieve weight allocation.

[0149] Please see Figure 5 The specific steps for obtaining the multi-level performance clustering structure are as follows:

[0150] Based on the light source and driver category parameters in the dynamic weight parameter set, the chip brightness value and voltage stability value are extracted. The chip brightness value and voltage stability value are sorted according to the light source and driver category. The sorting difference value within the group is calculated and the grouping of the difference value meets the target threshold range is selected to generate the sorting result of the light source and driver category.

[0151] By extracting the brightness and voltage stability values ​​of each chip individually, and grouping them according to light source and driver categories, the ranking difference value of the chip brightness and voltage stability values ​​within each group is calculated. The calculation method involves summing the absolute values ​​of the differences between each pair of ranked values ​​and normalizing the results to reduce the impact of group size on the difference calculation. Taking the chip brightness value as an example, assuming the parameter ranking within a group is P1, P2, P3, the calculation formula is as follows: The group ranking difference value is calculated using a normalization formula, where n is the number of chip brightness values ​​in the current group. After completing the above steps, the group ranking difference value is calculated by repeating the above process with the voltage stability value. The group ranking result of light source and driver category is compared with the threshold range to determine whether the ranking difference value meets the set conditions. The threshold is calculated by calling the mean and standard deviation of the ranking difference value distribution in the historical light source and driver category. In the example, the ranking difference threshold range is set to 0.1 and 0.5. Groups that meet the conditions are filtered to construct a group ranking result with higher ranking stability, and the group ranking result of light source and driver category is generated.

[0152] Extract the usage duration value of the lifespan category from the grouping and sorting results of the light source and drive categories, calculate the grouping difference of the usage duration value within the group to the light source and drive categories, and through weighted cumulative analysis of the grouping difference and lifespan category offset parameter, filter the groups whose difference offset values ​​meet the threshold range, and generate the grouping offset and weight matching parameter.

[0153] The correlation coefficient is obtained through the formula The calculation shows that x represents the usage duration value of the lifespan category, and y represents the difference in ranking between the light source and the drive category. and Let x and y be the means, respectively. The ∑ symbol represents the summation of each data point within a group. The calculated correlation coefficient is then weighted and accumulated with the group offset parameter, which is the ratio of each value within a group to its mean. The formula for calculating this parameter is: Finally, by judging whether the weighted cumulative offset value meets the preset threshold condition, the results that meet the group difference condition are filtered out, and the group offset and weight matching parameters are generated.

[0154] Using the group offset and weight matching parameters, a weighted normalized index of the in-group offset value and dynamic weight parameters is calculated, using the formula:

[0155]

[0156] Calculate multi-level performance clustering metrics, integrate the clustering metrics with dynamic weight parameters, and generate a multi-level performance clustering structure.

[0157] Among them, C p A represents a multi-level performance clustering metric. i B represents the offset parameter within the group. i D represents the dynamic weight parameter. i R represents the group difference value. i The association weights for the offset parameters, |A i ·B i +D i | represents the absolute value of the offset parameter, weight parameter, and difference value. The sum of squares of differences between groups and the sum of squares of weights are calculated, where k is the total number of cluster groups, i and j are parameter indices, and n and m are the total number of parameters within and between groups.

[0158] formula:

[0159]

[0160] The advantage of the formula is that by combining the group offset parameter with the dynamic weight parameter, it comprehensively considers the normalized weight accumulation effect of the offset value, weight parameter and inter-group differences, and realizes the optimized analysis of the group offset for weight rationality matching verification.

[0161] Detailed explanation of the formula and its calculation derivation:

[0162] Set group offset parameter A i = [0.15, 0.20, 0.25, 0.10], dynamic weight parameter B i =[1.1,1.2,0.9,1.3], Group difference value D i = [0.05, 0.07, 0.02, 0.09], correlation weight R i =[0.8,0.9,0.7,0.85], inter-group differences The total number of cluster groups k = 4. First, calculate the numerator:

[0163] The numerator is calculated as: |0.15·1.1+0.05|·0.8+|0.20·1.2+0.07|·0.9+|0.25·0.9+0.02|·0.7+|0.10·1.3+0.09|·0.85.

[0164] = (0.165·0.8)+(0.31·0.9)+(0.245·0.7)+(0.22·0.85);

[0165] = 0.132 + 0.279 + 0.1715 + 0.187;

[0166] =0.7695;

[0167] Calculation of the denominator:

[0168]

[0169] Calculation formula result:

[0170]

[0171] The results show that the clustering validation index generated by combining group offset and weight parameters is 0.0833, which reflects that the performance matching between the current groups is close to the set conditions and can be further used to generate multi-level performance clustering structures.

[0172] Please see Figure 6 The specific steps for obtaining the dynamically adjusted value set for category management are as follows:

[0173] Clustering boundary parameters for chip brightness, voltage stability, and usage time are extracted from the multi-level performance clustering structure. By comparing the distribution range of the multi-class boundary parameters, the initial parameter range of the classification boundary is determined, and the clustering boundary parameters are generated.

[0174] By sequentially analyzing the classification characteristics of chip brightness and voltage stability values, each data group is categorized into boundary parameters based on its actual distribution range. The upper and lower boundaries of each group are analyzed in detail. The average difference and dispersion of brightness values ​​in each group are calculated using statistical methods. These results serve as the basic parameters for the upper and lower bounds of the classification. Voltage stability values ​​undergo similar processing. By collecting the fluctuation range of actual operating voltage under standard conditions, the minimum and maximum voltage ranges under different operating modes are calculated. Simultaneously, based on usage time, lifespan performance is divided according to different modes. By analyzing the peak and critical points of lifespan performance in the data groups, the boundary lifespan range of each group is calculated, and the corresponding upper and lower boundary values ​​are extracted. The comprehensive results of these classification boundaries are labeled with light source, driver, and lifespan category parameters to form clustering classification boundary parameters.

[0175] The dynamic weight adjustment value is interactively analyzed with the clustering classification boundary parameters. By calculating the offset correction amount of the dynamic weight for each classification boundary value, the corrected classification boundary value set is called to generate the corrected classification boundary parameters.

[0176] First, based on the upper and lower limit data in the clustering classification boundary parameters, and combined with the dynamic weight values, the correction calculation is performed on each boundary point. Using the range of change of each group of dynamic weight values, the offset correction magnitude of each classification boundary parameter is calculated in turn. For the calculation of the offset magnitude, the product of the clustering classification boundary parameter and the dynamic weight value is weighted and accumulated to generate the offset correction amount of each classification boundary. The corrected upper and lower boundaries are respectively assigned to the new classification boundary parameters. At the same time, the results of each classification group are filtered to exclude the upper and lower limit data that do not meet the requirements of the adjusted classification interval. All corrected classification upper and lower limit values ​​are called to form the complete corrected classification boundary parameters.

[0177] By combining the corrected classification boundary parameters with dynamic weight adjustment values, synchronous optimization of light source, drive, and lifetime categories is achieved, using the formula:

[0178]

[0179] Calculate the dynamic adjustment index for the classification boundary, integrate the adjustment index into the corrected classification boundary value, and generate a set of dynamic adjustment values ​​for classification management.

[0180] Among them, T opt A represents the dynamic adjustment index of the classification boundary. i B represents the corrected classification boundary parameter. i C represents the dynamic weight adjustment value. i For boundary offset correction, |A i -B i To correct the absolute deviation between the boundary and the weight adjustment value, F k is the optimized clustering index for the classification groups, p is the total number of classification groups, i, j, k are parameter indices, and n, m, p are the total number of parameters within, between, and within the classification groups, respectively.

[0181] formula:

[0182]

[0183] The advantage of the formula is that by simultaneously integrating and correcting the classification boundary parameters, dynamic weight values, and classification interval adjustment results, it improves the dynamic adaptability of the classification boundary and achieves multi-dimensional optimization of the overall parameters through cluster optimization results, thereby improving classification accuracy.

[0184] Detailed explanation of the formula and its calculation derivation:

[0185] A iThe corrected classification boundary parameters are calculated by combining the initial upper and lower limit data of each classification boundary in the actual environment with the offset correction amount of the dynamic weight. For example, if the upper and lower limits of the classification boundary brightness parameters are 220 and 300 respectively, and the dynamic weight offset is 10, then the corrected upper and lower limits are 220+10 and 300-10.

[0186] B i This represents the dynamic weight adjustment value. The dynamic weight value is calculated based on the dynamic performance of historical light source parameters. For example, the dynamic weight value range is adjusted to [0.8, 1.2] based on the normalized brightness performance range.

[0187] C i The offset correction amount represents the classification boundary. This offset correction amount is calculated by the absolute difference between the upper and lower bounds and the dynamic weight adjustment values, such as |A i -B i The offset result of | is 20, and the squared cumulative correction amount is 400;

[0188] D j and E j These represent the differences within and between classification groups, respectively. They were calculated using the sum of squares of the brightness differences within groups and the sum of squares of the parameter differences between groups. The sum of squares of the brightness differences within groups [5, 10, 15] was 350, and the sum of squares of the brightness differences between groups [20, 25, 30] was 1375.

[0189] F k The clustering optimization index is calculated by considering the clustering interval and intra-group consistency of multiple light sources. For example, the total optimization value is 2.5 obtained by weighted averaging the optimization values ​​of each interval.

[0190] Formula calculation process:

[0191] Substitute in specific values:

[0192]

[0193] in:

[0194] A1=230, A2=240, A3=250, B1=220, B2=230, B3=240, C1=10, C2=

[0195] 10, C3 = 10, F1 = 2.0, F2 = 2.5, F3 = 3.0.

[0196] calculate:

[0197]

[0198] T opt=6.77 + 2.5 = 9.27;

[0199] The results show that after the modified classification boundary parameters are optimized and adjusted by dynamic weights, the classification accuracy is improved. The resulting set of dynamically adjusted classification values ​​can meet the dynamic change requirements of different parameters, and at the same time, it has good adaptability to classification stability and consistency.

[0200] The cloud-based LED tube quality management method, implemented based on the aforementioned cloud-based LED tube quality management system, includes the following steps:

[0201] S1: Based on cloud data stream, collect chip brightness value, voltage stability value, and usage duration parameters. Classify chip brightness value and voltage stability value as light source and driving parameters, and classify usage duration value as lifespan parameter. Perform standardization operations through data classification, and perform data layering and grouping according to parameter characteristics to generate a multi-dimensional performance standardization set.

[0202] S2: Based on the multidimensional performance standardization set, calculate the difference between the chip brightness value and voltage stability value to obtain the correlation value. At the same time, calculate the cumulative change rate of usage time and brightness value, merge the two into a fuzzy correlation matrix, normalize the matrix data, and generate an initial weight parameter set.

[0203] S3: Based on the initial weight parameter set, analyze the deviation between the external ambient temperature and the chip brightness and voltage value, extract the offset adjustment value, normalize all offset values, classify the weight categories according to the results, and generate a dynamic weight parameter set;

[0204] S4: Based on the dynamic weight parameter set, perform grouping and sorting of chip brightness value and voltage stability value, calculate the difference between usage duration value groups, verify data offset according to weight matching parameters, and generate multi-level performance clustering structure through clustering verification process;

[0205] S5: Based on a multi-level performance clustering structure, analyze the classification boundaries of chip brightness, voltage stability, and usage time values ​​in the clusters, perform dynamic weight adjustment, simultaneously optimize the classification boundaries of light source, driver, and lifespan categories, and generate a set of dynamic adjustment values ​​for classification management.

[0206] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A cloud-based LED tube quality management system, characterized in that, The system includes: The data standardization processing module acquires chip brightness, voltage stability, and usage duration parameters based on cloud data streams, standardizes the parameter range, classifies chip brightness and voltage stability values ​​into light source and drive parameters, classifies usage duration values ​​into lifespan parameters, and hierarchically and group the classified data to generate a multi-dimensional performance standardization set. The parameter correlation and fuzzy matrix generation module, based on the multidimensional performance standardization set, extracts the chip brightness value and voltage stability value and performs difference calculation to obtain parameter correlation. It calculates the cumulative change rate of the usage duration value and chip brightness value, merges the correlation and change rate to form a fuzzy correlation matrix, normalizes the matrix data, and generates an initial weight parameter set. Based on the initial weight parameter set, the dynamic weight generation module performs difference analysis by combining the external environment temperature change with the chip brightness value and voltage stability value, extracts the offset adjustment value between the environmental fluctuation value and the usage time value, normalizes all offset values ​​and matches the weight category, and generates a dynamic weight parameter set. The multi-level performance clustering module, based on the dynamic weight parameter set, groups and sorts the chip brightness values ​​and voltage stability values ​​of the light source and driver categories, calculates the group differences by combining the usage time values ​​of the lifespan category, extracts the group offset and weight matching parameters to verify the clustering, and generates a multi-level performance clustering structure. The classification boundary optimization and dynamic adjustment module, based on the multi-level performance clustering structure, extracts the classification boundary parameters of chip brightness value, voltage stability value, and usage time value from the clustering results, and corrects the classification boundary by combining the dynamic weight adjustment value, thereby completing the synchronous optimization and dynamic adjustment of light source, driver, and lifespan categories, and generating a set of dynamic adjustment values ​​for classification management.

2. The cloud-based LED tube quality management system according to claim 1, characterized in that, The multidimensional performance standardization set includes chip brightness value, voltage stability value, and usage time value. The initial weight parameter set specifically includes parameter correlation, cumulative change rate, and fuzzy correlation matrix. The dynamic weight parameter set includes environmental fluctuation value, offset adjustment value, and weight category. The multi-level performance clustering structure includes group offset, weight matching parameter, and clustering verification result. The classification management dynamic adjustment value set specifically includes classification boundary parameter, dynamic weight adjustment value, and synchronous optimization and dynamic adjustment result.

3. The cloud-based LED tube quality management system according to claim 2, characterized in that, The specific steps for obtaining the multidimensional performance standardization set are as follows: Based on cloud data streams, the chip brightness value, voltage stability value, and usage time value are standardized. The chip brightness value and voltage stability value are combined after being normalized by the range of their respective distribution ranges to generate light source and driving parameters. The usage time value is normalized by the ratio of cumulative usage time to normal usage to generate lifespan parameters. Using the light source and driving parameters and the lifetime parameters, the layering process is performed based on their distribution characteristics in multidimensional space. The distribution center point of each layer is calculated by calling the mean and standard deviation of the parameter value distribution. The layering range is determined by comparing the distance between the center points, and an initial set of multi-layer performance is generated. The parameters within the initial set of multi-layer performance are grouped, and the correlation index of the parameters within each group is judged to meet the minimum threshold condition using the following formula: Calculate the comprehensive score for each group, filter the groups based on the goodness-of-fit threshold of the group index values, and generate a multidimensional performance standardized set; Where G represents the combined score of the parameters after grouping, P i w represents the normalized value of the i-th parameter in the group. i This represents the weight coefficient of the corresponding parameter, where i is the index of the parameter in the current group and n is the total number of parameters in the current group.

4. The cloud-based LED tube quality management system according to claim 3, characterized in that, The specific steps for obtaining the initial weight parameter set are as follows: Based on the chip brightness value and voltage stability value in the multidimensional performance standardization set, the correlation parameter is calculated by the difference operation of the corresponding positions of the two, and it is determined whether the absolute value of the difference of each pair of parameters meets the specified threshold range. The parameter combinations that meet the conditions are included in the chip brightness and voltage stability difference correlation parameter set. The cumulative change rate of the usage duration value and the chip brightness value is calculated. The change rate parameter is generated by standardizing the ratio of the cumulative change value to the initial parameter value. The linear correlation coefficient between the change rate and the cumulative parameter is calculated. It is determined whether the correlation coefficient meets the set range. Parameters that meet the conditions are screened into the usage duration and brightness change rate parameter set. The parameter set relating chip brightness and voltage stability difference and the parameter set relating usage duration and brightness change rate are merged using the following formula: Calculate the comprehensive correlation value of the positions of multiple matrices and generate a fuzzy correlation matrix; Among them, M ij P represents the comprehensive correlation value between the i-th row and the j-th column in the fuzzy correlation matrix. i Q represents the chip's brightness value. j Represents the voltage stability value, R ij The cumulative rate of change correlation weights are given by: i being the brightness value index, j being the voltage stability value index, n being the total number of parameters, and P. k Q represents the k-th parameter value in the chip's brightness values. k Represents the k-th parameter value in the voltage stability value, |P i -Q j | represents the absolute difference between the chip's brightness value and its voltage stability value; Normalization is performed on the comprehensive correlation values ​​in the fuzzy correlation matrix. Matrix parameters whose comprehensive correlation values ​​meet the preset threshold range are selected. The weight normalization values ​​of the selected matrix parameters are calculated. The normalization values ​​and the initial parameter values ​​of the corresponding matrix positions are called to generate an initial weight parameter set.

5. The cloud-based LED tube quality management system according to claim 4, characterized in that, The specific steps for obtaining the dynamic weight parameter set are as follows: Based on the chip brightness value, voltage stability value and external ambient temperature value in the initial weight parameter set, the difference between the chip brightness value and the external ambient temperature value and the difference between the voltage stability value and the external ambient temperature value are calculated respectively. Parameter combinations that satisfy the absolute value of the difference within the set threshold range are selected to generate an environmental fluctuation difference parameter set. Extract usage duration values ​​and environmental fluctuation differences from the set of environmental fluctuation difference parameters, calculate the adjustment offset value of the change trend of the usage duration values ​​to the environmental fluctuation difference, and generate an offset adjustment parameter set by accumulating the change trend and normalization ratio. The offset adjustment parameter set is used to match weight categories. This is achieved by comprehensively calculating the offset adjustment value, weight parameter value, and environmental fluctuation parameter using the following formula: The combined value of multiple matching categories is called to calculate and generate a dynamic weight parameter set; Among them, W dyn P represents the dynamic weight parameter. i T represents the offset adjustment parameter. j For the corresponding weight category, The sum of squares of environmental fluctuation parameters. This is a normalized calculation of the sum of squares of the weight values ​​and the offset values, where i, j, k, and m are parameter indices, and n is the total number of parameters.

6. The cloud-based LED tube quality management system according to claim 5, characterized in that, The specific steps for obtaining the multi-level performance clustering structure are as follows: Based on the light source and driver category parameters in the dynamic weight parameter set, the chip brightness value and voltage stability value are extracted. The chip brightness value and voltage stability value are sorted according to the light source and driver category. The sorting difference value within the group is calculated and the grouping of the difference value that meets the target threshold range is selected to generate the sorting result of the light source and driver category. Extract the usage duration value of the lifespan category from the grouping and sorting results of the light source and drive categories, calculate the grouping difference of the usage duration value within the group with respect to the light source and drive categories, and through weighted cumulative analysis of the grouping difference and lifespan category offset parameter, filter the groups whose difference offset values ​​meet the threshold range, and generate grouping offset and weight matching parameters. Using the group offset and weight matching parameters, a weighted normalized index of the intra-group offset value and dynamic weight parameters is calculated, using the formula: Calculate multi-level performance clustering metrics, integrate the clustering metrics with dynamic weight parameters, and generate a multi-level performance clustering structure. Among them, C p A represents a multi-level performance clustering metric. i B represents the offset parameter within the group. i D represents the dynamic weight parameter. i R represents the group difference value. i The association weights for the offset parameters, |A i ·B i +D i | represents the absolute value of the offset parameter, weight parameter, and difference value. The summation of the sum of squares of differences between groups and the sum of squares of weights is calculated, where k is the total number of cluster groups, i and j are parameter indices, and n and m are the total number of parameters within and between groups.

7. The cloud-based LED tube quality management system according to claim 6, characterized in that, The specific steps for obtaining the dynamically adjusted value set for classification management are as follows: The clustering boundary parameters of chip brightness value, voltage stability value and usage time value are extracted from the multi-level performance clustering structure. By comparing the distribution range of the multi-classification boundary parameters, the initial parameter range of the classification boundary is determined, and the clustering classification boundary parameters are generated. The dynamic weight adjustment value is interactively analyzed with the clustering classification boundary parameters. By calculating the offset correction amount of the dynamic weight for each classification boundary value, the corrected classification boundary value set is called to generate the corrected classification boundary parameters. By combining the corrected classification boundary parameters and dynamic weight adjustment values, synchronous optimization of light source, drive, and lifetime categories is achieved, using the formula: Calculate the dynamic adjustment index for the classification boundary, integrate the adjustment index into the corrected classification boundary value, and generate a set of dynamic adjustment values ​​for classification management. Among them, T opt A represents the dynamic adjustment index of the classification boundary. i B represents the corrected classification boundary parameters. i C represents the dynamic weight adjustment value. i For boundary offset correction, |A i -B i To correct the absolute deviation between the boundary and the weight adjustment value, F k is the optimized clustering index for the classification groups, p is the total number of classification groups, i, j, k are parameter indices, and n, m, p are the total number of parameters within, between, and within the classification groups, respectively.

8. A cloud-based LED tube quality management method, characterized in that, The cloud-based LED tube quality management system according to any one of claims 1-7 includes the following steps: S1: Based on cloud data stream, collect chip brightness value, voltage stability value, and usage duration parameters. Classify chip brightness value and voltage stability value as light source and driving parameters, and classify usage duration value as lifespan parameter. Perform standardization operations through data classification, and perform data layering and grouping according to parameter characteristics to generate a multi-dimensional performance standardization set. S2: Based on the multidimensional performance standardization set, calculate the difference between the chip brightness value and voltage stability value to obtain the correlation value. At the same time, calculate the cumulative change rate of usage time and brightness value, merge the two into a fuzzy correlation matrix, normalize the matrix data, and generate an initial weight parameter set. S3: Based on the initial weight parameter set, analyze the deviation between the external ambient temperature and the chip brightness and voltage value, extract the offset adjustment value, normalize all offset values, classify weight categories according to the results, and generate a dynamic weight parameter set. S4: Based on the dynamic weight parameter set, perform grouping and sorting of chip brightness value and voltage stability value, calculate the difference between usage duration value groups, verify data offset according to weight matching parameter, and generate multi-level performance clustering structure through clustering verification process; S5: Based on the multi-level performance clustering structure, analyze the classification boundaries of chip brightness value, voltage stability value and usage time value in the cluster, perform dynamic weight adjustment, simultaneously optimize the classification boundaries of light source, driver and lifespan categories, and generate a set of dynamic adjustment values ​​for classification management.

Images