A product quality tracing method and system
By obtaining and analyzing the historical data of sweater products, building and updating quality detection functions, and adjusting the sampling rate, the problem of ignoring other product quality in batch coding methods is solved, and the accuracy and credibility of sweater product quality traceability is improved.
Patent Information
- Application Number
- CN202410699845.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-31
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-05-31
AI Technical Summary
The existing batch coding method ignores the quality impact of other products except the sampled products in the same batch in the sweater product quality traceability, resulting in low accuracy and reliability of quality traceability results.
By obtaining the raw material quality evaluation indicators, process parameter change characteristic values and quality inspection evaluation values of historical batches, the initial product quality detection function is constructed, and the function is updated according to the characteristic values and fluctuations of process parameters change, and the product expected quality detection function is obtained. , adjust the sampling rate to obtain traceability code information and trace product quality.
It improves the accuracy and credibility of product quality traceability, enhances the reliability of the expected results of finished product quality, improves the detection efficiency, and effectively characterizes the quality information of all products in the batch products.
Smart Images

Figure CN118485452B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of product traceability, and in particular to a product quality traceability method and system. Background Art
[0002] With the continuous development of industrial production, product quality control and traceability have become an important part of the production process of enterprises. Product quality traceability refers to tracking and recording the quality information of products in various links such as manufacturing, processing, distribution and sales through certain methods and technologies, so that quality problems can be quickly defined and solved when necessary to ensure product quality and safety. As a common clothing category, sweater products have a wide range of consumer demand in the market. The production process of sweater products involves multiple links, including raw material procurement, spinning, weaving, dyeing, knitting, finishing, etc., so the quality traceability of sweater products is particularly important.
[0003] The existing method of batch coding is to achieve the quality traceability of sweater products. Batch coding means to uniformly evaluate all products in the current batch according to the sampling test results, and batch code based on the evaluation results. However, in the process of batch coding, sampling testing only evaluates the quality of sampled products, ignoring the impact of the quality of other products in the same batch except the sampled products, resulting in low credibility of the batch coding method used to achieve product quality traceability, and thus low accuracy and reliability of product quality traceability results. Summary of the invention
[0004] In order to solve the technical problem that the above-mentioned batch coding method ignores the influence of the quality of other products in the same batch except the sampled products, resulting in low accuracy of product quality traceability results, the purpose of the present invention is to provide a product quality traceability method and system, and the technical solutions adopted are as follows:
[0005] An embodiment of the present invention provides a product quality tracing method and system, the method comprising the following steps:
[0006] Obtain the raw material quality evaluation indexes of several historical batches of products at the raw material procurement stage, the process parameter change characteristic values at the production and processing stage, the quality inspection evaluation value of each product at the quality inspection stage, and obtain the quality inspection evaluation value of each sampled product at the quality inspection stage for the current batch of products;
[0007] Determine the degree of compensation for raw material quality caused by process parameter changes based on the raw material quality evaluation index and process parameter change characteristic value of each historical batch of products;
[0008] According to the quality inspection evaluation value of each product in each historical batch of products, determine the degree of fluctuation of the quality inspection evaluation values of all products in each historical batch of products, and construct the product initial quality detection function;
[0009] According to the process parameter change characteristic value of each historical batch of products, the fluctuation degree and the compensation degree, the product initial quality detection function is updated to obtain the product expected quality detection function;
[0010] According to the batch number of the current batch of products, the expected quality detection function of the product and the quality inspection evaluation value of each sampled product in the current batch of products, the adjusted sampling rate is obtained, and then the traceability code information of the current batch of products when batch coding is obtained to trace the product quality.
[0011] Furthermore, the step of obtaining the raw material quality evaluation index includes:
[0012] For any historical batch of products, obtain the parameter set of the historical batch of products at the raw material procurement stage, use the parameter set of the raw material procurement stage as input data, input it into a pre-built and trained raw material quality evaluation model, and obtain the raw material quality evaluation index of the historical batch of products at the raw material procurement stage.
[0013] Furthermore, the step of obtaining the process parameter variation characteristic value includes:
[0014] For any historical batch of products, obtain each type of process parameter in the parameter set of the historical batch of products in the production and processing stage, perform curve fitting on the same type of process parameters in the parameter set of the historical batch of products in the production and processing stage, and obtain each type of process parameter curve in the time series;
[0015] Calculate the variance of each type of process parameter curve, and use the average value of all variances as the first variation factor of the process parameter; calculate the similarity between each pair of process parameter curves, and use the average value of all similarities as the second variation factor of the process parameter;
[0016] Combined with the first change factor and the second change factor of the process parameter, determine the change characteristic value of the process parameter of the historical batch of products in the production and processing stage; wherein the first change factor, the second change factor and the change characteristic value are positively correlated.
[0017] Furthermore, according to the raw material quality evaluation index and process parameter change characteristic value of each historical batch of products, the degree of compensation of the process parameter change on the raw material quality is determined, including:
[0018] The set of raw material quality evaluation indicators of all historical batches of products is used as the quality evaluation indicator set, and the set of process parameter change characteristic values of all historical batches of products is used as the change characteristic value set;
[0019] The correlation coefficient is calculated according to the quality evaluation index set and the change characteristic value set, and the correlation coefficient is used as the compensation degree of the process parameter change on the raw material quality.
[0020] Furthermore, the step of obtaining the quality inspection evaluation value includes:
[0021] Obtain standard values of several evaluation factors of the product; for any product in any historical batch of products or any sampled product in the current batch of products, obtain each evaluation factor of the product in the parameter set of the batch of products in the quality inspection stage, calculate the difference between each evaluation factor of the product and the standard value of the corresponding evaluation factor, record it as the fluctuation factor, and then obtain the quality inspection evaluation value of the product; wherein, the quality inspection evaluation value is negatively correlated with the fluctuation factor.
[0022] Furthermore, according to the quality inspection evaluation value of each product in each historical batch of products, the fluctuation degree of the quality inspection evaluation value of all products in each historical batch of products is determined, including:
[0023] Pair every two products in the same historical batch of products, calculate the difference between the quality inspection evaluation values of each pair of products, and then calculate the average value of the difference between the quality inspection evaluation values of all pairs of products in the same historical batch of products as the degree of fluctuation of the quality inspection evaluation values of all products in the corresponding historical batch of products.
[0024] Furthermore, according to the quality inspection evaluation value of each product in each historical batch of products, the product initial quality inspection function is constructed, including:
[0025] Calculate the average of the quality inspection evaluation values of all products in each historical batch of products during the quality inspection stage as the initial quality inspection value of the corresponding historical batch of products;
[0026] A linear fit is performed on the batch number and the initial quality detection value of each historical batch of products to obtain a fitting function, and the fitting function is used as the product initial quality detection function.
[0027] Furthermore, according to the process parameter change characteristic value of each historical batch of products, the fluctuation degree and the compensation degree, the product initial quality detection function is updated to obtain the product expected quality detection function, including:
[0028] Normalize the sequence of process parameter change characteristic values of each historical batch of products to obtain a normalized sequence, which is recorded as a change characteristic value sequence;
[0029] Normalize the sequence composed of the fluctuation degree of all product quality inspection evaluation values in each historical batch of products to obtain a normalized sequence, which is recorded as the fluctuation degree sequence;
[0030] Calculate the average value of the difference between the data at the same position in the variation characteristic value sequence and the fluctuation degree sequence as the first correction weight of the product initial quality detection function;
[0031] Calculating the product of the first correction weight and the degree of compensation of the raw material quality by the process parameter change, normalizing the product of the first correction weight and the degree of compensation, and obtaining the second correction weight of the product initial quality detection function;
[0032] The second correction weight is multiplied by the slope in the product expected quality detection function to obtain the product expected quality detection function.
[0033] Furthermore, according to the batch number of the current batch of products, the expected quality detection function of the products, and the quality inspection evaluation value of each sampled product in the current batch of products, the adjusted sampling rate is obtained, including:
[0034] The average of the quality inspection evaluation values of all sampled products in the current batch of products is used as the sampling quality evaluation value of the current batch of products; the batch number of the current batch of products is used as input and input into the product quality expectation detection function to obtain the expected quality evaluation value of the current batch of products;
[0035] If the expected quality assessment value of the current batch of products is less than the sampling quality assessment value, the preset sampling rate is increased, and then the sampling quality assessment value is updated, and the expected quality assessment value of the current batch of products and the updated sampling quality assessment value are iteratively compared until the expected quality assessment value of the current batch of products is not less than the updated sampling quality assessment value, and the adjusted sampling rate is obtained.
[0036] Furthermore, the step of obtaining the raw material quality evaluation model includes:
[0037] Build a classification model and obtain training and test data sets;
[0038] Using the training data set to train the constructed classification model to obtain a trained classification model;
[0039] Use the test data set to evaluate the trained classification model, and tune and optimize the classification model based on the evaluation results to obtain the raw material quality evaluation model;
[0040] The input data for each training is the parameter set of the historical batch products in the raw material procurement stage, the label data is the raw material quality evaluation index of the historical batch products determined by manual scoring, and the loss function is the cross entropy loss function.
[0041] A product quality tracing system comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a product quality tracing method.
[0042] The present invention has the following beneficial effects:
[0043] In order to ensure the accuracy and credibility of the product quality traceability process and provide effective data support for improving the traceability accuracy of the batch coding method of product quality traceability, the present invention provides a product quality traceability method and system, which relates to the field of product traceability technology and can be applied to product quality traceability. Changes in process parameters of batch products during the production and processing stage can not only compensate for quality problems of raw materials, but also affect the quality of finished products. In order to further improve the accuracy of quality expectations for the entire batch of products, the degree of compensation for raw material quality by process parameter changes is quantified through raw material quality evaluation indicators and process parameter change characteristic values, and the degree of fluctuation of the quality inspection evaluation values of all products in each historical batch of products is quantified. Based on the degree of compensation and the degree of fluctuation, the initial quality detection function of the product is updated to obtain the expected product quality detection function, which enhances the reliability and accuracy of the expected quality results of future batches of products. The preset sampling rate is adjusted through the quality inspection evaluation value of each sampled product in the current batch of products, the batch number of the current batch of products and the expected product quality detection function, and batch coding is performed according to the adjusted sampling rate. On the one hand, there is no need to detect the quality information of all products in the current batch of products, which improves the detection efficiency; on the other hand, the product information corresponding to the adjusted sampling rate can better represent the quality information of all products in the current batch of products, which effectively ensures the accuracy and credibility of the product quality traceability process. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0045] Figure 1 This is a flow chart of a product quality tracing method according to an embodiment of the present invention;
[0046] Figure 2 The flowchart is a step of determining characteristic values of process parameter changes of historical batch products in an embodiment of the present invention. DETAILED DESCRIPTION
[0047] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the technical solutions proposed by the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0048] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0049] The scenarios targeted by the present invention may be:
[0050] As a common clothing category, sweater products not only meet most of the needs in the market, but also with the development of industrial production, the quality requirements of clothing categories are increasing. Sweater products involve quality control in many aspects such as raw materials, production processes, and production equipment. Product quality traceability can ensure that companies can effectively monitor product quality and ensure that product quality meets standards. Through product quality traceability, it can provide good data support for the optimization of sweater production processes.
[0051] In the process of sweater product quality traceability, we are faced with technical challenges such as diversity of data sources, complex information management, and high requirements for data accuracy. We need to solve problems in data accuracy, information sharing, and identification code management to achieve full traceability of the sweater production process and improve the quality and safety of sweater products. Specifically, the traceability code is used for product quality traceability, and the finished product quality inspection results are used as the traceability code for the product quality in the batch to which they belong. The quality inspection results of the sampled products obtained by sampling inspection in the traceability process ignore the quality impact of other products in the same batch except the sampled products, resulting in uncertainty in the traceability code information of product quality. It is necessary to predict the product quality in the production process and correct the batch coding method based on the prediction results to ensure the accuracy of the traceability information.
[0052] In order to ensure the accuracy and credibility of sweater product quality traceability, an embodiment of the present invention provides a product quality traceability method, such as Figure 1 As shown, the following steps are included:
[0053] S1, obtain the raw material quality evaluation index of several historical batches of products in the raw material procurement stage, the process parameter change characteristic value in the production and processing stage, the quality inspection evaluation value of each product in the quality inspection stage, and obtain the quality inspection evaluation value of each sampled product in the quality inspection stage of the current batch of products.
[0054] S11, obtaining the raw material quality evaluation index of each historical batch of products at the raw material procurement stage.
[0055] In this embodiment, the number of historical batch products can be set to 30, the products produced can be sweater products, and the production stages of sweater products can include raw material procurement stage, production and processing stage, and quality inspection stage. The number of historical batch products and product types can be set by the implementer according to the actual situation, and are not specifically limited.
[0056] The calculation process of the quality evaluation index of the raw materials of each historical batch of products is consistent. In order to reduce unnecessary descriptions, taking any historical batch of products as an example, the neural network model is used to determine the quality evaluation index of the raw materials of the historical batch of products. The specific implementation process may include:
[0057] The parameter set of the historical batch of products in the raw material procurement stage is obtained, and the parameter set of the raw material procurement stage is used as input data, which is input into a pre-built and trained raw material quality evaluation model to obtain the raw material quality evaluation index of the historical batch of products.
[0058] It should be noted that in the raw material procurement stage, the raw material types can be wool, cashmere, cotton yarn, etc.; in the procurement process, it is necessary to record the source, batch, purchase quantity, quality parameters, etc. of the raw materials, and the quality parameters include fiber length, fiber strength, fiber fineness, raw material color, moisture content, etc. The parameter set in the raw material procurement stage is the quality parameter set of the raw materials, and the number of types of quality parameters in the quality parameter set can be determined according to the specific actual situation.
[0059] Among them, the steps for obtaining the raw material quality evaluation model include:
[0060] Construct a classification model, and obtain a training data set and a test data set; use the training data set to train the constructed classification model, and in the training process, use the gradient descent method to train until the loss function converges to obtain a trained classification model; use the test data set to evaluate the trained classification model, adjust and optimize the classification model according to the evaluation results, and then obtain a raw material quality evaluation model. The implementation process of the classification model training is a prior art, which is not within the protection scope of the present invention and will not be elaborated in detail here.
[0061] The input data for each training is a set of parameters of historical batch products in the raw material procurement stage, such as fiber length, fiber strength, fiber fineness, raw material color, etc.; the ratio of the training data set to the test data set can be set to 7:3; the label data is the raw material quality evaluation index of historical batch products determined in the form of manual scoring; the loss function is the cross entropy loss function; the classification network structure can be a five-layer fully connected neural network.
[0062] S12, obtaining the process parameter variation characteristic values of each historical batch of products in the production and processing stage.
[0063] It should be noted that during the production and processing stage, sweater products are produced and processed, and the processing process includes spinning, weaving, dyeing, finishing and other links. During the production and processing process, the Internet of Things devices and sensors are used to monitor the process parameters in real time, including temperature, humidity, pressure, speed, etc. The parameter set in the production and processing stage is the process parameter set in the production and processing process. The number of types of production parameters in the production parameter set can be determined according to the specific actual situation.
[0064] In this embodiment, the calculation process of the process parameter change characteristic value of each historical batch of products is consistent. In order to reduce unnecessary description and facilitate understanding, taking any historical batch of products as an example, the process parameter change characteristic value of the historical batch of products is determined by using the data change stability and similarity. The steps for determining the process parameter change characteristic value are as follows: Figure 2 As shown, the specific implementation process may include:
[0065] S121, curve fitting is performed on the same type of process parameters in the parameter set of the historical batch of products in the production and processing stage to obtain a process parameter curve of each type in the time series.
[0066] In this embodiment, the parameter set of the production and processing stage includes different types of process parameters, such as temperature, humidity, pressure, speed and other process parameters. Each type of process parameter changes continuously with the processing process, so each type of process parameter has its corresponding time series, and the number of data in each time series remains consistent.
[0067] In order to facilitate the subsequent analysis of the changing characteristics of the time series, the least squares method is used to perform curve fitting on each time series, and the fitting curve corresponding to each time series can be obtained. The fitting curve is used as the process parameter curve to obtain each type of process parameter curve in the time series. Among them, the implementation process of the least squares method is a prior art and is not within the scope of protection of the present invention, and will not be elaborated here; the horizontal axis of the process parameter curve is the acquisition time of the process parameter, and the vertical axis is the process parameter.
[0068] S122, calculating the variance of each type of process parameter curve, and taking the average value of all variances as the first variation factor of the process parameter.
[0069] As an example, the calculation formula for the first variation factor of the process parameter of the rth historical batch product may be:
[0070] Where w1 ris the first change factor of the process parameter corresponding to the rth historical batch of products in the production and processing stage, M is the number of types of process parameters corresponding to the rth historical batch of products in the production and processing stage, m is the type sequence number of the process parameter corresponding to the rth historical batch of products in the production and processing stage, σ rm 2 It is the variance of the mth type of process parameter curve corresponding to the rth historical batch of products in the production and processing stage.
[0071] In the calculation formula of the first change factor, the first change factor is obtained by quantifying the variance of the process parameter curve itself, which can be used to represent the overall change of the process parameters during the production process. The larger the fluctuation variance of the process parameter curve, the worse the stability of the process parameter curve, and the larger the first change factor.
[0072] S123, calculating the similarity between each pair of process parameter curves, and taking the average value of all similarities as the second variation factor of the process parameters.
[0073] As an example, the implementation process of the second variation factor of the process parameter may include:
[0074] In this embodiment, in order to analyze the similarity of trend changes between any two process parameter curves, any two process parameter curves can be combined into a pair to obtain multiple pairs of process parameter curves; the existing technology for calculating the similarity between two curves can be used to determine the degree of similarity between each pair of process parameter curves, and then the average value of all similarities obtained is used as the second change factor of the process parameter.
[0075] Existing technologies for calculating the similarity between two curves include but are not limited to: mean square error algorithm, Pearson correlation coefficient algorithm, dynamic time warping algorithm and edit distance algorithm. The similarity calculation process is existing technology and is not within the scope of protection of the present invention, and will not be elaborated here.
[0076] In another example, the process of implementing the second variation factor of the process parameter may include:
[0077] The frequency of occurrence of each data point in each pair of process parameter curves on its own curve is counted. According to the frequency of occurrence of each data point in each pair of process parameter curves on its own curve, the change similarity between the two process parameter curves is analyzed, and the second change factor is quantified. The calculation formula of the second change factor can be:
[0078] In the formula, w2 ris the second change factor of the process parameter corresponding to the rth historical batch of products in the production and processing stage, n is the logarithm of the process parameter curve, i is the sequence number of each pair of process parameter curves, N is the number of data points in the process parameter curve, v is the sequence number of the data point in the process parameter curve, Q v is the frequency of occurrence of the vth data point in the i-th pair of process parameter curves corresponding to the rth historical batch of products in the production and processing stage, P v is the frequency of occurrence of the vth data point in the other process parameter curve of the ith pair corresponding to the rth historical batch of products in the production and processing stage, ln is an exponential function with a natural constant as the base, and || is a function for finding the absolute value.
[0079] In the calculation formula of the second change factor, Q v and P v It can show the distribution probability of the vth data point in the i-th pair of process parameter curves on the corresponding curve, and each data point in each pair of process parameter curves presents a one-to-one comparison; It can represent the information difference between a pair of process parameter curves. The smaller the information difference is, the higher the similarity of the changes of a pair of process parameter curves is, and the smaller the second change factor is.
[0080] S124, combining the first change factor and the second change factor of the process parameter to determine the process parameter change characteristic value of the historical batch product.
[0081] In this embodiment, the first change factor, the second change factor and the change characteristic value are positively correlated. The positive correlation means that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases. The specific relationship can be a multiplication relationship, an addition relationship, or the power of an exponential function. Therefore, the product of the first change factor and the second change factor can be used as the process parameter change characteristic value of the historical batch of products.
[0082] It should be noted that when calculating the characteristic value of process parameter changes, not only the stability of changes in a single type of process parameter over time is taken into account, that is, the first change factor is determined, but also the similarity between a single type of process parameter and other types of process parameters other than itself is taken into account, that is, the second change factor is determined. This effectively improves the comprehensiveness and reliability of the calculation of the change characteristic value, and can accurately reflect the characteristics of process parameter changes in each historical batch of products during the processing and production stage.
[0083] S13, obtaining the quality inspection evaluation value of each product in each historical batch of products during the quality inspection stage.
[0084] First of all, it should be noted that in the quality inspection stage, the sweater products are subject to quality inspection, which includes appearance inspection, size measurement, tensile test, etc. The data generated in the quality inspection process include inspection results, qualification certificates, defect records, etc. The evaluation factors of the finished product in the quality inspection process include the size of each component of the sweater finished product, the tensile strength of the fabric, the fiber content, etc. The set of evaluation factors of each product in a historical batch of products is the parameter set of the historical batch of products in the quality inspection stage.
[0085] In this embodiment, the calculation process of the quality inspection evaluation value of each product in each historical batch of products is the same. In order to reduce unnecessary description, taking any historical batch of products as an example, the quality inspection evaluation value of each product in the historical batch of products is determined by using the difference between the evaluation factor and its standard value. The specific implementation process may include:
[0086] S131, obtaining standard values of several evaluation factors of the product.
[0087] In this embodiment, the standard values of the evaluation factors of the sweater product may include: the standard size of each component of the finished sweater, and the standard size of each component is recorded as D'; the tensile strength of the fabric of the finished sweater is tested, and the minimum tensile strength that meets the product requirements is recorded as L'; the minimum fiber content that meets the production requirements of the finished sweater is recorded as Among them, the various components of the product may include clothing length, sleeve length, shoulder width, etc. The number of components can be set by the implementer according to specific actual conditions and is not specifically limited.
[0088] It should be noted that during the quality inspection stage, the standard values of the same type of evaluation factors for each product in the same batch are the same. The number of types of each evaluation factor of the product can be set by the implementer according to the specific situation. Here, only the size, tensile strength and fiber content of each component of the product are analyzed.
[0089] S132, based on each evaluation factor of each product in the parameter set of the historical batch of products in the quality inspection stage, calculate the difference between each evaluation factor of each product and the standard value of the corresponding evaluation factor, record it as the fluctuation factor, and then obtain the quality inspection evaluation value of each product.
[0090] As an example, the calculation formula for the quality inspection evaluation value of the e-th product in the historical batch of products can be:
[0091] In the formula, E eis the quality inspection evaluation value of the e-th product in the historical batch of products, exp is an exponential function with a natural constant as the base, B is the number of components of the e-th product, b is the serial number of the component of the e-th product, and D eb is the size of the bth component of the eth product in the historical batch of products, D' b is the standard size of the bth component of the product, is the fiber content of the e-th product in the historical batch of products, is the minimum fiber content of the product, L e is the tensile strength of the e-th product in this historical batch of products, L' is the minimum tensile strength of the product, and || is the absolute value function.
[0092] In the calculation formula of the quality inspection evaluation value, Indicates the average difference between the finished product's appearance dimensions and the marked dimensions during the quality inspection process. The smaller the average difference, the more the finished product meets the production and processing standards. It indicates the ratio of the difference in fabric fiber content to the difference in tensile strength of the product. The smaller the ratio, the better the quality of the finished product. The smaller it is, the better the quality of the e-th product is. Using exp(-) can achieve negative correlation processing. Of course, the implementer can also use other negative correlation processing methods to achieve it; | D eb -D' b |、 and |L e -L'| are all fluctuation factors. Each fluctuation factor is negatively correlated with the quality inspection evaluation value, so the product of each fluctuation factor is negatively correlated.
[0093] S14, obtaining the quality inspection evaluation value of each sampled product in the quality inspection stage of the current batch of products.
[0094] In this embodiment, based on the evaluation factors of each sampled product in the current batch of products at the quality inspection stage, the quality inspection evaluation values of all sampled products in the current batch of products can be determined with reference to step S13, which will not be repeated here.
[0095] It is worth mentioning that the quality parameters of raw materials, process parameters during processing, and evaluation factors of finished products can be collected directly from production equipment or production lines and transmitted to the data management system through data transmission technology.
[0096] At this point, this embodiment has obtained the raw material quality evaluation indicators of historical batches of products in the raw material procurement stage, the process parameter change characteristic values in the production and processing stage, the quality inspection evaluation values of each product in the quality inspection stage, and the quality inspection evaluation values of each sampled product in the current batch of products in the quality inspection stage.
[0097] S2, based on the raw material quality evaluation index and process parameter change characteristic value of each historical batch of products, determine the degree of compensation of the process parameter change on the raw material quality.
[0098] First of all, it should be explained that the quality evaluation index can show the quality level of the raw materials of each historical batch of products. The larger the quality evaluation index, the higher the quality of the raw materials of the corresponding historical batch of products. The change characteristic value can show the change characteristics of the same type of process parameters over time during the processing and production of each historical batch of products. The larger the change characteristic value, the more obvious the change in the process parameters of the corresponding historical batch of products during the processing and production. The purpose of calculating the quality evaluation index and the change characteristic value is to analyze the correlation between the two. Through the correlation, the degree of compensation of the process parameter changes for the quality of raw materials can be analyzed.
[0099] If the quality of raw materials is poor, the process parameters need to be adjusted frequently during the sweater product processing to compensate for the insufficient quality of raw materials. By analyzing the correlation between the quality evaluation index and the change characteristic value, the degree of compensation of the process parameter change on the quality of raw materials can be quantified. The greater the correlation, the greater the degree of compensation of the process parameter change on the quality of raw materials. The initial quality detection function of the product can be updated according to the compensation degree, that is, the second correction weight used to determine the initial quality detection function of the product.
[0100] S21, taking a set of raw material quality evaluation indicators of all historical batches of products as a quality evaluation indicator set, and taking a set of process parameter change characteristic values of all historical batches of products as a change characteristic value set.
[0101] S22, calculating the correlation coefficient according to the quality evaluation index set and the change characteristic value set, and taking the correlation coefficient as the degree of compensation of the process parameter change on the raw material quality.
[0102] As an example, the calculation formula for the degree of compensation of the process parameter change on the quality of the raw materials can be:
[0103] Where ρ is the degree of compensation of raw material quality by process parameter changes, Cov(U,Z) is the covariance of the two sets, U is the set of quality evaluation indicators, Z is the set of change characteristic values, and σ U is the standard deviation of the quality evaluation index set, σ Z is the standard deviation of the set of varying eigenvalues.
[0104] In the calculation formula of compensation degree, It indicates the correlation between the quality changes of sweater raw materials and the changes of process parameters in all historical batches of products. The stronger the correlation, the better the process parameters can compensate for the quality problems of raw materials.
[0105] So far, this embodiment has obtained the degree of compensation of the raw material quality by the change of process parameters.
[0106] S3, according to the quality inspection evaluation value of each product in each historical batch of products, determine the fluctuation degree of the quality inspection evaluation values of all products in each historical batch of products, and construct the product initial quality inspection function.
[0107] It should be noted that during the production and processing of sweaters, process parameters need to be adjusted frequently to compensate for the deficiencies in the quality of raw materials. However, fluctuations in process parameters during the production and processing of sweaters will also have a certain impact on the quality of sweater products. That is, the stability of process parameters can make the quality inspection results of products stable. Therefore, the credibility of process parameter fluctuations can be evaluated based on changes in product quality inspection results.
[0108] Among them, there are two situations of fluctuations in process parameters during the production and processing. One is that the stability of process parameter changes is poor in order to compensate for the poor quality of raw materials, so as to overcome the adverse consequences caused by the low quality of raw materials to a certain extent; the other is that the instability of process parameters is not to compensate for the quality of raw materials, but because of abnormalities in the production and processing process, resulting in poor quality of the finished product.
[0109] S31, determining the degree of fluctuation of the quality inspection evaluation values of all products in each historical batch of products according to the quality inspection evaluation value of each product in each historical batch of products.
[0110] In this embodiment, every two products in the same historical batch of products are grouped into a pair, and the difference between the quality inspection evaluation values of each pair of products is calculated. Then, the average value of the difference between the quality inspection evaluation values of all pairs in the same historical batch of products is calculated as the degree of fluctuation of the quality inspection evaluation values of all products in the corresponding historical batch of products.
[0111] As an example, the calculation formula for the fluctuation degree of the quality inspection evaluation values of all products in the rth historical batch of products can be:
[0112] In the formula, R r is the fluctuation degree of the quality inspection evaluation value of all products in the rth historical batch of products, V is the number of pairs of products in the rth historical batch of products, h is the serial number of each pair of products in the rth historical batch of products, E e is the quality inspection evaluation value of a product in the hth pair of the rth historical batch of products, E g is the quality inspection evaluation value of the other product in the hth pair of the rth historical batch of products, |Ee -E g | h is the difference between the quality inspection evaluation values of the hth pair of products in the rth historical batch of products, and || is the absolute value function.
[0113] In the calculation formula of volatility, |E e -E g | h It indicates the difference in quality inspection evaluation values of any two paired products in the same historical batch of products. The greater the difference in quality inspection evaluation values between each pair of products, the more drastic the fluctuation in the quality inspection evaluation values of the historical batch of products.
[0114] S32, constructing a product initial quality inspection function according to the quality inspection evaluation value of each product in each historical batch of products.
[0115] S321, calculating the average of the quality inspection evaluation values of all products in each historical batch of products as the initial quality inspection value of the products in the corresponding historical batch of products.
[0116] In this embodiment, the product initial quality inspection function can be used to evaluate the finished product quality inspection status of all products in the entire historical batch of products. Therefore, when constructing the product initial quality inspection function, it is necessary to take the average value of the quality inspection evaluation values of all products in each historical batch of products as the basic data, recorded as the initial quality inspection value, and each historical batch of products has its corresponding initial quality inspection value.
[0117] S322, performing linear fitting on the batch number and initial quality inspection value of each historical batch of products to obtain a fitting function, and using the fitting function as the product initial quality inspection function.
[0118] In this embodiment, the product initial quality inspection function can be used to evaluate the initial quality inspection status of all products in different batches of products. The horizontal axis of the linear fitting is the batch number of the historical batch products, and the vertical axis is the initial quality inspection value of the products in each historical batch of products. The product initial quality inspection function can be expressed as y=kx+b, x is the batch number of the independent variable, y is the initial quality inspection value of the dependent variable, k is the slope, and b is the intercept.
[0119] So far, this embodiment has obtained the fluctuation degree of all product quality inspection evaluation values in each historical batch of products and the product initial quality inspection function.
[0120] S4, according to the characteristic value, fluctuation degree and compensation degree of the process parameter changes of each historical batch of products, the initial quality detection function of the product is updated to obtain the expected quality detection function of the product.
[0121] It should be noted that the product initial quality detection function only considers the quality of the finished products of the same historical batch of products, ignoring the correlation between the changes in process parameters and the quality of raw materials and the fluctuation degree of quality inspection evaluation values, resulting in low accuracy of the finished product quality analysis results of the current batch of products. Therefore, it is necessary to update the product initial quality detection function using the change characteristic value, fluctuation degree and compensation degree. The specific implementation process may include:
[0122] S41, normalizing the sequence consisting of the process parameter change characteristic values of each historical batch of products to obtain a normalized sequence, which is recorded as a change characteristic value sequence.
[0123] S42, normalizing the sequence consisting of the fluctuation degrees of the quality inspection evaluation values of all products in each historical batch of products to obtain a normalized sequence, which is recorded as a fluctuation degree sequence.
[0124] In this embodiment, in order to unify the dimensions of different types of data, both the change characteristic value and the fluctuation degree are normalized, and the value ranges of the data in the obtained change characteristic value sequence and the fluctuation degree sequence are both between 0 and 1. The normalization process can be a maximum and minimum value normalization process, and the normalization in the implementation steps of this embodiment can all adopt the maximum and minimum value normalization process. In other embodiments of the present invention, other normalization methods can be selected according to the specific range of values, which will not be described in detail.
[0125] S43, calculating the average value of the difference between the data at the same position in the variation characteristic value sequence and the fluctuation degree sequence as the first correction weight of the product initial quality detection function.
[0126] In another example, the calculation process of the first correction weight may include:
[0127] The least square method is used to perform curve fitting on the change characteristic value sequence and the fluctuation degree sequence to obtain two fitting curves, and the first correction weight is obtained by calculating the DTW (Dynamic Time Warping) distance value of the two fitting curves. Among them, the least square method and the DTW distance calculation process are both prior arts, which are not within the protection scope of the present invention and will not be elaborated in detail here.
[0128] S44, calculating the product of the first correction weight and the degree of compensation of the process parameter change on the raw material quality, normalizing the product of the first correction weight and the degree of compensation, and obtaining the second correction weight of the product initial quality detection function.
[0129] As an example, the calculation formula of the second correction weight of the product initial quality detection function may be:
[0130] Where ω is the second correction weight of the product initial quality detection function, th is the normalization function, H is the number of all historical batches of products, r is the batch number of each historical batch of products, and w r is the normalized value of the process parameter change characteristic value of the rth historical batch of products in the change characteristic value sequence, R r is the normalized value of the fluctuation degree of the quality inspection evaluation values of all products in the rth historical batch of products in the fluctuation degree sequence, | | is the absolute value function, and ρ is the compensation degree of the process parameter change on the raw material quality.
[0131] In the calculation formula of the second correction weight, It indicates the correlation between the fluctuation degree of the characteristic value of process parameter change and the quality inspection evaluation value in all production batches. The larger the value, the higher the synchronous change degree of the quality inspection evaluation value of the finished product when the characteristic value of the process parameter changes fluctuates; represents the correction weight for the expected finished product quality, The larger the value, the stronger the correlation between the change of process parameters and the change of raw material quality during the production process, that is, the larger the value, and the smaller the correlation between the change of process parameters and the change of quality inspection evaluation value, that is, The larger it is, the more it means that the changes in process parameters during the production process are mainly to compensate for the quality problems of raw materials, and the quality changes of the finished product are relatively stable, and the greater the degree of correction of the finished product quality inspection results is needed.
[0132] S45, multiplying the second correction weight by the slope in the product expected quality detection function to obtain the product expected quality detection function.
[0133] In this embodiment, the second correction weight is multiplied by the slope k in the expected product quality detection function y=kx+b to complete the update of the expected product quality detection function, and the updated expected product quality detection function is used as the expected product quality detection function.
[0134] So far, this embodiment has obtained a product expected quality detection function that can be used to determine the expected quality evaluation value of each historical batch of products.
[0135] S5, according to the batch number of the current batch of products, the expected quality detection function of the products and the quality inspection evaluation value of each sampled product in the current batch of products, obtain the adjusted sampling rate, and then obtain the traceability code information when the current batch of products is batch coded, so as to trace the product quality.
[0136] It should be noted that the existing batch coding method has the problem of low accuracy, and the batch coding method needs to be corrected, that is, the sampling rate needs to be adjusted. Batch coding based on the adjusted sampling rate, on the one hand, does not need to detect the quality information of all products in the current batch of products, which improves the detection efficiency; on the other hand, the product information corresponding to the adjusted sampling rate can better represent the quality information of all products in the current batch of products, which effectively ensures the accuracy and credibility of the product quality traceability process.
[0137] S51, adjusting the preset sampling rate according to the quality inspection evaluation value of each sampled product in the current batch of products and the expected detection function of product quality to obtain an adjusted sampling rate.
[0138] In this embodiment, the expected quality evaluation value obtained by the expected product quality detection function and the sampling quality evaluation value corresponding to the sampled product are combined to determine whether the sampling rate set for the current batch is applicable to the finished product quality of the current batch of products. If not, the preset sampling rate is adjusted until the sampling rate is applicable to the current batch of products. The preset sampling rate can be set to 10%, and the implementer can set the size of the preset sampling rate according to the specific actual situation without specific limitation.
[0139] S511, obtain the quality inspection evaluation value of each sampled product in the current batch of products, and take the average of the quality inspection evaluation values of all sampled products in the current batch of products as the sampling quality evaluation value of the current batch of products; take the batch number of the current batch of products as input, input it into the product quality expectation detection function, and obtain the expected quality evaluation value of the current batch of products.
[0140] In this embodiment, the expected quality assessment value can characterize the expected quality of the finished products of the current batch of products during the quality inspection stage. When subsequently predicting the expected quality assessment value of the next batch of products, the batch number of the next batch of products can also be input into the product quality expectation detection function to determine it, which effectively improves the quality inspection efficiency of the finished products corresponding to the batch products.
[0141] S512, if the expected quality assessment value of the current batch of products is less than the sampling quality assessment value, it is determined that there is an abnormality in the batch coding of the current batch of products, that is, the sampling rate in the batch coding process is not good, and the preset sampling rate needs to be increased. The larger the sampling rate, the more comprehensive the quality assessment of the current batch of products; then the sampling quality assessment value is updated according to the increased sampling rate, and the expected quality assessment value of the current batch of products and the updated sampling quality assessment value are compared iteratively and continuously until the expected quality assessment value of the current batch of products is not less than the updated sampling quality assessment value, then the comparison is stopped to obtain the adjusted sampling rate at this time.
[0142] It should be noted that if the expected quality assessment value is lower than the sampling quality assessment value, it means that there may be potential risks or uncertainties in the product quality of the current batch of products, or that the sampled products fail to accurately capture fluctuations or changes in product quality. By increasing the preset sampling rate, the risk of product quality problems being ignored or missed can be reduced, and the accuracy of quality testing can be improved to ensure that quality problems are discovered and handled in a timely manner. Therefore, when the expected quality assessment value is lower than the sampling quality assessment value, it is necessary to adjust the product preset sampling rate, which can help improve detection accuracy, reduce quality detection risks, and thus improve the accuracy of product quality traceability.
[0143] S52, according to the adjusted sampling rate, obtain the traceability code information of the current batch of products and conduct product quality traceability.
[0144] In this embodiment, after obtaining the adjusted sampling rate, the quality inspection data obtained by the adjusted sampling rate is used as the traceability code information uniformly assigned to the current batch of products to complete the batch coding process; according to the traceability code information of the current batch of products, the quality traceability of the current batch of products can be achieved, which effectively ensures the accuracy and credibility of the traceability of the current batch of products. The process of realizing product quality traceability based on traceability code information is a prior art and is not within the scope of protection of the present invention, and will not be elaborated here.
[0145] Thus, this embodiment achieves product quality traceability with higher accuracy and credibility.
[0146] An embodiment of the present invention further provides a product quality tracing system, including a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a product quality tracing method.
[0147] In summary, a parameter set of several historical batches of products at different production stages is obtained, and according to the parameter sets of several historical batches of products at different production stages, the raw material quality evaluation index, process parameter change characteristic value and the degree of fluctuation of the quality inspection evaluation value of all products of each historical batch of products are determined; the degree of compensation of the process parameter change on the quality of raw materials is determined in combination with the quality evaluation index and the change characteristic value, and the initial quality detection function of the product is updated using the degree of fluctuation and the degree of compensation to obtain the expected quality detection function of the product; according to the quality inspection evaluation value of the sampled product and the expected product quality detection function, the preset sampling rate is adjusted to obtain the adjusted sampling rate, and then the traceability code information of the current batch of products when batch coding is obtained, which ensures the transparency and credibility of the traceability method, effectively reduces the difficulty of product quality control and traceability, ensures the accuracy and consistency of quality information, and helps to achieve the reliability of product quality traceability.
[0148] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.
Claims
1. A product quality tracing method, characterized in that: The following steps are involved: Obtain the raw material quality evaluation indexes of several historical batches of products at the raw material procurement stage, the process parameter change characteristic values at the production and processing stage, the quality inspection evaluation value of each product at the quality inspection stage, and obtain the quality inspection evaluation value of each sampled product at the quality inspection stage for the current batch of products; Determine the degree of compensation for raw material quality caused by process parameter changes based on the raw material quality evaluation index and process parameter change characteristic value of each historical batch of products; According to the quality inspection evaluation value of each product in each historical batch of products, determine the degree of fluctuation of the quality inspection evaluation values of all products in each historical batch of products, and construct the product initial quality detection function; According to the process parameter change characteristic value of each historical batch of products, the fluctuation degree and the compensation degree, the product initial quality detection function is updated to obtain the product expected quality detection function; According to the batch number of the current batch of products, the expected quality detection function of the product and the quality inspection evaluation value of each sampled product in the current batch of products, the adjusted sampling rate is obtained, and then the traceability code information of the current batch of products when batch coding is obtained to trace the product quality; The steps of obtaining the raw material quality evaluation index include: For any historical batch of products, obtain a parameter set of the historical batch of products at the raw material procurement stage, use the parameter set of the raw material procurement stage as input data, input it into a pre-built and trained raw material quality evaluation model, and obtain the raw material quality evaluation index of the historical batch of products at the raw material procurement stage; The step of obtaining the process parameter variation characteristic value comprises: For any historical batch of products, obtain each type of process parameter in the parameter set of the historical batch of products in the production and processing stage, perform curve fitting on the same type of process parameters in the parameter set of the historical batch of products in the production and processing stage, and obtain each type of process parameter curve in the time series; Calculate the variance of each type of process parameter curve, and use the average value of all variances as the first variation factor of the process parameter; calculate the similarity between each pair of process parameter curves, and use the average value of all similarities as the second variation factor of the process parameter; Determine the process parameter change characteristic value of the historical batch product in the production and processing stage in combination with the first change factor and the second change factor of the process parameter; wherein the first change factor, the second change factor and the change characteristic value are positively correlated; According to the raw material quality evaluation index and process parameter change characteristic value of each historical batch of products, determine the degree of compensation of process parameter changes on raw material quality, including: The set of raw material quality evaluation indicators of all historical batches of products is used as the quality evaluation indicator set, and the set of process parameter change characteristic values of all historical batches of products is used as the change characteristic value set; Calculate the correlation coefficient based on the quality evaluation index set and the change characteristic value set, and use the correlation coefficient as the degree of compensation of the process parameter change on the quality of the raw materials; The step of obtaining the quality inspection evaluation value comprises: Obtain standard values of several evaluation factors of the product; for any product in any historical batch of products or any sampled product in the current batch of products, obtain each evaluation factor of the product in the parameter set of the batch of products in the quality inspection stage, calculate the difference between each evaluation factor of the product and the standard value of the corresponding evaluation factor, record it as the fluctuation factor, and then obtain the quality inspection evaluation value of the product; wherein the quality inspection evaluation value is negatively correlated with the fluctuation factor; Based on the quality inspection evaluation value of each product in each historical batch of products, determine the degree of fluctuation of the quality inspection evaluation values of all products in each historical batch of products, including: Pair every two products in the same historical batch of products, calculate the difference between the quality inspection evaluation values of each pair of products, and then calculate the average value of the difference between the quality inspection evaluation values of all pairs of products in the same historical batch of products as the fluctuation degree of the quality inspection evaluation values of all products in the corresponding historical batch of products; According to the quality inspection evaluation value of each product in each historical batch of products, the product initial quality inspection function is constructed, including: Calculate the average of the quality inspection evaluation values of all products in each historical batch of products during the quality inspection stage as the initial quality inspection value of the corresponding historical batch of products; Performing linear fitting on the batch number and initial quality test value of each historical batch of products to obtain a fitting function, and using the fitting function as the product initial quality test function; According to the process parameter change characteristic value of each historical batch of products, the fluctuation degree and the compensation degree, the product initial quality detection function is updated to obtain the product expected quality detection function, including: Normalize the sequence of process parameter change characteristic values of each historical batch of products to obtain a normalized sequence, which is recorded as a change characteristic value sequence; Normalize the sequence composed of the fluctuation degree of all product quality inspection evaluation values in each historical batch of products to obtain a normalized sequence, which is recorded as the fluctuation degree sequence; Calculate the average value of the difference between the data at the same position in the variation characteristic value sequence and the fluctuation degree sequence as the first correction weight of the product initial quality detection function; Calculating the product of the first correction weight and the degree of compensation of the raw material quality by the process parameter change, normalizing the product of the first correction weight and the degree of compensation, and obtaining the second correction weight of the product initial quality detection function; Multiplying the second correction weight by the slope in the product expected quality detection function to obtain the product expected quality detection function; According to the batch number of the current batch of products, the expected quality detection function of the product, and the quality inspection evaluation value of each sampled product in the current batch of products, the adjusted sampling rate is obtained, including: The average of the quality inspection evaluation values of all sampled products in the current batch of products is used as the sampling quality evaluation value of the current batch of products; the batch number of the current batch of products is used as input and input into the product quality expectation detection function to obtain the expected quality evaluation value of the current batch of products; If the expected quality assessment value of the current batch of products is less than the sampling quality assessment value, the preset sampling rate is increased, and then the sampling quality assessment value is updated, and the expected quality assessment value of the current batch of products and the updated sampling quality assessment value are iteratively compared until the expected quality assessment value of the current batch of products is not less than the updated sampling quality assessment value, and the adjusted sampling rate is obtained.
2. A product quality traceability system, characterized in that: It comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a product quality tracing method as claimed in claim 1.
