Product manufacturing information management system and method based on big data

Abstract

The invention discloses a product manufacturing information management system and method based on big data, relates to the technical field of factory production management, and solves the technical problem that an existing product manufacturing information management system is low in efficiency of tracing defective products. The data acquisition module is used for acquiring production data of each stage of production equipment through data acquisition equipment connected with the data acquisition module, and acquiring product feedback information through a user side; the data analysis module is used for acquiring a plurality of product types and corresponding product production data through a database; generating a plurality of product feature data according to the product type and the product production data; the defect tracing module is used for acquiring product feedback information and product feature data and generating a defect tracing report according to the product feedback information and the product feature data; according to the invention, the production stage in which abnormity may occur can be quickly positioned, the excluded data volume is reduced, the troubleshooting accuracy is improved, and the tracing efficiency of defective products is further improved.

Description

Technical Field

[0001] The present application belongs to the field of factory production management technology, and relates to factory production management technology, specifically a product manufacturing information management system and method based on big data. Background Art

[0002] With the further development of science and technology, especially the popularization and rapid development of information technology represented by digital technology and computer technology, the world today has entered the information age.

[0003] Existing product manufacturing information management systems often focus on the aggregation of product production data. When a problem occurs with a product, it is often necessary to check the product's production process in sequence. Since most product production processes are relatively large and each part has a large amount of corresponding production data, in this case, the traditional method is used to check the entire production process of the product, which will result in a large amount of data to be processed. At the same time, since the existing investigation is mainly carried out manually, the probability of errors or omissions is high, which makes the traceability of defective products low. Therefore, a product manufacturing information management system and method based on big data is needed. Summary of the invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes a product manufacturing information management system and method based on big data, which is used to solve the technical problem that the existing product manufacturing information management system has low efficiency in tracing defective products.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides a product manufacturing information management system based on big data, including: a data acquisition module, a data analysis module, a defect tracing module, a product early warning module and a database;

[0006] The data acquisition module: acquires production data of each stage of the production equipment through the data acquisition device connected thereto, and acquires product feedback information through the user terminal;

[0007] The data analysis module: obtains a number of product types and their corresponding product production data through a database; generates a number of product feature data according to the product types and product production data;

[0008] The defect tracing module: obtains product feedback information and product feature data, and generates a defect tracing report based on the product feedback information and product feature data;

[0009] The product early warning module obtains production data at each stage and generates defect alarms based on the stage production data and defect feature data.

[0010] The present application obtains several product types and their corresponding product production data; extracts several production parameters from the stage production data corresponding to each production stage in each product production data; integrates each production parameter into a feature vector, and performs cluster analysis on the feature vectors of the production data of each stage to obtain several cluster clusters and cluster centers; generates product codes of corresponding product types according to the cluster clusters corresponding to the production data of each stage; generates corresponding product feature data according to the product code, and the corresponding product type and type keywords; obtains product feedback information; and generates a defect traceability report according to the product feedback information and the product feature data; the present application performs cluster analysis on the data of each production stage in the product production data corresponding to the recorded various product types, so as to distinguish the production data of each stage of various product types, so that when the defect type is subsequently searched, the stage where the abnormality may occur can be quickly located, the amount of excluded data can be reduced, the accuracy of the investigation can be increased, and the traceability efficiency of defective products can be improved.

[0011] Preferably, the generating of a number of product characteristic data according to the product type and product production data includes:

[0012] Extracting several production parameters from the stage production data corresponding to each production stage in the production data of each product; integrating each production parameter into a feature vector, and performing cluster analysis on the feature vectors of the production data of each stage to obtain several clusters and cluster centers;

[0013] Generate product codes of corresponding product types according to the clusters corresponding to the production data of each stage; extract product types and type keywords in the product production data, and generate corresponding product feature data according to the product codes and the corresponding product types and type keywords.

[0014] Preferably, the clustering analysis of the characteristic vectors of the production data at each stage to obtain a number of clusters and cluster centers includes the following steps:

[0015] Step 1: Obtain the domain density of the feature points corresponding to each feature vector in the domain corresponding to the set radius; take the feature vector whose domain density is greater than the set density threshold as the initial core point;

[0016] Step 2: Obtain the Euclidean distance between each feature point and each initial core point, and assign the feature point to the initial cluster of the initial core point with the smallest Euclidean distance, so as to form several initial clusters with each initial core point;

[0017] Step 3: Merge the initial clusters whose Euclidean distance between each initial cluster is less than the set distance threshold and whose sum of cluster density between adjacent clusters is greater than the corresponding density threshold to generate clusters, and take the point with the largest cluster density in the cluster as the cluster center.

[0018] Preferably, generating product codes of corresponding product types according to clusters corresponding to production data of each stage includes:

[0019] Numbering a number of clusters corresponding to the stage production data to obtain a number of stage cluster numbers corresponding to the corresponding stage production data;

[0020] The stage cluster number to which the production data of each stage corresponds to each product production data is obtained, and the stage cluster numbers are integrated according to the order of the stage production data to obtain the product code.

[0021] Preferably, the generating corresponding product feature data according to the product code, and the corresponding product type and type keyword includes:

[0022] Extract each product code corresponding to each product production data of the same product type, obtain the same type number of the product code in the product type and the different type number of the product code in other product types, generate the reference coefficient of the corresponding product code in the product type according to the same type number and the different type number; sort each product code in descending order of the reference coefficient to obtain a type code table; integrate the type code table and the corresponding type keywords into type characteristic data of the corresponding product type;

[0023] Integrate each type of feature data and its corresponding product type into product feature data.

[0024] Preferably, generating a reference coefficient of the corresponding product code in the product type according to the same type quantity and the different type quantity includes:

[0025] Get the same code number and the different code number, and mark them as TB and YB respectively.

[0026] The reference coefficient CX of the corresponding product code in the product type is calculated by the formula CX=α1×exp(β1×TB / TZB)+α2×exp(-β2×YB / TB); wherein α1 and α2 are proportional coefficients used to adjust the proportion of the two items in the formula, and the specific values ​​are set according to experience; β1 and β2 are adjustment coefficients used to adjust the distribution of the corresponding two items, respectively, and the specific values ​​are set according to experience; TZB is the total number of product codes corresponding to the product type.

[0027] Preferably, generating a defect tracing report based on product feedback information and product feature data includes:

[0028] Obtain product feedback information; extract feedback keywords from product feedback information and convert feedback keywords into feedback word vectors;

[0029] Extract type keywords from each type feature data in the product feature data, and convert the type keywords into type word vectors; calculate the cosine similarity between each feedback word vector and the type word vector corresponding to each type keyword, and take the product type corresponding to the type keyword with the largest cosine similarity value as the product defect type; generate a defect traceability report based on the product defect type and its corresponding type feature data.

[0030] Preferably, generating a defect tracing report according to the product defect type and its corresponding type characteristic data includes:

[0031] Obtain a type code table in the type feature data of a product type that is normal, and mark the product code in the type code table as a normal code; integrate the stage cluster numbers corresponding to each normal code in the same production stage into a normal stage cluster set corresponding to the production stage;

[0032] Extract the type code table in the type feature data corresponding to the product defect type; extract the stage cluster number of each production stage corresponding to each product code in the type code table; determine whether each of the stage cluster numbers is in the normal stage cluster set of the corresponding production stage, if yes, set the corresponding production stage as the normal stage; if not, set the corresponding production stage as the abnormal stage, and generate the abnormal probability coefficient of the corresponding production stage according to the stage cluster number that is not in the normal stage cluster set and the reference degree coefficient of its corresponding product code;

[0033] Trace back the defect reports of each abnormal stage and its corresponding abnormal probability coefficient.

[0034] Preferably, the step of generating the abnormal probability coefficient of the corresponding production stage according to the reference coefficient of the stage cluster number not in the normal stage cluster set and its corresponding product code includes:

[0035] Get the stage cluster number that is not in the normal stage cluster set, and the reference coefficient of the corresponding product code, and mark them as CXi; calculate the abnormal probability coefficient YG of the corresponding production stage by formula YG=∑(CXi) / ∑(CXj); where i is the number of the stage cluster number that is not in the normal stage cluster set, i=1, 2, ..., I, I is the total number of stage cluster numbers that are not in the normal stage cluster set; j is the number of the stage cluster numbers of the corresponding product type in the corresponding production stage, j=1, 2, ..., J, J is the total number of the stage cluster numbers of the corresponding product type in the corresponding production stage.

[0036] Preferably, the generating of defect alarms according to the stage production data and defect characteristic data comprises:

[0037] Acquire the production data of each stage, integrate the production parameters in the production data of each stage into the corresponding feature vector; obtain the stage cluster number of the cluster cluster to which each feature vector belongs; and integrate the stage cluster number corresponding to each production stage into a partial product code, and use the partial product code to query the type code with the highest overlap in the type code table; when the product type corresponding to the type code is abnormal and the reference coefficient is greater than the set coefficient threshold, a defect alarm for the corresponding product type is generated.

[0038] The second aspect of the present application provides a product manufacturing information management method based on big data, comprising the following steps:

[0039] Step 1: Obtain several product types and their corresponding product production data;

[0040] Step 2: extract several production parameters from the stage production data corresponding to each production stage in the production data of each product; integrate each production parameter into a feature vector, perform cluster analysis on the feature vectors of the production data of each stage to obtain several clusters and cluster centers;

[0041] Step 3: Generate product codes for corresponding product types based on the clusters corresponding to the production data at each stage;

[0042] Step 4: Generate corresponding product feature data according to the product code, the corresponding product type and type keyword;

[0043] Step 5: Get product feedback information;

[0044] Step 6: Generate a defect traceability report based on product feedback information and product feature data.

[0045] Compared with the prior art, the beneficial effects of this application are:

[0046] 1. The present application obtains several product types and their corresponding product production data; extracts several production parameters from the stage production data corresponding to each production stage in each product production data; integrates each production parameter into a feature vector, and performs cluster analysis on the feature vectors of the production data of each stage to obtain several cluster clusters and cluster centers; generates product codes of corresponding product types according to the cluster clusters corresponding to the production data of each stage; generates corresponding product feature data according to the product code, and the corresponding product type and type keywords; obtains product feedback information; generates a defect traceability report according to the product feedback information and the product feature data; The present application performs cluster analysis on the data of each production stage in the product production data corresponding to the recorded various product types, so as to distinguish the production data of each stage of various product types, so that when the defect type is subsequently searched, the stage where the abnormality may occur can be quickly located, the amount of excluded data can be reduced, the accuracy of the investigation can be increased, and the traceability efficiency of defective products can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 This is a schematic diagram of the principle of the product manufacturing information management system in this application;

[0049] Figure 2 This is a schematic diagram of the steps of the product manufacturing information management method in this application. DETAILED DESCRIPTION

[0050] The technical solution of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0051] See also Figure 1 , the first aspect of the present application provides a product manufacturing information management system based on big data, including: a data acquisition module, a data analysis module, a defect tracing module, a product early warning module and a database;

[0052] Data acquisition module: obtains production data of each stage of production equipment through the data acquisition equipment connected to it, and obtains product feedback information through the user end; stage production data is the relevant parameters collected by the product in each production stage, such as in the stamping part production process, the relevant parameters in the stamping stage include the size of the stamping part and the stamping machine pressure, etc.; stage production data also includes the production equipment and responsible personnel of the corresponding production stage; data acquisition equipment includes pressure acquisition equipment, etc., specifically which data acquisition equipment is related to the parameters that need to be measured in the production stage, and product feedback information is the feedback information of users on product defects after the corresponding product is put into use, such as the product is prone to cracking, etc.

[0053] The data analysis module: obtains several product types and their corresponding product production data through a database; generates several product feature data according to the product types and product production data; the product types include normal types and various abnormal types, the product production data corresponds to the product types, and is the data generated during the production process of the corresponding products, including several stages of production data; the product feature data includes several product types and their corresponding type feature data, the type feature data includes type keywords and a type coding table, the type keywords are the characteristic description keywords of the corresponding product types; the type coding table is the codes corresponding to various production situations in each production stage of the corresponding product type and the occurrence probability of each code;

[0054] The defect tracing module: obtains product feedback information and product feature data, and generates a defect tracing report based on the product feedback information and product feature data; the defect tracing report includes the possible production stage at which the corresponding product may have a problem and what the problem may be, and the relevant personnel can query the relevant parameters of the corresponding product at the corresponding production stage based on the report;

[0055] The product early warning module obtains production data at each stage and generates defect alarms based on the stage production data and defect feature data.

[0056] It can be understood that this embodiment is applicable to various processes of product production, including the production process of product raw materials and various processes of product production.

[0057] The present embodiment obtains several product types and their corresponding product production data; extracts several production parameters from the stage production data corresponding to each production stage in each product production data; integrates each production parameter into a feature vector, performs cluster analysis on the feature vector of each stage production data to obtain several cluster clusters and cluster centers; generates product codes of corresponding product types according to the cluster clusters corresponding to the production data of each stage; generates corresponding product feature data according to the product code, and the corresponding product type and type keyword; obtains product feedback information; generates a defect traceability report according to the product feedback information and the product feature data; since the occurrence of defective products may be caused by problems with certain production parameters in a certain stage or multiple stages of the product production stage; therefore, the present embodiment performs cluster analysis on the data of each production stage in the product production data corresponding to the recorded various product types to distinguish the production data of each stage of various product types, so that when the defect type is subsequently searched, the stage where the abnormality may occur can be quickly located, the amount of excluded data can be reduced, the accuracy of the investigation can be increased, and the traceability efficiency of the defective products can be improved.

[0058] Generate a number of product characteristic data according to product type and product production data, including: extracting a number of production parameters from the stage production data corresponding to each production stage in each product production data; the production parameters are various parameters related to product production collected in the corresponding production stage, such as pressure, temperature and size, etc.; integrate each production parameter into a characteristic vector, specifically, after performing corresponding preprocessing on each production parameter, mark each production parameter as SCg, and generate a characteristic vector ST={SC1, SC2, ..., SCg, ..., SCG} according to the production parameter, where G is the total number of production parameters in the corresponding production stage; perform cluster analysis on the characteristic vectors of the production data of each stage to obtain a number of cluster clusters and cluster centers; including the following steps:

[0059] Step 1: Obtain the domain density of the feature points corresponding to each feature vector in the domain corresponding to the set radius; take the feature vector whose domain density is greater than the set density threshold as the initial core point; the radius and density threshold are both set based on experience; the domain density can be the ratio of the domain area to the number of feature points in the domain;

[0060] Step 2: Obtain the Euclidean distance between each feature point and each initial core point, and assign the feature point to the initial cluster of the initial core point with the smallest Euclidean distance, so as to form several initial clusters with each initial core point;

[0061] Step 3: Merge the initial clusters whose Euclidean distance between each initial cluster is less than the set distance threshold and whose sum of cluster density between adjacent clusters is greater than the corresponding density threshold to generate cluster clusters, and take the point with the largest cluster density in the cluster as the cluster center; the cluster density can be the sum of the number of feature points and initial core points in the corresponding cluster area ratio.

[0062] Generate product codes of corresponding product types according to the clusters corresponding to the production data of each stage; extract product types and type keywords in the product production data, and generate corresponding product feature data according to the product codes and the corresponding product types and type keywords.

[0063] Generating a product code of a corresponding product type according to the clusters corresponding to the production data of each stage, including: numbering a number of clusters corresponding to the stage production data to obtain a number of stage cluster numbers corresponding to the corresponding stage production data;

[0064] The stage cluster number to which the production data of each stage corresponds to each product production data is obtained, and the stage cluster numbers are integrated according to the order of the stage production data to obtain the product code.

[0065] In this embodiment, the corresponding product has three production stages. The stage cluster number corresponding to the first production stage of a certain product is JC1=5, the stage cluster number corresponding to the second production stage is JC2=7, and the stage cluster number corresponding to the third production stage is JC3=2; 5 is the encoding of the characteristic vector of the production parameters in the first production stage of the corresponding product at the corresponding cluster center; 7 is the encoding of the characteristic vector of the production parameters in the second production stage of the corresponding product at the corresponding cluster center; 2 is the encoding of the characteristic vector of the production parameters in the third production stage of the corresponding product at the corresponding cluster center; the product code at this time is {5, 7, 2}.

[0066] This embodiment performs cluster analysis on the correlations of various production stages of the product, thereby facilitating the subsequent establishment of the correlation between product defects and various production stages, and further facilitating the subsequent rapid tracing of product quality; timely adjusting parameters of relevant production stages; and improving product production quality.

[0067] Generate corresponding product characteristic data according to the product code, and the corresponding product type and type keyword, including: extract each product code corresponding to each product production data of the same product type, obtain the same type number of the product code in the product type and the different type number of the product code in other product types, and generate the reference degree coefficient of the corresponding product code in the product type according to the same type number and the different type number; the reference degree coefficient can be approximately understood as the probability of the corresponding product code appearing in the corresponding product type; or understood as the matching degree between the product code and the corresponding product type; the larger the reference degree coefficient, the higher the corresponding matching degree; sort each product code in descending order of the reference degree coefficient to obtain a type code table; integrate the type code table and the corresponding type keyword into the type characteristic data of the corresponding product type;

[0068] Integrate each type of feature data and its corresponding product type into product feature data.

[0069] Generate a reference coefficient of the corresponding product code in the product type according to the same type quantity and the different type quantity, including: obtaining the same code quantity and the different code quantity, and marking them as TB and YB respectively,

[0070] The reference coefficient CX of the corresponding product code in the product type is calculated by the formula CX=α1×exp(β1×TB / TZB)+α2×exp(-β2×YB / TB); wherein α1 and α2 are proportional coefficients used to adjust the proportion of the two items in the formula, and the specific values ​​are set according to experience; β1 and β2 are adjustment coefficients used to adjust the distribution of the corresponding two items, respectively, and the specific values ​​are set according to experience; TZB is the total number of product codes corresponding to the product type.

[0071] This embodiment calculates the reference coefficient of each product code and the corresponding product type through the above formula; when the probability of the product code appearing in the corresponding product type is higher, or the number of times the product code appears in other product types is lower, the product code is more capable of representing the corresponding product type, that is, the higher the degree of matching between the product code and the corresponding product type, the higher the corresponding reference coefficient; otherwise, it is lower.

[0072] Generate a defect traceability report based on product feedback information and product feature data, including: obtaining product feedback information; extracting feedback keywords from product feedback information, and converting feedback keywords into feedback word vectors;

[0073] Extract the type keywords in each type feature data in the product feature data, and convert the type keywords into type word vectors; calculate the cosine similarity between each feedback word vector and the type word vector corresponding to each type keyword, and use the product type corresponding to the type keyword with the largest cosine similarity value as the product defect type; generate a defect traceability report based on the product defect type and its corresponding type feature data; when multiple feedback keywords appear, and a product type has multiple type keywords, then obtain the sum of the cosine similarities of each feedback keyword and the type keyword in the corresponding product type, and then calculate the sum of the corresponding sums of each feedback keyword as the cosine similarity value of the product feedback information and the corresponding product type, and use the product type corresponding to the largest cosine similarity value as the corresponding product defect type. In another embodiment, when multiple feedback keywords appear, and a product type has multiple type keywords, then obtain the maximum value of the cosine similarity of each feedback keyword and the type keyword in the corresponding product type, and then calculate the sum of the corresponding maximum values ​​of each feedback keyword as the cosine similarity value of the product feedback information and the corresponding product type, and use the product type corresponding to the largest cosine similarity value as the corresponding product defect type; this method can avoid inaccurate matching results caused by a large difference in the number of type keywords.

[0074] Generate a defect traceability report according to the product defect type and its corresponding type feature data, including: obtaining a type code table in the type feature data of the product type being normal, and marking the product code in the type code table as a normal code; integrating the stage cluster numbers corresponding to each normal code in the same production stage into a normal stage cluster set corresponding to the production stage;

[0075] Extract the type code table in the type feature data corresponding to the product defect type; extract the stage cluster number of each production stage corresponding to each product code in the type code table; determine whether each of the stage cluster numbers is in the normal stage cluster set of the corresponding production stage, if yes, set the corresponding production stage as the normal stage; if not, set the corresponding production stage as the abnormal stage, and generate the abnormal probability coefficient of the corresponding production stage according to the stage cluster number that is not in the normal stage cluster set and the reference degree coefficient of its corresponding product code;

[0076] Each abnormal stage and its corresponding abnormal probability coefficient are integrated into a defect tracing report; specifically, the abnormal probability coefficient corresponding to each abnormal stage, as well as the stage cluster number that does not belong to the normal stage cluster set under each abnormal stage and its corresponding reference coefficient are integrated into a defect tracing report.

[0077] The abnormal probability coefficient of the corresponding production stage is generated according to the stage cluster number that is not in the normal stage cluster set and the reference coefficient of its corresponding product code, including:

[0078] Get the stage cluster number that is not in the normal stage cluster set, and the reference coefficient of the corresponding product code, and mark them as CXi; calculate the abnormal probability coefficient YG of the corresponding production stage by formula YG=∑(CXi) / ∑(CXj); where i is the number of the stage cluster number that is not in the normal stage cluster set, i=1, 2, ..., I, I is the total number of stage cluster numbers that are not in the normal stage cluster set; j is the number of the stage cluster numbers of the corresponding product type in the corresponding production stage, j=1, 2, ..., J, J is the total number of the stage cluster numbers of the corresponding product type in the corresponding production stage.

[0079] In this embodiment, the abnormal probability coefficient under the abnormal stage is calculated by the above formula; when the sum of the reference degree coefficients of the numbers in the normal stage cluster set of each stage cluster number under the corresponding abnormal stage is lower, it means that the probability of abnormality occurring in the corresponding stage is greater, and the corresponding abnormal probability coefficient is greater.

[0080] A defect alarm is generated according to stage production data and defect characteristic data, including: obtaining production data of each stage, integrating production parameters in the production data of each stage into corresponding characteristic vectors; obtaining the stage cluster number of the cluster cluster to which each characteristic vector belongs; and integrating the stage cluster number corresponding to each production stage into a partial product code, and using the partial product code to query the type code with the highest overlap in the type code table; when the reference degree coefficient of the product type corresponding to the type code is abnormal and is greater than the set coefficient threshold, a defect alarm of the corresponding product type is generated.

[0081] See also Figure 2 The second aspect of the present application provides a product manufacturing information management method based on big data, comprising the following steps:

[0082] Step 1: Obtain several product types and their corresponding product production data;

[0083] Step 2: extract several production parameters from the stage production data corresponding to each production stage in the production data of each product; integrate each production parameter into a feature vector, perform cluster analysis on the feature vectors of the production data of each stage to obtain several clusters and cluster centers;

[0084] Step 3: Generate product codes for corresponding product types based on the clusters corresponding to the production data at each stage;

[0085] Step 4: Generate corresponding product feature data according to the product code, the corresponding product type and type keyword;

[0086] Step 5: Get product feedback information;

[0087] Step 6: Generate a defect traceability report based on product feedback information and product feature data.

[0088] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0089] How this application works:

[0090] The present application obtains several product types and their corresponding product production data; extracts several production parameters from the stage production data corresponding to each production stage in each product production data; integrates each production parameter into a feature vector, and performs cluster analysis on the feature vectors of the production data of each stage to obtain several cluster clusters and cluster centers; generates product codes of corresponding product types according to the cluster clusters corresponding to the production data of each stage; generates corresponding product feature data according to the product code, and the corresponding product type and type keywords; obtains product feedback information; and generates a defect traceability report according to the product feedback information and the product feature data; the present application performs cluster analysis on the data of each production stage in the product production data corresponding to the recorded various product types, so as to distinguish the production data of each stage of various product types, so that when the defect type is subsequently searched, the stage where the abnormality may occur can be quickly located, the amount of excluded data can be reduced, the accuracy of the investigation can be increased, and the traceability efficiency of defective products can be improved.

[0091] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical method of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A product manufacturing information management system based on big data, comprising: Data acquisition module, data analysis module, defect tracing module and database; characterized in that, The data acquisition module is used to acquire product feedback information through the user terminal; The data analysis module: obtains a number of product types and their corresponding product production data through a database; generates a number of product feature data according to the product types and product production data; The defect tracing module obtains product feedback information and product feature data, and generates a defect tracing report based on the product feedback information and product feature data.

2. A product manufacturing information management system based on big data according to claim 1, characterized in that: The generating of a number of product characteristic data according to the product type and product production data includes: Extracting several production parameters from the stage production data corresponding to each production stage in the production data of each product; integrating each production parameter into a feature vector, and performing cluster analysis on the feature vectors of the production data of each stage to obtain several clusters and cluster centers; Generate product codes of corresponding product types according to clusters corresponding to production data of each stage; extract product types and type keywords in product production data, and generate corresponding product feature data according to product codes and corresponding product types and type keywords.

3. A product manufacturing information management system based on big data according to claim 2, characterized in that: The cluster analysis of the characteristic vectors of the production data at each stage to obtain a number of clusters and cluster centers includes the following steps: Step 1: Obtain the domain density of the feature points corresponding to each feature vector in the domain corresponding to the set radius; take the feature vector whose domain density is greater than the set density threshold as the initial core point; Step 2: Obtain the Euclidean distance between each feature point and each initial core point, and assign the feature point to the initial cluster of the initial core point with the smallest Euclidean distance, so as to form several initial clusters with each initial core point; Step 3: Merge the initial clusters whose Euclidean distance between each initial cluster is less than the set distance threshold and whose sum of cluster density between adjacent clusters is greater than the corresponding density threshold to generate clusters, and take the point with the largest cluster density in the cluster as the cluster center.

4. A product manufacturing information management system based on big data according to claim 3, characterized in that: The generating of product codes of corresponding product types according to the clusters corresponding to the production data of each stage includes: Numbering a number of clusters corresponding to the stage production data to obtain a number of stage cluster numbers corresponding to the corresponding stage production data; The stage cluster number to which the production data of each stage corresponds to each product production data is obtained, and the stage cluster numbers are integrated according to the order of the stage production data to obtain the product code.

5. The product manufacturing information management system based on big data according to claim 1 is characterized in that: The generating of corresponding product feature data according to the product code, the corresponding product type and the type keyword includes: Extract each product code corresponding to each product production data of the same product type, obtain the same type number of the product code in the product type and the different type number of the product code in other product types, generate the reference coefficient of the corresponding product code in the product type according to the same type number and the different type number; sort each product code in descending order of the reference coefficient to obtain a type code table; integrate the type code table and the corresponding type keywords into type characteristic data of the corresponding product type; Integrate each type of feature data and its corresponding product type into product feature data.

6. A product manufacturing information management system based on big data according to claim 5, characterized in that: The step of generating a reference coefficient of the corresponding product code in the product type according to the same type quantity and the different type quantity includes: Get the same code number and the different code number, and mark them as TB and YB respectively. The reference coefficient CX of the corresponding product code in the product type is calculated by the formula CX=α1×exp(β1×TB / TZB)+α2×exp(-β2×YB / TB); wherein α1 and α2 are proportional coefficients; β1 and β2 are adjustment coefficients; TZB is the total number of product codes corresponding to the product type.

7. The product manufacturing information management system based on big data according to claim 1 is characterized in that: The defect tracing report generated according to the product feedback information and product feature data includes: Obtain product feedback information; extract feedback keywords from product feedback information and convert feedback keywords into feedback word vectors; Extract type keywords from each type feature data in the product feature data, and convert the type keywords into type word vectors; calculate the cosine similarity between each feedback word vector and the type word vector corresponding to each type keyword, and take the product type corresponding to the type keyword with the largest cosine similarity value as the product defect type; generate a defect traceability report based on the product defect type and its corresponding type feature data.

8. The product manufacturing information management system based on big data according to claim 1 is characterized in that: The defect tracing report is generated according to the product defect type and its corresponding type characteristic data, including: Obtain a type code table in the type feature data of a product type that is normal, and mark the product code in the type code table as a normal code; integrate the stage cluster numbers corresponding to each normal code in the same production stage into a normal stage cluster set corresponding to the production stage; Extract the type code table in the type feature data corresponding to the product defect type; extract the stage cluster number of each production stage corresponding to each product code in the type code table; determine whether each of the stage cluster numbers is in the normal stage cluster set of the corresponding production stage, if yes, set the corresponding production stage as the normal stage; if not, set the corresponding production stage as the abnormal stage, and generate the abnormal probability coefficient of the corresponding production stage according to the stage cluster number that is not in the normal stage cluster set and the reference degree coefficient of its corresponding product code; Integrate each abnormal stage and its corresponding abnormal probability coefficient into a defect tracing report.

9. A product manufacturing information management system based on big data according to claim 8, characterized in that: The generating of the abnormal probability coefficient of the corresponding production stage according to the reference coefficient of the stage cluster number not in the normal stage cluster set and its corresponding product code includes: Get the stage cluster number that is not in the normal stage cluster set, and the reference coefficient of the corresponding product code, and mark them as CXi; calculate the abnormal probability coefficient YG of the corresponding production stage by formula YG=∑(CXi) / ∑(CXj); where i is the number of the stage cluster number that is not in the normal stage cluster set, i=1, 2, ..., I, I is the total number of stage cluster numbers that are not in the normal stage cluster set; j is the number of the stage cluster numbers of the corresponding product type in the corresponding production stage, j=1, 2, ..., J, J is the total number of the stage cluster numbers of the corresponding product type in the corresponding production stage.

10. A product manufacturing information management method based on big data, based on the product manufacturing information management system based on big data according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step 1: Obtain several product types and their corresponding product production data; Step 2: extract several production parameters from the stage production data corresponding to each production stage in the production data of each product; integrate each production parameter into a feature vector, perform cluster analysis on the feature vectors of the production data of each stage to obtain several clusters and cluster centers; Step 3: Generate product codes for corresponding product types based on the clusters corresponding to the production data at each stage; Step 4: Generate corresponding product feature data according to the product code, the corresponding product type and type keyword; Step 5: Get product feedback information; Step 6: Generate a defect traceability report based on product feedback information and product feature data.

Images