Lipstick quality inspection method and system

Abstract

The present application relates to a lipstick quality inspection method and system, and the method includes the following steps: obtaining the quality inspection requirements of the target lipstick, and generating a preliminary quality inspection plan through intelligent matching; collecting the raw material data and production data of the target lipstick, screening abnormal data to correct the preliminary quality inspection plan to obtain the target quality inspection plan; generating a scheduling instruction based on the target quality inspection plan, delivering the target lipstick to each quality inspection device in turn, and issuing equipment parameter information to each quality inspection device for parameter setting according to the target quality inspection plan; generating a control instruction according to each quality inspection process information, controlling each quality inspection device to perform quality inspection on the target lipstick in turn, and generating a quality inspection report to send to the management personnel. The present application has the effect of effectively improving the efficiency and quality of lipstick quality inspection.

Description

Technical Field

[0001] The present application relates to the field of lipstick quality inspection, and in particular to a lipstick quality inspection method and system. Background Art

[0002] The quality inspection of lipstick is a multi-level and multi-faceted process, involving inspections of microorganisms, mineral oils, chemical composition, physical properties, etc., to ensure that the product meets the standard requirements in terms of safety, stability and user experience.

[0003] At present, the production lines of lipstick manufacturers have not yet reached the level of full automation. Many links require full-time manual intervention online. The quality inspection part is one of the bottlenecks for increasing production capacity, and there is an upper limit to the efficiency of manual quality inspection. Manpower cannot be online for a long time, which not only causes physical harm to quality inspection workers, but also reduces the accuracy of quality inspection over time. Even when workers have full energy, manual quality inspection may still result in missed inspections and false inspections. The statistics of production data are also limited to manual records.

[0004] The Chinese patent application with publication number CN114549476A discloses a transfer learning target detection method for lipstick quality inspection, which belongs to the field of deep learning target detection technology. A transfer learning target detection method for lipstick quality inspection is based on the YOLO v5 network structure and optimizes the network structure. The application adopts Fine-tuning transfer learning to enhance the generalization performance of the model, and visually inspects the lipstick through the machine learning model to determine the appearance defects of the lipstick. The Chinese patent application with publication number CN117990644A discloses a lipstick quality identification method, which performs physical and chemical testing on the lipstick, and detects the color, smell, DSC and IR of the lipstick.

[0005] It can be seen from this that the existing lipstick quality inspection often only conducts quality inspection on a certain quality inspection or multiple quality inspection items of the same type, and is unable to meet the actual quality inspection needs of lipstick according to actual production conditions, resulting in low quality inspection efficiency and quality of lipstick. Summary of the invention

[0006] In order to solve the problems of low efficiency and quality of existing lipstick quality inspection, the present application provides a lipstick quality inspection method and system.

[0007] In the first aspect, the present application provides a lipstick quality inspection method, which adopts the following technical solution:

[0008] A lipstick quality inspection method comprises the following steps:

[0009] The quality inspection requirements of the target lipstick are obtained, and a preliminary quality inspection plan is generated through intelligent matching. The quality inspection requirements include one or more of physical and chemical inspection requirements, bionic inspection requirements, microbial inspection requirements, harmful substance inspection requirements, performance inspection requirements, appearance defect inspection requirements, and specific inspection requirements; the preliminary quality inspection plan includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information and equipment parameter information;

[0010] Collect the raw material data and production data of the target lipstick, filter out abnormal data, and modify the preliminary quality inspection plan to obtain the target quality inspection plan;

[0011] Generate scheduling instructions based on the target quality inspection plan, transport the target lipstick to each quality inspection device in turn, and send equipment parameter information to each quality inspection device for parameter setting according to the target quality inspection plan;

[0012] Generate control instructions based on each quality inspection process information, control each quality inspection equipment to perform quality inspection on the target lipstick in turn, and generate a quality inspection report and send it to the management staff.

[0013] Preferably, the obtaining of the quality inspection requirements of the target lipstick and the intelligent matching to generate a preliminary quality inspection plan specifically include the following steps:

[0014] Obtaining the demand information uploaded by the user, wherein the demand information includes general demand information and remarks information;

[0015] Perform data cleaning on demand information and extract the quality inspection requirements of the target lipstick;

[0016] The quality inspection requirements are input into the preset solution matching model for intelligent matching to generate a preliminary quality inspection solution; the solution matching model is a machine learning model obtained by iterative training with historical quality inspection data.

[0017] Preferably, the collecting of raw material data and production data of the target lipstick, screening of abnormal data to amend the preliminary quality inspection plan to obtain the target quality inspection plan specifically comprises the following steps:

[0018] Collect the raw material data of the target lipstick, and identify and filter out abnormal raw material data through a preset solution correction model; the solution correction model is based on the ChatGPT4.0 model and is trained through historical production data and historical quality inspection data;

[0019] Collect the production data of the target lipstick and identify and filter out abnormal production data through the preset solution correction model;

[0020] The solution modification model generates the estimated yield of each quality inspection item based on the raw material abnormality data and the production abnormality data, and calculates the predicted defect rate of each quality inspection item through the preset defect rate calculation formula; the defect rate calculation formula is specifically:

[0021] Y=1-(Xi+Pi) / 2;

[0022] Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item;

[0023] According to the predicted value of the defective rate of each quality inspection item, the quality inspection items of the preliminary quality inspection plan are sorted from large to small, and the quality inspection process information of each quality inspection item is integrated in sequence to obtain the target quality inspection plan, which includes quality inspection equipment for at least one quality inspection item as well as quality inspection process information, equipment parameter information and complete quality inspection process information.

[0024] Preferably, the collecting of the raw material data of the target lipstick and identifying and filtering out the abnormal raw material data through the preset scheme correction model specifically include the following steps:

[0025] Collecting raw material data of the target lipstick, the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, the raw material image data includes raw material in-warehouse image, raw material out-warehouse image and raw material online image;

[0026] The raw material purchase chain data and raw material inventory data are cleaned and screened through the preset solution correction model to obtain abnormal inventory data;

[0027] The solution modification model identifies the raw material image data based on the raw material inventory data, verifies the extracted abnormal inventory data based on the identification results, and determines whether there are new abnormal representations. If so, an alarm message is sent to the management personnel and image abnormality data is generated;

[0028] The abnormal inventory data and image abnormal data are packaged to generate raw material abnormal data.

[0029] Preferably, it also includes: generating monitoring and early warning information based on the quality inspection report through a preset negative feedback model and sending it to the management personnel, wherein the monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts and production data mismatch prompts.

[0030] In the second aspect, the present application provides a lipstick quality inspection system, which adopts the following technical solution:

[0031] A lipstick quality inspection system, comprising:

[0032] A scheme matching module is used to obtain the quality inspection requirements of the target lipstick, and to generate a preliminary quality inspection scheme through intelligent matching. The quality inspection requirements include one or more of physical and chemical inspection requirements, bionic inspection requirements, microbial inspection requirements, harmful substance inspection requirements, performance inspection requirements, appearance defect inspection requirements, and specific inspection requirements; the preliminary quality inspection scheme includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information and equipment parameter information;

[0033] The scheme correction module is used to collect the raw material data and production data of the target lipstick, filter abnormal data, and correct the preliminary quality inspection scheme to obtain the target quality inspection scheme;

[0034] The production scheduling module is used to generate scheduling instructions based on the target quality inspection plan, transport the target lipstick to each quality inspection device in sequence, and send equipment parameter information to each quality inspection device for parameter setting according to the target quality inspection plan;

[0035] The quality inspection control module is used to generate control instructions according to the quality inspection process information, control each quality inspection device to perform quality inspection on the target lipstick in turn, and generate a quality inspection report and send it to the management personnel.

[0036] Preferably, the solution matching module includes:

[0037] A demand acquisition unit, used to acquire demand information uploaded by a user, wherein the demand information includes regular demand information and remark information;

[0038] The data processing unit is used to clean the demand information and extract the quality inspection requirements of the target lipstick;

[0039] The solution matching unit is used to input the quality inspection requirements into a preset solution matching model for intelligent matching to generate a preliminary quality inspection solution; the solution matching model is a machine learning model obtained by iterative training through historical quality inspection data.

[0040] Preferably, the scheme modification module includes:

[0041] The raw material extraction unit is used to collect the raw material data of the target lipstick, and identify and filter out abnormal raw material data through a preset solution correction model; the solution correction model is based on the ChatGPT4.0 model and is trained through historical production data and historical quality inspection data;

[0042] The production extraction unit is used to collect the production data of the target lipstick and identify and filter out abnormal production data through a preset solution correction model;

[0043] The defect prediction unit is used for the solution correction model to generate the estimated yield of each quality inspection item based on the raw material abnormal data and the production abnormal data, and calculate the defect rate prediction value of each quality inspection item through the preset defect rate calculation formula; the defect rate calculation formula is specifically:

[0044] Y=1-(Xi+Pi) / 2;

[0045] Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item;

[0046] The scheme modification unit is used to sort the quality inspection items of the preliminary quality inspection scheme from large to small according to the predicted defective rate value of each quality inspection item, and integrate the quality inspection process information of each quality inspection item in sequence to obtain a target quality inspection scheme, wherein the target quality inspection scheme includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information, equipment parameter information and complete quality inspection process information.

[0047] Preferably, the raw material extraction unit comprises:

[0048] A data collection subunit, used to collect raw material data of the target lipstick, wherein the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, wherein the raw material image data includes raw material in-warehouse image, raw material out-warehouse image and raw material online image;

[0049] A data processing subunit is used to clean and filter the raw material purchase chain data and raw material inventory data through a preset solution correction model to obtain abnormal inventory data;

[0050] The image verification subunit is used to identify the raw material image data based on the raw material inventory data through the solution correction model, verify the extracted abnormal inventory data based on the recognition result, and determine whether there is a new abnormal representation. If so, an alarm message is sent to the management personnel and image abnormality data is generated;

[0051] The data packaging subunit is used to package the abnormal inventory data and the abnormal image data to generate the abnormal raw material data.

[0052] Preferably, it also includes: a negative feedback module, which is used to generate monitoring and early warning information based on the quality inspection report through a preset negative feedback model and send it to the management personnel, wherein the monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts and production data mismatch prompts.

[0053] In summary, the present application includes at least one of the following beneficial technical effects:

[0054] 1. According to the actual quality inspection needs of the target lipstick, combined with the actual production and raw material storage conditions of the enterprise, a quality inspection plan is generated for its customized decision. On the basis of meeting the various quality inspection needs of different lipsticks, quality inspection equipment is dispatched, equipment parameters are set, quality inspection processes are controlled, and production and raw material management defects are corrected, so as to realize intelligent automation and all-round high-precision quality inspection of lipsticks, so as to effectively improve the efficiency and quality of lipstick quality inspection;

[0055] 2. By setting up a scheme correction model, the abnormal items in the raw material data and production data are extracted, and the number of defective products that may be caused is predicted based on these abnormal items, and then the estimated yield of each quality inspection item of the target lipstick is generated. Then, the defective rate of each quality inspection item of the target lipstick is calculated and determined in combination with the historical quality inspection data. The preliminary quality inspection scheme is corrected based on the defective rate, and the various quality inspection items of the preliminary quality inspection scheme are sorted, and the quality inspection items with high estimated defective rates are prioritized. Defective products are eliminated as much as possible, and the quality inspection volume of subsequent quality inspection processes is reduced. On the basis of saving manpower and material resources, the effect of improving the quality inspection efficiency of lipstick is achieved;

[0056] 3. Through the setting of negative feedback model, after completing the quality inspection, the defects in raw material storage and lipstick production are determined based on the quality inspection report, and corresponding suggestions are generated by decision-making, which is conducive to the positive optimization of raw material storage, lipstick production and lipstick quality inspection processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a method flow chart of a lipstick quality inspection method in an embodiment of the present application;

[0058] Figure 2 is a flow chart of a method for generating a preliminary quality inspection plan in an embodiment of the present application;

[0059] Figure 3 is a flow chart of a method for correcting a preliminary quality inspection plan to obtain a target quality inspection plan in an embodiment of the present application;

[0060] Figure 4 is a flow chart of a method for screening out abnormal data of raw materials in an embodiment of the present application;

[0061] Figure 5 It is a system block diagram of a lipstick quality inspection system in an embodiment of the present application.

[0062] Explanation of the accompanying drawings: 1. Scheme matching module; 11. Demand acquisition unit; 12. Data processing unit; 13. Scheme matching unit; 2. Scheme correction module; 21. Raw material extraction unit; 211. Data acquisition subunit; 212. Data processing subunit; 213. Image verification subunit; 214. Data packaging subunit; 22. Production extraction unit; 23. Defect prediction unit; 24. Scheme correction unit; 3. Production scheduling module; 4. Quality inspection control module; 5. Negative feedback module. DETAILED DESCRIPTION

[0063] The following is combined with Figure 1-5 This application is described in further detail.

[0064] The present application embodiment discloses a lipstick quality inspection method. Figure 1 , a lipstick quality inspection method, comprising the following steps:

[0065] S1. Generate a preliminary quality inspection plan: obtain the quality inspection requirements of the target lipstick, and generate a preliminary quality inspection plan through intelligent matching. The quality inspection requirements include one or more of physical and chemical testing requirements, bionic testing requirements, microbial testing requirements, harmful substance testing requirements, performance testing requirements, appearance defect testing requirements, and specific testing requirements; the preliminary quality inspection plan includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information and equipment parameter information;

[0066] S2. Correction to obtain the target quality inspection plan: Collect the raw material data and production data of the target lipstick, filter out abnormal data, correct the preliminary quality inspection plan to obtain the target quality inspection plan;

[0067] S3, dispatching equipment and setting parameters: generating dispatch instructions based on the target quality inspection plan, delivering the target lipstick to each quality inspection equipment in sequence, and issuing equipment parameter information to each quality inspection equipment for parameter setting according to the target quality inspection plan;

[0068] S4, perform quality inspection on the target lipstick: generate control instructions according to each quality inspection process information, control each quality inspection device to perform quality inspection on the target lipstick in turn, and generate a quality inspection report and send it to the management personnel;

[0069] S5. Negative feedback suggestions: Based on the quality inspection report, the monitoring and early warning information is generated through the preset negative feedback model and sent to the management personnel. The monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts, and production data mismatch prompts. According to the actual quality inspection needs of the target lipstick, combined with the actual production and raw material storage conditions of the enterprise, a quality inspection plan is generated for its customized decision-making. On the basis of meeting the various quality inspection needs of different lipsticks, the quality inspection equipment is dispatched, the equipment parameters are set, the quality inspection process is controlled, and the production and raw material management defects are corrected, so as to realize intelligent automation and all-round high-precision quality inspection of lipsticks, so as to effectively improve the efficiency and quality of lipstick quality inspection.

[0070] In addition, the negative feedback model is a neural network model that is obtained through deep learning using historical raw material data, historical production data, historical raw material data, and production and inventory correction data as samples. Through the setting of the negative feedback model, after completing the quality inspection, the defects in raw material storage and lipstick production are determined based on the quality inspection report, and corresponding suggestions are generated for decision-making, which is conducive to the positive optimization of raw material storage, lipstick production, and lipstick quality inspection processes. The correctness of production data and raw material collection is verified, and management personnel are prompted to calibrate data collection equipment. On the basis of achieving high-precision detection of raw material storage and lipstick production, the accuracy of the quality inspection plan is improved, and then the production weaknesses can be more targeted for priority quality inspection, and the quality inspection efficiency is improved on the basis of reducing the workload of quality inspection.

[0071] Reference Figure 2 The obtaining of the quality inspection requirements of the target lipstick and the intelligent matching to generate a preliminary quality inspection plan specifically include the following steps:

[0072] A1. Obtaining demand information: Obtaining demand information uploaded by the user, the demand information includes general demand information and remarks information;

[0073] A2. Obtain quality inspection requirements: perform data cleaning on the demand information and extract the quality inspection requirements of the target lipstick;

[0074] A3. Generate a preliminary quality inspection plan: Input the quality inspection requirements into the preset solution matching model for intelligent matching to generate a preliminary quality inspection plan; the solution matching model is a machine learning model, which is obtained through iterative training of historical quality inspection data. Through the above steps, a quality inspection plan can be customized according to the user's actual quality inspection needs. Even in the face of unconventional quality inspection requirements, it can be pointed out through the remarks information, and then a customized quality inspection plan is generated, which improves the universality of solution matching on the basis of improving user satisfaction.

[0075] Reference Figure 3 The collecting of raw material data and production data of the target lipstick, screening of abnormal data, and correction of the preliminary quality inspection plan to obtain the target quality inspection plan specifically include the following steps:

[0076] B1. Screen out abnormal raw material data: Collect the raw material data of the target lipstick, and identify and screen out abnormal raw material data through a preset solution correction model; the solution correction model is based on the ChatGPT4.0 model and is trained through historical production data and historical quality inspection data;

[0077] B2. Screen out abnormal production data: Collect the production data of the target lipstick, and identify and screen out abnormal production data through the pre-set solution correction model;

[0078] B3. Calculate the predicted value of the defective rate of each quality inspection item: The solution modification model generates the estimated yield of each quality inspection item based on the raw material abnormality data and the production abnormality data, and calculates the predicted value of the defective rate of each quality inspection item through the preset defective rate calculation formula; the specific defective rate calculation formula is:

[0079] Y=1-(Xi+Pi) / 2;

[0080] Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item;

[0081] B4. Correction to obtain the target quality inspection plan: Sort the quality inspection items of the preliminary quality inspection plan from large to small according to the predicted value of the defective rate of each quality inspection item, and integrate the quality inspection process information of each quality inspection item in order to obtain the target quality inspection plan, which includes the quality inspection equipment and quality inspection process information, equipment parameter information and complete quality inspection process information of at least one quality inspection item. By setting up a plan correction model, the abnormal items in the raw material data and production data are extracted, and the number of defective products that may be caused is predicted based on these abnormal items, and then the expected yield of each quality inspection item of the target lipstick is generated, and then the defective rate of each quality inspection item of the target lipstick is calculated and determined in combination with the historical quality inspection data. The preliminary quality inspection plan is corrected based on the defective rate, and the various quality inspection items of the preliminary quality inspection plan are sorted, and the quality inspection items with high predicted defective rates are prioritized, and defective products are eliminated as much as possible, reducing the quality inspection volume of subsequent quality inspection processes, saving manpower and material resources, and achieving the effect of improving the quality inspection efficiency of lipstick.

[0082] Reference Figure 4 The method of collecting the raw material data of the target lipstick and identifying and filtering out the abnormal raw material data through the preset scheme correction model specifically includes the following steps:

[0083] C1. Collecting raw material data of the target lipstick, the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, the raw material image data includes raw material inbound image, raw material outbound image and raw material online image;

[0084] C2. Screening to obtain abnormal inventory data: Using the preset solution correction model, the raw material purchase chain data and raw material inventory data are cleaned and screened to obtain abnormal inventory data;

[0085] C3. Verify the extracted abnormal inventory data: The solution modification model recognizes the raw material image data based on the raw material inventory data, and verifies the extracted abnormal inventory data based on the recognition result to determine whether there is any new abnormal representation. If so, an alarm message is sent to the management personnel and abnormal image data is generated;

[0086] C4. Package and generate abnormal raw material data: Package abnormal inventory data and abnormal image data to generate abnormal raw material data. Through the setting of the solution correction model, based on the powerful data processing and image recognition capabilities of the ChatGPT4.0 model, the inventory process of raw materials is accurately detected, and abnormal inventory data generated by various unexpected factors before the production of raw materials is screened out, which is helpful for the subsequent accurate prediction of the yield rate of each quality inspection item. Then, the raw material image data is identified through the solution correction model, which can effectively verify the correctness of the raw material data, further improve the data accuracy, and effectively expose the defects of raw material storage and scheduling process data collection, which is helpful for better raw material storage and data monitoring, and facilitates the provision of accurate raw material data to achieve accurate correction of the preliminary quality inspection plan, so as to improve the efficiency of lipstick quality inspection.

[0087] The present application also discloses a lipstick quality inspection system. Figure 5 , a lipstick quality inspection system, comprising:

[0088] Solution matching module 1 is used to obtain the quality inspection requirements of the target lipstick, and to generate a preliminary quality inspection solution through intelligent matching. The quality inspection requirements include one or more of physical and chemical testing requirements, bionic testing requirements, microbial testing requirements, harmful substance testing requirements, performance testing requirements, appearance defect testing requirements, and specific testing requirements; the preliminary quality inspection solution includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information and equipment parameter information;

[0089] The scheme correction module 2 is used to collect the raw material data and production data of the target lipstick, filter abnormal data, and correct the preliminary quality inspection scheme to obtain the target quality inspection scheme;

[0090] Production scheduling module 3, used to generate scheduling instructions based on the target quality inspection plan, transport the target lipstick to each quality inspection equipment in sequence, and send equipment parameter information to each quality inspection equipment for parameter setting according to the target quality inspection plan;

[0091] The quality inspection control module 4 is used to generate control instructions according to each quality inspection process information, control each quality inspection device to perform quality inspection on the target lipstick in turn, and generate a quality inspection report and send it to the management personnel;

[0092] The negative feedback module 5 is used to generate monitoring and early warning information based on the quality inspection report through a preset negative feedback model and send it to the management personnel. The monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts, and production data mismatch prompts. According to the actual quality inspection requirements of the target lipstick, combined with the actual production and raw material storage conditions of the enterprise, a quality inspection plan is generated for its customized decision-making. On the basis of meeting the various quality inspection requirements of different lipsticks, the quality inspection equipment is dispatched, the equipment parameters are set, the quality inspection process is controlled, and the production and raw material management defects are corrected, so as to realize intelligent automation and all-round high-precision quality inspection of lipsticks, so as to effectively improve the efficiency and quality of lipstick quality inspection.

[0093] Reference Figure 5 , the scheme matching module 1 includes:

[0094] The demand acquisition unit 11 is used to acquire the demand information uploaded by the user, wherein the demand information includes regular demand information and remark information;

[0095] The data processing unit 12 is used to clean the demand information and extract the quality inspection requirements of the target lipstick;

[0096] The solution matching unit 13 is used to input the quality inspection requirements into the preset solution matching model for intelligent matching to generate a preliminary quality inspection solution; the solution matching model is a machine learning model, which is obtained by iterative training through historical quality inspection data. Through the setting of the solution matching module 1, it is possible to customize the quality inspection solution according to the actual quality inspection requirements of the user. Even in the face of unconventional quality inspection requirements, it can be pointed out through the remarks information, and then the quality inspection solution is customized to generate, thereby improving the universality of solution matching on the basis of improving user satisfaction.

[0097] Reference Figure 5 , the scheme revision module 2 includes:

[0098] The raw material extraction unit 21 is used to collect the raw material data of the target lipstick, and identify and filter out the abnormal raw material data through a preset solution correction model; the solution correction model is based on the ChatGPT4.0 model and is trained through historical production data and historical quality inspection data;

[0099] The production extraction unit 22 is used to collect the production data of the target lipstick and identify and filter out the abnormal production data through the preset solution correction model;

[0100] The defect prediction unit 23 is used for the solution correction model to generate the estimated yield of each quality inspection item based on the raw material abnormality data and the production abnormality data, and calculate the defect rate prediction value of each quality inspection item through a preset defect rate calculation formula; the defect rate calculation formula is specifically:

[0101] Y=1-(Xi+Pi) / 2;

[0102] Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item;

[0103] The scheme correction unit 24 is used to sort the various quality inspection items of the preliminary quality inspection scheme from large to small according to the predicted value of the defective rate of each quality inspection item, and integrate the quality inspection process information of each quality inspection item in order to obtain the target quality inspection scheme, wherein the target quality inspection scheme includes at least one quality inspection item of quality inspection equipment and quality inspection process information, equipment parameter information and complete quality inspection process information. By setting a scheme correction model, the abnormal items in the raw material data and production data are extracted, and the number of defective products that may be caused is predicted based on these abnormal items, and then the expected yield of each quality inspection item of the target lipstick is generated, and then the defective rate of each quality inspection item of the target lipstick is calculated and determined in combination with the historical quality inspection data, and the preliminary quality inspection scheme is corrected based on the defective rate, and the various quality inspection items of the preliminary quality inspection scheme are sorted, and the quality inspection items with high predicted defective rates are given priority, and defective products are eliminated as much as possible, and the quality inspection amount of each subsequent quality inspection process is reduced, and the effect of improving the quality inspection efficiency of lipstick is achieved on the basis of saving manpower and material resources.

[0104] Reference Figure 5 , the raw material extraction unit 21 comprises:

[0105] The data collection subunit 211 is used to collect the raw material data of the target lipstick, wherein the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, wherein the raw material image data includes raw material in-warehouse image, raw material out-warehouse image and raw material online image;

[0106] The data processing subunit 212 is used to clean and filter the raw material purchase chain data and the raw material inventory data through a preset solution correction model to obtain abnormal inventory data;

[0107] The image verification subunit 213 is used to identify the raw material image data based on the raw material inventory data through the solution correction model, verify the extracted abnormal inventory data based on the recognition result, and determine whether there is a new abnormal representation. If so, send an alarm message to the management personnel and generate image abnormality data;

[0108] The data packaging subunit 214 is used to package abnormal inventory data and abnormal image data to generate abnormal raw material data. Through the setting of the solution correction model, based on the powerful data processing and image recognition capabilities of the ChatGPT4.0 model, the inventory process of raw materials is accurately detected, and the abnormal inventory data generated by various unexpected factors before the production of raw materials is screened out, which is helpful for the subsequent accurate prediction of the yield rate of each quality inspection item. Then, the raw material image data is identified through the solution correction model, which can effectively verify the correctness of the raw material data, further improve the data accuracy, and effectively expose the defects of raw material storage and scheduling process data collection, which is helpful for better raw material storage and data monitoring, and is convenient for providing accurate raw material data to realize accurate correction of the preliminary quality inspection plan, so as to improve the efficiency of lipstick quality inspection.

[0109] The embodiment of the present application also discloses a computer-readable storage medium, which stores a computer program that can be loaded by a processor and executed as in the above method. The computer-readable storage medium includes, for example: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0110] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in this field can still combine, add, delete or make other adjustments to the features in the various embodiments of the present invention according to the circumstances without conflict, without making creative work, so as to obtain different other technical solutions that do not deviate from the concept of the present invention in essence, and these technical solutions also belong to the scope of protection of the present invention.

Claims

1. A lipstick quality inspection method, characterized in that: The following steps are involved: The quality inspection requirements of the target lipstick are obtained, and a preliminary quality inspection plan is generated through intelligent matching. The quality inspection requirements include one or more of physical and chemical inspection requirements, bionic inspection requirements, microbial inspection requirements, harmful substance inspection requirements, performance inspection requirements, appearance defect inspection requirements, and specific inspection requirements; the preliminary quality inspection plan includes quality inspection equipment for at least one quality inspection item, as well as quality inspection process information and equipment parameter information; Collect the raw material data and production data of the target lipstick, filter out abnormal data, and modify the preliminary quality inspection plan to obtain the target quality inspection plan; Generate scheduling instructions based on the target quality inspection plan, transport the target lipstick to each quality inspection device in turn, and send device parameter information to each quality inspection device for parameter setting according to the target quality inspection plan; Generate control instructions based on each quality inspection process information, control each quality inspection equipment to inspect the target lipstick in turn, and generate a quality inspection report and send it to the management staff; The step of collecting the raw material data and production data of the target lipstick, screening abnormal data, and correcting the preliminary quality inspection plan to obtain the target quality inspection plan specifically includes the following steps: Collect the raw material data of the target lipstick, and identify and filter out abnormal raw material data through a preset solution correction model; the solution correction model is based on the ChatGPT4.0 model and is trained through historical production data and historical quality inspection data; Collect the production data of the target lipstick and identify and filter out abnormal production data through the preset solution correction model; The solution modification model generates the estimated yield of each quality inspection item based on the raw material abnormality data and the production abnormality data, and calculates the predicted defect rate of each quality inspection item through the preset defect rate calculation formula; the defect rate calculation formula is specifically: Y=1-(Xi+Pi) / 2; Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item; According to the predicted value of the defect rate of each quality inspection item, the quality inspection items of the preliminary quality inspection plan are sorted from large to small, and the quality inspection process information of each quality inspection item is integrated in sequence to obtain a target quality inspection plan, wherein the target quality inspection plan includes quality inspection equipment and quality inspection process information, equipment parameter information and complete quality inspection process information of at least one quality inspection item; The step of collecting the raw material data of the target lipstick and identifying and filtering out abnormal raw material data through a preset scheme correction model specifically includes the following steps: Collecting raw material data of the target lipstick, the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, the raw material image data includes raw material in-warehouse image, raw material out-warehouse image and raw material online image; The raw material purchase chain data and raw material inventory data are cleaned and screened through the preset solution correction model to obtain abnormal inventory data; The solution modification model identifies the raw material image data based on the raw material inventory data, verifies the extracted abnormal inventory data based on the identification results, and determines whether there are new abnormal representations. If so, an alarm message is sent to the management personnel and image abnormality data is generated; The abnormal inventory data and image abnormal data are packaged to generate raw material abnormal data.

2. A lipstick quality inspection method according to claim 1, characterized in that: The obtaining of the quality inspection requirements of the target lipstick and the intelligent matching to generate a preliminary quality inspection plan specifically include the following steps: Obtaining the demand information uploaded by the user, wherein the demand information includes general demand information and remarks information; Perform data cleaning on demand information and extract the quality inspection requirements of the target lipstick; The quality inspection requirements are input into the preset solution matching model for intelligent matching to generate a preliminary quality inspection solution; the solution matching model is a machine learning model obtained by iterative training with historical quality inspection data.

3. A lipstick quality inspection method according to claim 1, characterized in that: Also includes: Based on the quality inspection report, monitoring and early warning information is generated through a preset negative feedback model and sent to the management personnel. The monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts, and production data mismatch prompts.

4. A lipstick quality inspection system, characterized in that: include: A scheme matching module (1) is used to obtain the quality inspection requirements of the target lipstick and to generate a preliminary quality inspection scheme by intelligent matching. The quality inspection requirements include one or more of physical and chemical inspection requirements, bionic inspection requirements, microbial inspection requirements, harmful substance inspection requirements, performance inspection requirements, appearance defect inspection requirements, and specific inspection requirements; the preliminary quality inspection scheme includes quality inspection equipment for at least one quality inspection item, quality inspection process information, and equipment parameter information; The scheme correction module (2) is used to collect the raw material data and production data of the target lipstick, filter out abnormal data, and correct the preliminary quality inspection scheme to obtain the target quality inspection scheme; The production scheduling module (3) is used to generate scheduling instructions based on the target quality inspection plan, transport the target lipstick to each quality inspection device in sequence, and send device parameter information to each quality inspection device for parameter setting according to the target quality inspection plan; The quality inspection control module (4) is used to generate control instructions according to each quality inspection process information, control each quality inspection device to perform quality inspection on the target lipstick in turn, and generate a quality inspection report and send it to the management personnel; The scheme correction module (2) comprises: A raw material extraction unit (21) is used to collect raw material data of a target lipstick and identify and filter out abnormal raw material data through a preset solution correction model; the solution correction model is based on a ChatGPT4.0 model and is obtained through training with historical production data and historical quality inspection data; A production extraction unit (22) is used to collect production data of the target lipstick and identify and filter out abnormal production data through a preset solution correction model; The defect prediction unit (23) is used for the solution correction model to generate the estimated yield of each quality inspection item based on the raw material abnormality data and the production abnormality data, and to calculate the defect rate prediction value of each quality inspection item through a preset defect rate calculation formula; the defect rate calculation formula is specifically: Y=1-(Xi+Pi) / 2; Where Y is the defective rate of the quality inspection item, Xi is the expected yield of the i-th quality inspection item, and Pi is the historical average yield of the i-th quality inspection item; A scheme modification unit (24) is used to sort the quality inspection items of the preliminary quality inspection scheme from large to small according to the predicted defect rate value of each quality inspection item, and to integrate the quality inspection process information of each quality inspection item in order to obtain a target quality inspection scheme, wherein the target quality inspection scheme includes quality inspection equipment and quality inspection process information, equipment parameter information and complete quality inspection process information of at least one quality inspection item; The raw material extraction unit (21) comprises: A data collection subunit (211) is used to collect raw material data of the target lipstick, wherein the raw material data includes raw material purchase chain data, raw material inventory data and raw material image data, wherein the raw material image data includes raw material in-warehouse image, raw material out-warehouse image and raw material online image; A data processing subunit (212) is used to clean and filter the raw material purchase chain data and the raw material inventory data through a preset solution correction model to obtain abnormal inventory data; An image verification subunit (213) is used to identify the raw material image data based on the raw material inventory data through the solution correction model, verify the extracted abnormal inventory data based on the recognition result, and determine whether there is a new abnormal representation. If so, an alarm message is sent to the management personnel and image abnormality data is generated; The data packaging subunit (214) is used to package the abnormal inventory data and the image abnormal data to generate raw material abnormal data.

5. A lipstick quality inspection system according to claim 4, characterized in that: The solution matching module (1) comprises: A demand acquisition unit (11), used to acquire demand information uploaded by a user, wherein the demand information includes regular demand information and remark information; A data processing unit (12) is used to clean the demand information and extract the quality inspection requirements of the target lipstick; A solution matching unit (13) is used to input the quality inspection requirements into a preset solution matching model for intelligent matching to generate a preliminary quality inspection solution; the solution matching model is a machine learning model obtained by iterative training based on historical quality inspection data.

6. A lipstick quality inspection system according to claim 4, characterized in that: Also includes: The negative feedback module (5) is used to generate monitoring and early warning information based on the quality inspection report through a preset negative feedback model and send it to the management personnel. The monitoring and early warning information includes production equipment adjustment suggestions, raw material storage suggestions, raw material data mismatch prompts and production data mismatch prompts.

Images