SPC quality management method and device based on model driving and storage medium

Abstract

The invention provides an SPC quality management method and device based on model driving and a storage medium. The method comprises the steps that data collected in all stages in the production process is stored in a document type database; starting an SPC quality management service by adopting a triggering mechanism; extracting data from the document database, and cleaning the extracted data based on a pre-configured conversion model to generate a data analysis model; on the basis of the data analysis model and a pre-configured point location algorithm model, point location models of multiple SPC control charts are generated; and calling a pre-configured early warning model to carry out early warning monitoring on point locations in the point location models of the multiple SPC control charts, and generating an abnormality judgment report when abnormal conditions are monitored. Large-scale data can be efficiently processed, and the system performance and the data processing efficiency of the server can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to an SPC quality management method and device based on model driving and a storage medium. BACKGROUND

[0002] Statistical Process Control (SPC) is a process control tool with the aid of mathematical statistics method. SPC analyzes the data collected at each stage of the production process by applying statistical techniques, and adjusts the process to achieve the goal of improvement and quality assurance.

[0003] In the production process, related equipment will collect millions of data every day. Before further processing of these data, it is necessary to clean them to filter out meaningful data. Then, different algorithms are used to analyze these data to generate corresponding point data, and finally draw control charts.

[0004] However, due to the large amount of data, the server is under great pressure, resulting in program lag and delay, and thus cannot timely issue early warning, nor can it accurately compare and analyze historical data.

[0005] In summary, SPC faces an important challenge, that is, how to efficiently process large-scale data to improve the system performance and data processing efficiency of the server. SUMMARY

[0006] The embodiments of the present application provide an SPC quality management method and device based on model driving and a storage medium to solve the problems in the related art, and the technical solutions are as follows: In a first aspect, the embodiments of the present application provide an SPC quality management method based on model driving, comprising: storing the data collected at each stage of the production process to a document type database; starting an SPC quality management service by using a trigger mechanism; extracting data from the document type database, cleaning the extracted data based on a pre-configured conversion model, and generating a data analysis model; generating point models of multiple SPC control charts based on the data analysis model and a pre-configured point algorithm model; calling a pre-configured early warning model to monitor points in the point models of the multiple SPC control charts, and generating a difference report when an abnormal condition is monitored.

[0007] In an embodiment, starting the SPC quality management service by using the trigger mechanism comprises: The SPC quality management service is started by using a timing task or a message queue trigger.

[0008] In an embodiment, data is extracted from the document type database, the extracted data is cleaned based on a pre-configured conversion model, and a data analysis model is generated including: The extracted data is converted into a corresponding memory model according to the data type based on the conversion model; The different types of memory models are adapted to the corresponding processors, and the memory models adapted by the corresponding processors are customized and cleaned; Based on the cleaned different types of memory models, the data analysis model is generated.

[0009] In an embodiment, based on the data analysis model and a pre-configured point algorithm model, a point model of a plurality of SPC control charts is generated including: A point conversion task is generated, and the point conversion task is put into a message queue; When the message queue is consumed, based on the data analysis model and the point algorithm model, the point model of the plurality of SPC control charts is generated.

[0010] In an embodiment, when the message queue is consumed, based on the data analysis model and the point algorithm model, the point model of the plurality of SPC control charts is generated including: When the message queue is consumed, data in the memory model of the data analysis model is fed into the point algorithm model, and statistical point information and related auxiliary information in the memory model of the data analysis model are extracted by the point algorithm model; Based on the statistical point information and related auxiliary information, the point model of the plurality of SPC control charts is generated.

[0011] In an embodiment, the method further includes: The data analysis model is stored in a relational database; and / or, The point model of the plurality of SPC control charts is saved in a relational database.

[0012] In an embodiment, the method further includes: In response to a user's viewing operation on any one SPC control chart, a pre-configured chart model is called to generate a corresponding curve in the any one SPC control chart based on the point model of the any one SPC control chart.

[0013] In a second aspect, the embodiments of the present application further provide a model-driven SPC quality management device, including: a storage unit, configured to store data collected in each stage of a production process into a document type database; a starting unit, configured to start an SPC quality management service by using a trigger mechanism; a first generating unit, configured to extract data from the document type database, clean the extracted data based on a pre-configured conversion model, and generate a data analysis model; a second generating unit, configured to generate a point position model of a plurality of SPC control charts based on the data analysis model and a pre-configured point position algorithm model; a monitoring unit, configured to call a pre-configured early warning model to perform early warning monitoring on point positions in the point position model of the plurality of SPC control charts, and generate a difference report when an abnormal situation is monitored.

[0014] In an implementation, the starting unit, when starting the SPC quality management service by using the trigger mechanism, is specifically configured to: start the SPC quality management service by using a timing task or a message queue.

[0015] In an implementation, the first generating unit, when extracting data from the document type database, cleaning the extracted data based on the pre-configured conversion model, and generating the data analysis model, is specifically configured to: convert the extracted data into a corresponding memory model according to a data type based on the conversion model; adapt different types of memory models to corresponding processors, and customize cleaning of the memory models adapted by the corresponding processors; generate the data analysis model based on the cleaned different types of memory models.

[0016] In an implementation, the second generating unit, when generating the point position model of the plurality of SPC control charts based on the data analysis model and the pre-configured point position algorithm model, is specifically configured to: generate a point position conversion task, and put the point position conversion task into a message queue; when the message queue is consumed, generate the point position model of the plurality of SPC control charts based on the data analysis model and the point position algorithm model.

[0017] In an implementation, the second generating unit, when generating the point position model of the plurality of SPC control charts based on the data analysis model and the point position algorithm model when the message queue is consumed, is specifically configured to: When the message queue is consumed, data in the memory model of the data analysis model is fed into the point algorithm model, and statistical point information and its related auxiliary information in the memory model of the data analysis model are extracted by the point algorithm model; Based on the statistical point information and its related auxiliary information, a point model of the plurality of SPC control charts is generated.

[0018] In an implementation, the storage unit is further configured to: store the data analysis model into a relational database; and / or, save the point model of the plurality of SPC control charts into the relational database.

[0019] In an implementation, the apparatus further comprises a third generation unit configured to: in response to a user's viewing operation on any one of the SPC control charts, invoke a pre-configured chart model to generate a corresponding curve in the any one of the SPC control charts based on the point model of the any one of the SPC control charts.

[0020] In a third aspect, the embodiments of the present application further provide a computer apparatus, comprising a memory and a processor, the memory stores instructions, the instructions are loaded and executed by the processor to implement the method in any one of the embodiments of the above aspects, wherein the memory and the processor communicate with each other through an internal connection path.

[0021] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, the computer readable storage medium stores a computer program, when the computer program runs on a computer, the method in any one of the embodiments of the above aspects is implemented.

[0022] The advantages or beneficial effects of the above technical solutions at least include: (1) High elasticity of data access expansion function: the present application stores the data collected in each stage of the production process into the document type database, so that the server can more efficiently store and read the data collected in each stage of the production process, has high elasticity of data access expansion function, supports the server to efficiently process large-scale data, thereby improving the system performance and data processing efficiency of the server; (2) Data processing peak clipping and decoupling: the present application starts the SPC quality management service by using the trigger mechanism, which can separate storage and calculation for processing, so that the server realizes data processing peak clipping and decoupling, can reduce the calculation pressure of the server, avoids the server from responding slowly or even crashing due to overload, thereby improving the system performance and data processing efficiency of the server; (3) Improve the efficiency and accuracy of data processing: the present application can generate a data analysis model by cleaning and processing the data extracted from the document type database based on the pre-configured conversion model, which can clean and convert the data, standardize the data, and model the data, reducing unnecessary data volume and fields, which can improve the efficiency of subsequent data analysis, improve the processing efficiency and accuracy of SPC quality management, and thus improve the system performance and data processing efficiency of the server; (4) Has visualization and interactivity: the present application can generate a variety of SPC control chart point position models based on the data analysis model and the pre-configured point position algorithm model, which can make the data analysis model visualized, facilitate the user to interactively view the related control chart, and provide better user experience and interactivity, further improving the system performance of the server.

[0023] (5) Improve response speed: the present application can generate a variety of SPC control chart point position models by calling the pre-configured early warning model to monitor the points in the point position model, and generate a difference report when an abnormal situation is detected, which can directly use the point position model of the SPC control chart to calculate when early warning is needed, without logical operation or data processing, which can greatly improve the response speed of the server warning, so that the abnormal situation can be detected in time and appropriate measures can be taken, further improving the system performance and data processing efficiency of the server.

[0024] The above summary is only for the purpose of the description and is not intended to limit in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the present application will be readily apparent to those skilled in the art by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0025] In the drawings, like reference numerals refer to same or similar functionalities throughout the several views. The drawings are not necessarily to scale. It is to be understood that the drawings only depict several embodiments in accordance with the disclosure and should not be interpreted to limit the scope of the disclosure.

[0026] Figure 1 A flowchart example of a model-driven SPC quality management method provided by an embodiment of the present application; Figure 2 A flowchart example of another model-driven SPC quality management method provided by an embodiment of the present application; Figure 3 An example of an Xbar control chart provided by an embodiment of the present application; Figure 4A sample diagram of a data query interface provided by an embodiment of the present application is shown in FIG. 1. Figure 5 A structural block diagram of an SPC quality management device based on model driving provided by an embodiment of the present application is shown in FIG. 2. Figure 6 A structural block diagram of a computer device provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0027] In the following, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than limiting.

[0028] Figure 1 A flow chart of an SPC quality management method based on model driving according to an embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, the method can include the following steps: Figure 1 S110, storing the data collected in each stage of the production process to a document type database.

[0029] The prior art is to store the data collected in each stage of the production process in a relational database, such as MySQL. When using a relational database to store data, the format of the data to be stored needs to be known in advance before storage. If not, the data can only be converted into a text type format and then stored in the relational database, resulting in the defects of slow storage and slow reading.

[0030] Based on this, an embodiment of the present application selects to use a document type database to store the data collected in each stage of the production process. In this way, no matter what device the data collected in each stage of the production process comes from and how the format of the data is, the data can be directly stored in the document type database without format conversion, and the writing and reading speed is very fast and efficient. At the same time, since the document type database has good scalability, it supports increasing server nodes through horizontal horizontal expansion, and can easily cope with the growth of data volume.

[0031] That is, in the embodiment of the present application, by performing step S110, the data collected in each stage of the production process can be more efficiently stored and read, and the data access expansion function has high flexibility, so as to improve the processing capacity and throughput of the server, support the server to efficiently process large-scale data, and improve the system performance and data processing efficiency of the server.

[0032] S120, starting the SPC quality management service by using a trigger mechanism.

[0033] ​Current technology involves directly processing data collected from various stages of the production process using relevant equipment without storage, resulting in significant computational burden and making historical analysis impossible.

[0034] Based on this, the data processing flow in this application embodiment has been optimized and improved. Instead of directly transmitting data from the relevant device to the view display, the data storage and computation are processed separately, that is, the data storage and computation are decoupled. It can be understood that step S110 is executed first to store the data, and then step S120 is executed to perform subsequent data computation.

[0035] In step S120, the SPC quality management service can be started using a scheduled task or a message queue. For example, the SPC quality management service can be started by distributing computing tasks through a scheduled task or a message queue.

[0036] In this embodiment of the application, by executing step S120, data computation and data storage decoupling can be achieved, as well as peak shaving of data processing volume can be achieved, which can reduce the computing pressure on the server and prevent the server from slowing down or even crashing due to overload, thereby improving the system performance and data processing efficiency of the server.

[0037] S130. Extract data from the document-based database, clean the extracted data based on the pre-configured transformation model, and generate a data analysis model.

[0038] In one implementation, each time the SPC quality management service is started, the data needs to be cleaned first, i.e., step S130 is executed.

[0039] In one implementation, the process of step S130 may include the following sub-steps: S131. Based on this conversion model, the extracted data is converted into the corresponding memory model according to the data type.

[0040] That is, in step S131, each data type corresponds to a memory model.

[0041] As an example, this transformation model can divide the extracted data into two main data types based on data type: count data and measure data.

[0042] For measurement data, this transformation model numbers the measurement data in groups based on data continuity, forming different valid data subgroups. For example, the transformation model can segment the measurement data according to different pre-configured segmentation strategies such as taking the start data, taking the end data, taking all data, and taking random data, resulting in multiple subgroups. Then, valid data is judged based on a minimum of K subgroups. During this process, outrageous data caused by equipment malfunction or human error can be identified as invalid data, while the remaining data is identified as valid data. The valid data is then grouped into a valid data subgroup. At the same time, for data generated due to special reasons, a pre-set threshold can be used to determine whether the data exceeds the limit. If so, the data is discarded and not included in the valid data subgroup.

[0043] K can be set according to actual needs. For example, since process capability indices (such as Cp and Cpk) depend on the accurate estimation of process mean and standard deviation, K can be set to ≥25 to provide sufficient data points, make the standard deviation estimation more robust, and reduce sampling error. In this way, it can be ensured that the subsequent control limits will not change significantly due to a small amount of data fluctuation.

[0044] For count data, this transformation model randomly assigns numbers to groups, forming different valid data subgroups. During this process, it can determine whether the count data is acceptable or unacceptable, and the number of defects. Simultaneously, data affected by special reasons is discarded. For example, for inspecting the appearance defects of parts on an assembly line (count data), 10 parts are randomly inspected each shift, and the results are recorded as acceptable (0 defects) / unacceptable (≥1 defect). If, on a certain day, a part is found to have multiple scratches due to a sudden equipment malfunction (a special reason), the data for that part is discarded (discarded), and the remaining data forms a valid data subgroup. Subsequently, only normal data is used to plot a P-chart to analyze process stability. As another example, for detecting the number of scratches on the surface of bearings (discrete count data), 100 bearings are numbered. If a bearing with a certain number is found to be scratched due to operational error (a special reason), the data for that number is discarded, and the data for the remaining 99 bearing numbers are used to form 99 valid data subgroups. These can then be used to calculate the defect rate and plot a u-control chart.

[0045] As an example, this transformation model can perform corresponding transformations (including field name and type conversions) on each valid data subgroup to obtain the corresponding in-memory model. For instance, data generated from equipment temperature control testing may contain 30 fields (such as material model, temperature, color, etc.), but only 10 fields (such as material code, temperature, etc.) may actually be used for analysis. Based on this, for each valid data subgroup, the transformation model can assign the material model to the material code field, directly assign the temperature field, and ignore the color field, which does not require transformation.

[0046] In this example, the two data types, count data and measure data, can each correspond to multiple memory models.

[0047] It should be understood that, in the embodiments of this application, the transformation model is a predefined "transformation rule set", which can be understood as a rule engine used to drive data from "original form" to "target form". It is used to solve the semantic, type and redundancy differences between the original data and the target requirements through predefined field mapping, type conversion, filtering and other rules, so as to ensure that the data is usable and efficient in subsequent processing (such as analysis, modeling and system interaction). That is, the transformation model can reduce the complexity of data transformation and improve the efficiency and consistency of data processing through configurable rules.

[0048] It should be understood that, in the embodiments of this application, the memory model is a bridge rule connecting external storage and program logic. It can transform "static, heterogeneous external data" into "dynamic, program-adaptive internal data structure" through parsing, reconstruction and optimization, ultimately achieving efficient processing and logical decoupling.

[0049] In this embodiment of the application, by executing step S131, modeling can be performed according to the pre-configured transformation model, reducing unnecessary data volume and fields, and facilitating the improvement of the efficiency of subsequent analysis.

[0050] S132. Adapt different types of memory models to the corresponding processors, and have the corresponding processors perform customized cleaning on the adapted memory models.

[0051] That is, in step S132, each memory model is adapted to one or more processors.

[0052] As an example, a processor can remove unnecessary data from its adapted memory model and tag specific attributes, such as encoding, so that it can be used for fast searching or specific transformations, etc.

[0053] S133. Generate a data analysis model based on the different types of cleaned memory models.

[0054] This can be understood as the data analysis model being composed of different types of cleaned memory models.

[0055] As an example, a data analysis model can consist of data sources, rules, and definitions: (1) Data source Collected data from different groups is grouped into a single data source. Specifically, for data collected on tests such as temperature and drop resistance of mobile phones and computers, the same data source, i.e., the same view table, is used for differentiation through coding. For data collected on production line efficiency, such as the number of products produced per hour and yield rate per day, a separate view is generated, also differentiated through coding. For the same type of view, different parameters are used to differentiate between different views.

[0056] (2) Rules In this example, the rule can be a control chart rule.

[0057] For example, the control chart rules may include: 1) Sampling configuration: sampling time unit, method, number of subgroups, etc.; 2) Outlier detection rules: consecutive k-point increments, distance from the median line greater than k standard deviations, consecutive k-points alternating vertically, etc.; 3) Others: number of histogram groups, estimated standard deviation.

[0058] (3) Definition In this example, "definition" can refer to a control chart definition. The process of defining a control chart can include the following steps: Step 1: Select control chart rules; Step 2: Select the data source; Step 3: Select the analysis dimensions: sample fields, unqualified fields, etc. Step 4: Define custom variables: control the upper and lower limits, and define the mean and standard deviation; Step 5: Data Processing: Customize Data Precision; Step 6: Select the CPK level; Step 7: Inspect the platform: Record the personnel on the platform and the early warning notification method; Step 8: Select the stratification criteria, which are the filtering criteria for running the control chart; Step 9: Select the grouping dimension, such as grouping by time or by inspector. The specific grouping and sampling method can be set in the sampling configuration of the rule.

[0059] In this embodiment of the application, by performing step S130, the data can be standardized, reducing unnecessary data volume and fields, thereby improving the processing efficiency and accuracy of SPC quality management.

[0060] In one applicable scenario provided in the embodiments of this application, combined with Figure 1 and Figure 2 As shown in the embodiments of this application, the model-driven SPC quality management method may further include the following steps: S130a. Store the data analysis model in a relational database.

[0061] In one implementation, the relational database can be MySQL.

[0062] Since the data initially collected (i.e. the data collected in step S110) has been transformed into a data analysis model through step S130, the data is very concise and has become stable and reliable. The field names are consistent, and there will be no excessive complexity or missing data. At this point, storing the data analysis model in a relational database is suitable for various statistics and calculations. For example, MySQL can be used with indexes, or queries can be performed on related tables. In this way, the data can be integrated and queried more quickly and in a more diverse manner.

[0063] In this embodiment of the application, by executing step S130a, the powerful query function of relational databases can be utilized to perform more complex and flexible data analysis. For example, since relational databases support rich query languages ​​and aggregation operations, storing data analysis models in relational databases allows for multi-table association, data filtering, and aggregation calculations to meet different analysis needs. This can further improve the processing efficiency of SPC quality management and expand the analysis dimensions of SPC quality management.

[0064] It should be noted that the embodiments of this application do not limit the execution order between steps S130a and S140. That is, steps S130a and S140 can be executed simultaneously, or steps S130a can be executed first and then steps S140, or steps S140 can be executed first and then steps S130a.

[0065] S140, a point model that generates various SPC control diagrams based on a data analysis model and a pre-configured point algorithm model.

[0066] In one implementation, the process of step S140 may include the following sub-steps: S141. Generate a point conversion task and put the point conversion task into the message queue.

[0067] In practice, after generating the data analysis model, the point conversion task can be automatically generated and placed in the message queue for subsequent automatic point conversion processing.

[0068] S142. When the message queue is consumed, a point model for various SPC control diagrams is generated based on the data analysis model and the point algorithm model.

[0069] That is, when the message queue is consumed, the execution process of generating point models of various SPC control diagrams is automatically triggered.

[0070] In practical implementation, when the message queue is consumed, the data from the memory model of the data analysis model can be fed into the point algorithm model, and the point algorithm model can extract the statistical point information and related auxiliary information from the memory model.

[0071] As an example, the point transformation task includes generating a mean-range control chart (i.e., Taking the point model of the control chart as an example, when the message queue is consumed, each valid data subgroup obtained by data partitioning in the memory model can be fed into the point algorithm model. The point algorithm model can then follow the mean chart (i.e., The graph generates and outputs statistical point information (such as mean point) and related auxiliary information (such as 6 sigma line and control limits) for each valid data subgroup, and can generate and output statistical point information (such as range point) and related auxiliary information (such as control limits) for each valid data subgroup according to the range graph (R graph).

[0072] For example, after the point-based algorithm model obtains each valid data subgroup, the mean of each valid data subgroup can be calculated according to the following formula (1). And calculate the range point R for each valid data subgroup according to the following formula (2). At this time, the mean of each valid data subgroup can be calculated. The range point R is stored as a data point value in a database (such as a relational database).

[0073] (1) (2) Where n represents the total number of data points in each valid data subgroup, x1, x2, ..., x... n For each valid data subgroup, X max X is the maximum value of the data in each valid data subset. min The minimum value of data in each valid data subgroup.

[0074] For example, in this example, combining Figure 3 As shown, taking the mean data point (i.e., the data point corresponding to number 7434) as an example, the process of the point position algorithm model generating and outputting the mean point and its related control limits for a certain valid data subgroup can be as follows: 1) Extract data from a valid data subgroup based on the sampling frequency. For example, each time 3 data points need to be extracted (1.73, 4.55, 5.34). 2) Calculate the mean among the extracted data (e.g., Figure 3 (mean data points in the data) = (1.73 + 4.55 + 5.34) / 3 = 3.87 (This mean can be stored in the database as a point value). 3) After the mean is calculated, the point algorithm model can automatically plot a point on the graph and then connect them with a line to form a trend chart; 4) Calculate the range point R 7434 =X max -X min =5.34-1.73=3.61 (This range point can be stored in the database as a point value). 5) Figure 3 The data contains 25 means, numbered 7419-7434. The average of the ranges of these 25 means is calculated. =5.21; 6) Centerline calculation: As time progresses, the number of data points on the graph gradually increases according to the sampling frequency. At this point, the sum of the data from each mean is taken as the average to form the centerline. N is the total number of mean values. , … The mean: There are a total of 25 means, numbered 7419-7434. Calculate the mean of these 25 means. =6.65, which is the center line CL of the mean control chart. =6.65; 7) Calculation of upper and lower control limits (see...) Figure 3 (Red line in the middle) Upper control limit UCL of the mean control chart = +A2 =6.65 + A2 * 5.21 = 11.98; Lower control limit LCL of the mean control chart = -A2 =6.65 - A2 * 5.21 = 1.31; 8) Calculation of 6 Sigma line: ; ; ; ; ; In processes 7) and 8), A2 and d2 are constant values, which can be obtained by looking up a table. After the calculation is completed, Figure 3 The upper and lower control limits are drawn in the middle; Sigma stands for standard deviation and is used to measure the degree of data fluctuation.

[0075] 9) Finally, the point algorithm model extracts the mean point and its related auxiliary information (such as upper and lower control limits, center line).

[0076] In the example above, the calculation process by which the point location algorithm model generates and outputs statistical point location information (such as range points) and related auxiliary information (such as control limits) for each valid data subgroup may include: 1) Centerline calculation: The center line CL of the range control chart R = =5.21; 2) Calculation of upper and lower control limits: Upper control limit (UCL) of the range control chart R =D4 ; Lower control limit (LCL) of the range control chart R =max(0, D3) ); D4 and D3 are constant values, which can be obtained by looking up a table.

[0077] As another example, taking the point conversion task as an example, which includes a non-conforming rate control chart (i.e., a P control chart), when the message queue is consumed, each valid data subgroup obtained by data segmentation in the memory model can be fed into the point algorithm model. The point algorithm model can extract the statistical point information (such as non-conforming rate points) and its related auxiliary information (such as control limits) in each valid data subgroup.

[0078] For example, in this example, taking the nonconforming rate as an example, the process by which the point-based algorithm model generates and outputs the nonconforming rate points and their related control limits for a certain valid data subgroup can be as follows: 1) Data Acquisition: After the point-based algorithm model obtains the effective data subgroup, it records the total amount of data in the effective data subgroup (i.e., the sample size n per batch). i ) and the total number of non-conforming data (i.e., the number of non-conforming items); 2) Calculate the proportion: Calculate the non-conforming rate point P for each batch. i =Number of defective products / n i (These non-compliance rate points can be stored as point values ​​in the database.) 3) Centerline calculation: Centerline CL= k is the total number of batches; 4) Calculation of upper and lower control limits: Upper control limit UCL = CL + 3 ; Lower control limit LCL=CL-3 (LCL must not be negative); 5) Finally, the point-based algorithm model extracts the non-compliance rate points and their related auxiliary information (such as upper and lower control limits, center line).

[0079] In practice, various point models for SPC control diagrams can be generated based on the statistical point information and its related auxiliary information.

[0080] For example, a point model for a mean-range control chart can be generated based on the mean point and its related auxiliary information. Similarly, a point model for a nonconforming rate control chart can be generated based on the nonconforming rate point and its related auxiliary information.

[0081] For example, since equipment temperature control and inspection scenarios involve high costs per unit of production or inspection, the production process can be monitored using individual value data and moving range. In this case, the four types of measurement-based control charts in SPC8 can be used for monitoring and analysis of the production process. Specifically, these four types of measurement-based control charts can be generated in step S142. These four types of measurement-based control charts can be: a mean-range control chart (i.e.,...) Control charts), mean-standard deviation control chart (i.e., X̄-S control chart), individual-moving range control chart (i.e., I-MR control chart), and median-range control chart (i.e., (Control chart).

[0082] For example, in a scenario involving monitoring the non-conforming rate of parts on an assembly line (for detecting appearance defects), four types of count-based control charts in SPC8 can be used for monitoring and analysis of the production process. Specifically, these four types of count-based control charts can be generated in step S142. These four types of count-based control charts can include: a non-conforming rate control chart (i.e., p control chart), a non-conforming number control chart (i.e., np control chart), a defect number control chart (i.e., c control chart), and a unit defect number control chart (i.e., u control chart).

[0083] It should be understood that, in the embodiments of this application, the point location algorithm model is a set of standardized calculation rules or logic used to guide how to extract specific "points" (i.e., statistical indicators) and related auxiliary information (such as control limits, 6 sigma lines, etc.) from given data. It can be understood that the point location algorithm model is the core algorithmic support for generating the point location model.

[0084] It should be understood that, in the embodiments of this application, the point model is a specific instantiation of the point algorithm model, which combines data processing requirements and point algorithms to form a "computational framework" that can be directly applied to actual data. It can be understood that the point model is an intermediate carrier connecting the data analysis model and the generation of specific points.

[0085] In this embodiment of the application, by executing step S140, the data analysis model can be visualized, which facilitates users to interactively view the relevant control charts and provides a better user experience and interactivity.

[0086] In another applicable scenario provided in the embodiments of this application, combined with Figure 1 and Figure 2 As shown in the embodiments of this application, the model-driven SPC quality management method may further include the following steps: S140a. Save the point models of various SPC control charts to a relational database.

[0087] In practical implementation, by storing the point models of various SPC control charts in a relational database, it becomes suitable for various statistics and calculations. For example, MySQL can be indexed, or queries of related tables can be performed. In this way, data can be integrated and queried more quickly and in a more diverse manner.

[0088] In this embodiment of the application, by executing step S140a, operations such as multi-table association, data filtering and aggregation calculation can be performed to meet different analysis needs, thereby further improving the processing efficiency of SPC quality management and expanding the analysis dimensions of SPC quality management.

[0089] It should be noted that the embodiments of this application do not limit the execution order between steps S140a and S150. That is, steps S140a and S150 can be executed simultaneously, or steps S140a can be executed first and then steps S150, or steps S150 can be executed first and then steps S140a.

[0090] S150: Call the pre-configured early warning model to perform early warning monitoring on the points in the point models of various SPC control charts, and generate an anomaly report when an anomaly is detected.

[0091] In one implementation, an early warning model can be periodically invoked to monitor points in the point model of any SPC control chart, so as to generate an anomaly report when an anomaly is detected.

[0092] For example, the early warning model can simultaneously assess the stability of the mean control chart and the range control chart of the mean-range control chart, that is, whether the mean and the range fluctuate within the upper and lower control limits. If the points on the mean control chart or the range control chart are detected to exceed the corresponding upper and lower control limits, it is determined that there is an anomaly, and an anomaly report can be generated.

[0093] For example, the early warning model can support the following two anomaly detection methods: The first type Users can customize warning thresholds in the warning model, such as custom specification limits, control limits, CPK, etc. When a location exceeds the custom warning threshold, the warning model will detect the anomaly. The second type The early warning model determines anomalies and issues warnings based on the probability of a point falling within the upper or lower control limits and the probability of a trend occurring between consecutive points. For example, the probability of a point falling outside one control limit is (1-99.73%) / 2=0.135%, which the early warning model can basically consider a low-probability event. If the event occurs, it is likely due to a special cause. The probability of 9 consecutive points on the same side of the center line is P(N=9) =2*(0.9973 / 2)^9 = 0.38%, which the early warning model can consider a low-probability event, and the system will issue an early warning.

[0094] In the second method of anomaly detection, the early warning model will learn autonomously in the data analysis model based on the characteristics of materials and projects, and will also issue anomaly warnings when the calculated probability is a low-probability event.

[0095] It should be noted that the probabilities involved in the second method of anomaly detection can be set according to the national standard by default, or they can be adjusted by the user. Users can adjust the probabilities involved for each control chart.

[0096] In one implementation, when an anomaly is detected, an alert can be automatically sent via notification channels such as email or WeChat messages, along with a generated anomaly report.

[0097] In this embodiment of the application, by executing step S150, when a warning judgment is required, the point models of these various SPC control charts can be directly used for calculation without performing logical operations or data processing. This can greatly improve the response speed of the server warning, so that abnormal situations can be detected in a timely manner and corresponding measures can be taken.

[0098] In another applicable scenario provided in the embodiments of this application, combined with Figure 1 and Figure 2 As shown in the embodiments of this application, the model-driven SPC quality management method may further include the following steps: S160. In response to the user's viewing operation for any type of SPC control chart, call the pre-configured chart model based on the point model of any type of SPC control chart to generate the corresponding curve in any type of SPC control chart.

[0099] In one implementation, the process of step S160 may include the following sub-steps: S161. In response to the viewing operation, call the chart model to render the statistical point information in the point model of any SPC control chart, and draw the statistical point polyline in any SPC control chart.

[0100] In practical implementation, a data query interface can be provided, which can be as follows: Figure 4 As shown in the image, on this data query interface, users can select data query criteria, such as inspection station, control chart type, material or equipment code, and inspection items. After selecting the data query criteria, users can click the "Analyze" button on the data query interface to generate the viewing operation.

[0101] In practice, by responding to the viewing operation, the chart model can be automatically invoked to render the statistical point information in the point model of any SPC control chart, and draw the statistical point polyline in any SPC control chart.

[0102] As an example, taking any SPC control chart as a mean-range control chart, the chart model can render the mean and range points in the point model of the mean-range control chart, and draw the mean line in the mean-range control chart (e.g., Figure 4 The blue line above) and the range point line (such as...) Figure 4 (The blue broken line below).

[0103] S162. Call the chart model and combine it with the data analysis model and the relevant auxiliary information in the point model of any SPC control chart to render the relevant auxiliary curves corresponding to the statistical point polyline in any SPC control chart.

[0104] In practical implementation, the chart model can be rendered by combining the effective data subgroups in the data analysis model and the relevant auxiliary information in the point model of any SPC control chart, and the relevant auxiliary curves corresponding to the statistical point polylines can be drawn in any SPC control chart.

[0105] As an example, taking the mean-range control chart as an example, the chart model can be rendered by combining the effective data subgroups in the data analysis model and the relevant auxiliary information in the point model of the mean-range control chart. This will draw the relevant auxiliary curves corresponding to the statistical point line in the SPC control chart, such as the 6-sigma line, upper and lower control limits, and center line, ultimately forming... Figure 4 Mean-range control chart in the data.

[0106] It should be understood that, in the embodiments of this application, the chart model can be regarded as a data visualization engine, which can serve as an intermediate layer connecting the point model and the user interface, and is responsible for converting the abstract point model into a visual curve.

[0107] In this embodiment of the application, by executing step S160, users can easily view and understand the data analysis results through intuitive charts and visualization interfaces, and quickly discover problems and trends.

[0108] In summary, the model-driven SPC quality management method provided in this application has the following beneficial effects: (1) Highly flexible data access expansion function: This application stores the data collected at each stage of the production process into a document-type database, enabling the server to store and read the data collected at each stage of the production process more efficiently. It has a highly flexible data access expansion function, supports the server to efficiently process large-scale data, thereby improving the system performance and data processing efficiency of the server. (2) Peak shaving and decoupling of data processing: This application starts the SPC quality management service by using a trigger mechanism, which can separate storage and computing, so that the server can realize peak shaving and decoupling of data processing, which can reduce the computing pressure on the server and avoid the server from slowing down or even crashing due to overload, thereby improving the system performance and data processing efficiency of the server. (3) Improve the efficiency and accuracy of data processing: This application cleans and processes the data extracted from the document database based on a pre-configured transformation model to generate a data analysis model. It can clean and transform the data, standardize the data, and model the data, reducing unnecessary data volume and fields, which helps to improve the efficiency of subsequent data analysis and improve the processing efficiency and accuracy of SPC quality management, thereby improving the system performance and data processing efficiency of the server. (4) Visualization and interactivity: This application generates point models of various SPC control diagrams based on data analysis models and pre-configured point algorithm models, which enables the data analysis models to be visualized, facilitates users to interactively view relevant control diagrams, provides a better user experience and interactivity, and further improves the system performance of the server.

[0109] (5) Improved response speed: This application uses a pre-configured early warning model to monitor the points in the point models of various SPC control charts and generates an anomaly report when an anomaly is detected. When an early warning judgment is needed, the point models of these various SPC control charts can be used directly for calculation without logical operations or data processing. This can greatly improve the response speed of the server's early warning, so that anomalies can be detected in time and corresponding measures can be taken, further improving the system performance and data processing efficiency of the server.

[0110] Furthermore, the model-driven SPC quality management method provided in this application also has the following beneficial effects: (6) Scalable analytical capabilities: By storing the data analysis model in a relational database and / or storing the point models of various SPC control charts in a relational database, this application can utilize the powerful query function of the relational database to perform more complex and flexible data analysis, and can support operations such as multi-table association, data filtering and aggregation calculation to meet different analytical needs, thereby further improving the processing efficiency of SPC quality management and expanding the analytical dimensions of SPC quality management; (7) Problems and trends can be quickly identified: This application responds to the user's viewing operation for any SPC control chart by calling the pre-configured chart model based on the point model of any SPC control chart to generate the corresponding curve in any SPC control chart. This allows users to easily view and understand the data analysis results through intuitive charts and visualization interfaces, and can quickly identify problems and trends.

[0111] Figure 5 A structural block diagram of a model-driven SPC quality management device according to an embodiment of this application is shown. Figure 5 As shown, the device may include: Storage unit 210 is used to store data collected at each stage of the production process into a document-type database; The startup unit 220 is used to start the SPC quality management service using a triggering mechanism. The first generation unit 230 is used to extract data from a document-type database, clean the extracted data based on a pre-configured transformation model, and generate a data analysis model. The second generation unit 240 is used to generate point models for various SPC control diagrams based on the data analysis model and the pre-configured point algorithm model. The monitoring unit 250 is used to call the pre-configured early warning model to perform early warning monitoring on the points in the point model of various SPC control charts, and generate an anomaly report when an anomaly is detected.

[0112] In one implementation, when the startup unit 220 is used to start the SPC quality management service using a triggering mechanism, it is specifically used for: The SPC quality management service can be started using a scheduled task or message queue.

[0113] In one implementation, the first generation unit 230, when extracting data from a document-type database, cleaning the extracted data based on a pre-configured transformation model, and generating a data analysis model, specifically performs the following: Based on the transformation model, the extracted data is converted into the corresponding memory model according to the data type; Different types of memory models are adapted to the corresponding processors, and the corresponding processors perform customized cleaning on the adapted memory models. Data analysis models are generated based on the different types of cleaned memory models.

[0114] In one implementation, when the second generation unit 240 generates point models for various SPC control maps based on a data analysis model and a pre-configured point algorithm model, it is specifically used for: Generate a point conversion task and put the point conversion task into a message queue; When the message queue is consumed, point models of various SPC control diagrams are generated based on the data analysis model and the point algorithm model.

[0115] In one implementation, the second generation unit 240, when generating point models for various SPC control charts based on a data analysis model and a point algorithm model when the message queue is consumed, specifically performs the following: When the message queue is consumed, the data in the memory model of the data analysis model is fed into the point algorithm model, and the point algorithm model extracts the statistical point information and related auxiliary information from the memory model of the data analysis model. Based on statistical point information and related auxiliary information, point models for various SPC control diagrams are generated.

[0116] In one embodiment, the storage unit 210 is further configured to: Store data analysis models in a relational database; and / or, Save the point models of various SPC control charts to a relational database.

[0117] In one embodiment, the apparatus further includes a third generating unit 260, the third generating unit 260 being configured to: In response to a user's viewing operation for any type of SPC control chart, the system calls a pre-configured chart model based on the point model of any type of SPC control chart to generate the corresponding curve in that SPC control chart.

[0118] The functions of each unit in the model-driven SPC quality management device of this application embodiment can be found in the corresponding description in the above method, and will not be repeated here.

[0119] Figure 6 A structural block diagram of a computer device according to an embodiment of this application is shown. Figure 6As shown, the computer device includes a memory 310 and a processor 320. The memory 310 stores instructions, which are loaded and executed by the processor 320 to implement the model-driven SPC quality management method described in the above embodiments. The number of memories 310 and processors 320 can be one or more.

[0120] The computer device also includes: The communication interface 330 is used to communicate with external devices and perform data exchange and transmission.

[0121] If the memory 310, processor 320, and communication interface 330 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0122] Optionally, in a specific implementation, if the memory 310, processor 320 and communication interface 330 are integrated on a single chip, the memory 310, processor 320 and communication interface 330 can communicate with each other through an internal interface.

[0123] This application provides a computer-readable storage medium storing a computer program. When the computer program is run on a computer, it implements the method provided in this application.

[0124] This application also provides a chip, which includes a processor for calling and executing instructions stored in a memory, causing a communication device with the chip installed to perform the method provided in this application.

[0125] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.

[0126] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0127] Further, optionally, the aforementioned memory may include read-only memory and random access memory, and may also include non-volatile random access memory. The memory may be volatile or non-volatile, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. Many forms of RAM are available by way of example, but not limitation. Examples include static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0128] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0129] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0130] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0131] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. Furthermore, the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functionality involved.

[0132] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0133] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. All or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware, the program being stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiments.

[0134] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.

[0135] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A model-driven SPC quality management method, characterized in that, include: Data collected at each stage of the production process is stored in a document-based database; The SPC quality management service is started using a trigger mechanism; Data is extracted from the document-based database, and the extracted data is cleaned and processed based on a pre-configured transformation model to generate a data analysis model. Based on the data analysis model and the pre-configured point location algorithm model, point location models for various SPC control diagrams are generated. The pre-configured early warning model is invoked to monitor the points in the point models of the various SPC control diagrams, and an anomaly report is generated when an anomaly is detected.

2. The method according to claim 1, characterized in that, The SPC quality management service is started using a trigger mechanism, including: The SPC quality management service is started using a scheduled task or message queue.

3. The method according to claim 1, characterized in that, Data is extracted from the document-based database, and the extracted data is cleaned and processed based on a pre-configured transformation model to generate a data analysis model, including: Based on the aforementioned conversion model, the extracted data is converted into a corresponding memory model according to the data type. Different types of memory models are adapted to the corresponding processors, and the corresponding processors perform customized cleaning on the adapted memory models. The data analysis model is generated based on the different types of cleaned memory models.

4. The method according to claim 3, characterized in that, Based on the data analysis model and the pre-configured point location algorithm model, various point location models for generating SPC control maps include: Generate a point conversion task and put the point conversion task into a message queue; When the message queue is consumed, the point model of the various SPC control diagrams is generated based on the data analysis model and the point algorithm model.

5. The method according to claim 4, characterized in that, When the message queue is consumed, the point model for generating the various SPC control diagrams based on the data analysis model and the point algorithm model includes: When the message queue is consumed, the data in the memory model of the data analysis model is fed into the point algorithm model, and the point algorithm model extracts the statistical point information and related auxiliary information from the memory model of the data analysis model. Based on the statistical point information and its related auxiliary information, point models of the various SPC control diagrams are generated.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Store the data analysis model in a relational database; and / or, The point models of the various SPC control charts are saved to a relational database.

7. The method according to any one of claims 1-5, characterized in that, The method further includes: In response to a user's viewing operation for any type of SPC control chart, a pre-configured chart model is invoked based on the point model of the SPC control chart to generate the corresponding curve in the SPC control chart.

8. A model-driven SPC quality management device, characterized in that, include: Storage unit, used to store data collected at each stage of the production process into a document-based database; The startup unit is used to start the SPC quality management service using a triggering mechanism. The first generation unit is used to extract data from the document-type database, clean the extracted data based on a pre-configured transformation model, and generate a data analysis model. The second generation unit is used to generate point models for various SPC control diagrams based on the data analysis model and the pre-configured point algorithm model. The monitoring unit is used to call the pre-configured early warning model to perform early warning monitoring on the points in the point model of the various SPC control diagrams, and generate an anomaly report when an anomaly is detected.

9. A computer device, characterized in that, include: A memory and a processor, wherein the memory stores instructions which are loaded and executed by the processor to implement the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when run on a computer, implements the method as described in any one of claims 1-7.

Images