Data model construction method

Abstract

The invention discloses a data model construction method, relates to the technical field of data model construction, and effectively improves supervision efficiency and accuracy in a product production process. The method comprises the following steps: acquiring production material information required by a product, and equipment operation data and product state data in a production process, establishing a real-time product image model according to the product state data, establishing a real-time equipment image model according to the equipment operation data, and establishing a real-time product image model according to the real-time equipment image model; establishing a production supervision model and generating a product supervision pair according to the plurality of product production records, inputting the real-time product image model, the real-time equipment image model and the production data information into the production supervision model, further matching with the product supervision pair, and generating a corresponding product abnormal production record and an equipment abnormal state record; and a product production record is generated after product production is completed, and the production supervision model is updated according to the newly generated product production record.

Description

Technical Field

[0001] The present invention relates to the technical field of data model construction, and in particular to a data model construction method. Background Art

[0002] Production line product supervision refers to the comprehensive monitoring and management of products on the production line to ensure that product quality meets the specified standards and requirements. This is an indispensable part of modern industrial manufacturing and is of great significance for improving product quality, reducing production costs and enhancing corporate competitiveness.

[0003] Existing production supervision technologies include the following defects: Visual inspection technology: Using computer vision and image processing technology, cameras or laser scanners are used to inspect and identify products. This technology can monitor product appearance, size, and shape in real time, but when it is used to monitor complex or sophisticated products, it is easily affected by lighting or environmental factors, resulting in inaccurate monitoring and slight product defects.

[0004] Data analysis and machine learning: By collecting a large amount of production data and applying machine learning algorithms to generate data models, potential quality problems and production anomalies can be identified. This method can help predict equipment failures, optimize production processes, and improve product quality, but it is not suitable for small production lines.

[0005] How to improve the monitoring efficiency and accuracy of the product production process while increasing the scope of application of the data model is a difficulty in the existing technology. For this purpose, a data model construction method is provided. Summary of the invention

[0006] In order to solve the above technical problems, the purpose of the present invention is to provide a data model construction method.

[0007] In order to achieve the above object, the present invention provides the following technical solutions: A data model construction method, the data model construction method comprising the following steps: S1. Collect the production data information required for the product, as well as the product status data of the product at each moment on each production line during the normal production process, and collect the equipment operation data of each equipment on the production line at each moment; S2, setting up a cloud computing platform, and then the cloud computing platform establishes a product image model according to a plurality of pre-stored product production records, and then establishes a product production supervision model, and at the same time establishes a real-time product image model according to the product status data generated in step S1; S3, the cloud computing platform establishes an equipment image model based on the product production record, and then establishes an equipment supervision model, and at the same time establishes a real-time equipment image model based on the equipment operation data generated in step S1; S4. Establish a production supervision model, wherein the production supervision model is provided with a data input layer, a production supervision layer, and a result output layer, and the product production supervision model and the equipment supervision model generated in step S2 and step S3 are input into the production supervision layer; S5. Overlap the product production supervision model and the equipment supervision model corresponding to the same production line, thereby generating a product supervision pair; S6. Input the real-time product image model, the real-time equipment image model and the production data information into the input layer, and then match the real-time product image model and the real-time equipment image model with the product supervision pair according to the production data information, and then generate the corresponding product abnormal production record and equipment abnormal status record, and generate the product production record after the product production is completed; S7. The cloud computing platform updates the production supervision model according to the newly generated product production records.

[0008] Furthermore, the process of collecting the product status data includes: Set numbers for each production line and each type of product, set up a cloud computing platform, and then collect the production data information required for the production of each type of product, and set numbers for the production data information according to the product numbers; The production data information includes the types and quantities of raw materials required for production, the production line numbers of each type of raw material for processing, and the order in which the product components generated by each production line are summarized; Set up exclusive sensors and infrared sensing devices for each production equipment on each production line; When the product starts to be produced, the raw materials are placed on the corresponding numbered production lines according to the required quantity according to the production data information; When the production raw materials are placed on the corresponding production line, the infrared sensing device of the first production equipment on the production line collects infrared image data of the production raw materials; According to the processing time of each production equipment for the production of raw materials, set k i processing time, where k i Indicates the number is S i The number of production equipment on the production line, i is a natural number greater than 0, and i is less than or equal to n; During the processing time, each infrared sensing device collects 20 to 50 sets of product infrared image data according to the length of the processing time of the product located in the production equipment and integrates them to generate product status data.

[0009] Furthermore, the process of collecting the equipment operation data includes: The various sensors and infrared sensing devices located on the production line collect state change data of the production equipment and equipment infrared image data, where the state change data includes temperature change curves, voltage change curves, etc. The state change data and equipment infrared image data are integrated to generate equipment operation data.

[0010] Furthermore, the process of establishing the real-time product image model and the real-time device image model includes: Whenever the cloud computing platform receives product status data, it generates a corresponding real-time product image model based on the product infrared image data in the product status data, and marks the corresponding product and production line number; Whenever the cloud computing platform receives equipment status data, it generates a corresponding real-time equipment image model based on the equipment infrared image data in the equipment status data, and marks the corresponding product and production line numbers.

[0011] Furthermore, the process of establishing the product production supervision model includes: The cloud computing platform is preset with Num product production records, which include historical product status data and historical equipment status data; Establish a corresponding product image model according to the product infrared image data, and overlap Num product image models at the same position of the same production line, where Num is a natural number greater than 50; Obtain the position range of each model line in the product image model, distribute the position range normally, select the central area relative to 50% of the entire position range as the normal position range of the product in the corresponding production line, and mark it on the corresponding product image model; The k product image models on the same production line are connected in sequence according to the corresponding processing time sequence, and then the product production supervision model of the corresponding product on each production line is obtained, and the corresponding product and production line numbers are marked. At the same time, the product production data information is bound to each product production supervision model.

[0012] Furthermore, the process of establishing the equipment supervision model includes: The cloud computing platform extracts historical equipment operation data from product production records, and extracts Num pieces of equipment infrared image data from the historical equipment operation data, and then generates equipment image models of each production equipment and the corresponding normal position range; Extract state change data from historical equipment operation data, and establish the same number of rectangular coordinate systems according to the number of types of state change data, and map Num state change data of the same production equipment on the same production line of the same type of product into the rectangular coordinate system; The state change data with the largest number of overlaps is taken as the standard state change data, and w time intervals are divided on the x-axis of the rectangular coordinate system, where w is a natural number greater than 10; Then, the calculus absolute value between each state change data and the standard state change data in each time interval is obtained; Set an abnormal threshold. If the absolute value of the integral in a time interval is greater than or equal to the abnormal threshold, it is determined that the corresponding state change data in the time interval has an abnormal change interval. Then, based on the judgment result, the abnormal change interval of the state change data is deleted or directly eliminated; Integrate all the retained state change data to obtain the normal state change data range of the production equipment of the corresponding production line when producing the corresponding product, and mark the corresponding product and production line numbers; According to the front and rear position relationship of each production equipment on the corresponding production line, each equipment image model is connected in series, and the corresponding normal state change data range interval is associated with the corresponding equipment image model, thereby obtaining the equipment supervision model.

[0013] Furthermore, the functions of each layer in the production supervision model include: The cloud computing platform inputs the product production supervision model and the equipment supervision model into the production supervision layer in the production supervision model; The data input layer is used to receive product production records, production data data, real-time product image models, equipment operation data and real-time equipment image models, and then generate corresponding model matching instructions; The production supervision layer is used to call the corresponding product production supervision model and equipment supervision model according to the model matching instruction, and then generate the corresponding product judgment result and equipment judgment result by judging the real-time equipment operation data and product status data; The result output layer is used to receive product judgment results and equipment judgment results from the production supervision layer, and then generate maintenance prompts and send them to corresponding maintenance personnel.

[0014] Furthermore, the process of generating the product supervision pair includes: When the cloud computing platform generates the real-time product image model and real-time equipment image model of any product for the first time, it will search for the corresponding production data according to the product number; Inputting production material data, real-time product image model and real-time equipment image model into the data input layer of the production supervision model; The data input layer generates a model matching instruction according to the product number and production line number and production data information carried by the real-time product image model and the real-time equipment image model; The production supervision layer searches for the corresponding product production supervision model and equipment supervision model according to the model matching instructions, and then overlaps the product production supervision model and the equipment supervision model according to the product number and production line number they carry to generate a product supervision pair. The products generated by each production line in the model matching instructions are summarized in order, and each product supervision pair is spliced ​​in sequence.

[0015] Furthermore, the process of determining the status of products and production equipment according to the production supervision model includes: Match the corresponding product supervision pair according to the product number and production line number of the real-time product image model and the real-time equipment image model; Overlap and compare the real-time product image model with the product image model in the product supervision pair, and generate product abnormal production records based on the comparison results; Overlap and compare the real-time equipment image model with the equipment image model in the product supervision pair, and map the state change data to the corresponding normal state change data range for comparison, and generate equipment abnormal production records based on the comparison results; The result output layer outputs abnormal production records of products or abnormal status records of equipment, and then maintenance personnel repair the corresponding products or production equipment; When it is determined that the product production is completed, the production supervision layer generates a product production record based on all state change data, real-time product image model and real-time equipment image model received during the generation process and sends it to the cloud computing platform.

[0016] Furthermore, whenever the cloud computing platform receives a new product production record, the production supervision model is retrained using the methods of step S2, step S3, and step S4 to generate a new production supervision model.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention generates a corresponding product production supervision model and an equipment supervision model through a number of product production records, and then splices the product production supervision model and the equipment supervision model according to the product production data information, thereby improving the supervision efficiency and accuracy of products and production lines; 2. Whenever product production is completed, a new product production record is generated, and the production supervision model is updated according to the new product production record, which improves the timeliness of the production supervision model. At the same time, by establishing a corresponding equipment supervision model based on the equipment operation data of all production lines, the applicability and flexibility of the production supervision model are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0019] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0020] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention. Example

[0021] like Figure 1 As shown, a data model construction method, the data model construction method comprises the following steps: S1. Collect the production data information required for the product, as well as the product status data of the product at each moment on each production line during the normal production process, and collect the equipment operation data of each equipment on the production line at each moment; S2, setting up a cloud computing platform, and then the cloud computing platform establishes a product image model according to a plurality of pre-stored product production records, and then establishes a product production supervision model, and at the same time establishes a real-time product image model according to the product status data generated in step S1; S3, the cloud computing platform establishes an equipment image model based on the product production record, and then establishes an equipment supervision model, and at the same time establishes a real-time equipment image model based on the equipment operation data generated in step S1; S4: Establish a production supervision model, which includes a data input layer, a production supervision layer, and a result output layer. The product production supervision model and the equipment supervision model generated in step S2 and step S3 are input into the production supervision layer. S5. Overlap the product production supervision model and the equipment supervision model corresponding to the same production line, thereby generating a product supervision pair; S6. Input the real-time product image model, the real-time equipment image model and the production data information into the input layer, and then match the real-time product image model and the real-time equipment image model with the product supervision pair according to the production data information, and then generate the corresponding product abnormal production record and equipment abnormal status record, and generate the product production record after the product production is completed; S7. The cloud computing platform updates the production supervision model according to the newly generated product production records. Example

[0022] This embodiment is a further limitation of Embodiment 1, and step S1 is implemented by the following process: Number each production line in the production workshop S1, S2, ..., S n , and set numbers H1, H2, ..., H for each type of product m , where n and m are natural numbers greater than 0; Setting up a cloud computing platform to collect production data information required for the production of various types of products, numbering the production data information according to the product number, and sending it to the cloud computing platform; It should be noted that the production data information includes the types and quantities of the raw materials required for production, the production line numbers of the various types of raw materials for processing, and the order in which the product components generated by each production line are summarized; Each production equipment on each production line is equipped with a dedicated sensor, wireless signal device and infrared sensing device, where the sensor can be a temperature sensor, voltage sensor, etc.; When the product starts to be produced, the raw materials are placed on the corresponding numbered production lines according to the required quantity according to the production data information; When the production raw materials are placed on the corresponding production line, the infrared sensing device of the first production equipment on the production line collects infrared image data of the production raw materials; According to the processing time of each production equipment for the production of raw materials, set k i processing time, where k i Indicates the number is S i The number of production equipment on the production line, i is a natural number greater than 0, and i is less than or equal to n. It should be noted that due to the different processing steps of each production equipment, the processing time of each production equipment for the production of raw materials is not exactly the same; During the processing time, each infrared sensing device collects 20 to 50 infrared image data of the product according to the processing time length of the production equipment and integrates them to generate product status data; At the same time, each sensor and infrared sensing device collects the state change data of the production equipment where it is located and the infrared image data of the equipment, wherein the state change data includes the temperature change curve, the voltage change curve, etc., and the state change data and the infrared image data of the equipment are integrated to generate the equipment operation data; It should be noted that the product infrared image data and the equipment infrared image data are generated synchronously, and after each set of product infrared image data and equipment infrared image data is generated, the state change data within the current time node is intercepted to generate real-time equipment operation data and real-time product status data, and the equipment operation data and generated product status data are sent to the cloud computing platform in real time. Example

[0023] This embodiment is a further limitation of Embodiment 1, and Step S2 is implemented by the following process: The cloud computing platform is preset with Num copies of product production records, and then the cloud computing platform extracts historical product status data generated during the production and processing of the product on each production line from the Num copies of product production records, wherein the product production records include historical product status data and historical equipment status data; The product infrared image data in the historical product status data is divided according to the number of processing moments of the production line, and then a corresponding product image model is established according to the product infrared image data, and Num product image models at the same position of the same production line are overlapped with each other, where Num is a natural number greater than 50; Then, the position range of each model line in the product image model is obtained, the position range is normally distributed, and the central area relative to 50% of the entire position range is selected as the normal position range of the product in the corresponding production line position, and marked on the corresponding product image model; Connect the k product image models on the same production line in sequence according to the corresponding processing time sequence, and then obtain the product production supervision model of the corresponding product on each production line, and mark the corresponding product and production line numbers; At the same time, the cloud computing platform binds the production data information of the products to the production supervision model of each product; At the same time, whenever the cloud computing platform receives product status data, it generates a corresponding real-time product image model based on the product infrared image data in the product status data, and marks the corresponding product and production line numbers. Example

[0024] This embodiment is a further limitation of Embodiment 1, and Step S3 is implemented by the following process: The cloud computing platform extracts historical equipment operation data from the product production records, and then extracts Num pieces of equipment infrared image data from the historical equipment operation data, and uses the method of generating product image models in Example 3 to generate equipment image models of each production equipment and the corresponding normal position range; Further, the state change data is extracted from the historical equipment operation data, and according to the number of types of the state change data, the same number of rectangular coordinate systems are established, and the Num state change data of the same production equipment on the same production line of the same type of product are mapped in the rectangular coordinate system; The state change data with the largest number of overlaps is taken as the standard state change data, and w time intervals are divided on the x-axis of the rectangular coordinate system, where w is a natural number greater than 10; Then, the calculus absolute value between each state change data and the standard state change data in each time interval is calculated; Set an abnormal threshold. If the absolute value of the integral in a time interval is greater than or equal to the abnormal threshold, it is determined that the corresponding state change data in the time interval has an abnormal change interval. If the absolute value of the integral in a time interval is less than the abnormal threshold, it is judged that the corresponding state change data in the time interval is normal; If there are more than three abnormal change intervals in the state change data, the corresponding state change data will be eliminated; If there are no more than three abnormal change intervals in the state change data, the abnormal change intervals in the corresponding state change data are removed, and the state change data after the removal is retained; If the state change data does not have an abnormal change interval, the state change data is directly retained; Integrate all the retained state change data to obtain the normal state change data range of the production equipment of the corresponding production line when producing the corresponding product, and mark the corresponding product and production line numbers; According to the front and rear position relationship of each production equipment on the corresponding production line, each equipment image model is connected in series, and the corresponding normal state change data range interval is associated with the corresponding equipment image model, thereby obtaining the equipment supervision model; At the same time, whenever the cloud computing platform receives equipment status data, it generates a corresponding real-time equipment image model based on the equipment infrared image data in the equipment status data, and marks the corresponding product and production line numbers. Example

[0025] This embodiment is a further limitation of Embodiment 1, and step S4 is implemented by the following process: Establishing a production supervision model, wherein the production supervision model is provided with a data input layer, a production supervision layer, and a result output layer, and then the cloud computing platform inputs all generated product production supervision models and equipment supervision models into the production supervision layer in the production supervision model; The data input layer is used to receive product production records, production data data, real-time product image models, equipment operation data and real-time equipment image models, and then generate corresponding model matching instructions; The production supervision layer is used to call the corresponding product production supervision model and equipment supervision model according to the model matching instruction, and then generate the corresponding product judgment result and equipment judgment result by judging the real-time equipment operation data and product status data; The result output layer is used to receive product judgment results and equipment judgment results from the production supervision layer, and then generate maintenance prompts and send them to corresponding maintenance personnel. Example

[0026] This embodiment is a further limitation of Embodiment 1, and step S5 is implemented by the following process: When the cloud computing platform generates the real-time product image model and real-time equipment image model of any product for the first time, it will search for the corresponding production data according to the product number; Then the cloud computing platform inputs the production material data, the real-time product image model and the real-time equipment image model into the data input layer of the production supervision model; The data input layer summarizes the order of products processed by each production line in the real-time product image model and the real-time equipment image model according to the product number and production line number, and the production data information, generates a model matching instruction and sends it to the production supervision layer; The production supervision layer searches for the corresponding product production supervision model and equipment supervision model according to the model matching instruction, and then overlaps the product production supervision model and the equipment supervision model according to the product number and production line number they carry to generate a product supervision pair; The products generated by each production line in the model matching instructions are summarized in order, and each product supervision pair is spliced ​​in sequence. Example

[0027] This embodiment is a further limitation of Embodiment 1, and step S6 is implemented by the following process: After receiving the state change data, the real-time product image model and the real-time equipment image model, the production supervision layer in the production supervision module matches the corresponding product supervision pair according to the product number and production line number carried by the real-time product image model and the real-time equipment image model; Overlap and compare the real-time product image model with the product image model in the product supervision pair. If more than 10% of the model lines in the real-time product image model are not within the normal position range, it is determined that there is a production abnormality in the current production line, and then a product abnormal production record is generated; If the model lines that are not in the normal position range in the real-time product image model are less than or equal to 10%, it is determined that the product is generated normally on the current production line and no operation is performed; Furthermore, the real-time equipment image model is overlapped and compared with the equipment image model in the product supervision pair. If more than 10% of the model lines in the real-time equipment image model are not within the normal position range, it is determined that the production equipment has an abnormal operation, and then an equipment abnormal status record is generated; If the model lines that are not in the normal position range in the real-time device image model are less than or equal to 10%, it is judged that the pen is running normally and no operation is performed; At the same time, the state change data is mapped to the corresponding normal state change data range. If more than 10% of the state change data is not within the normal state change data range, it is judged that the equipment has an abnormal operation, and then the equipment abnormal state record is generated; If the state change data is less than or equal to 10% and is not within the normal state change data range, it is judged that the operation of the pen is normal and no operation is performed; The production supervision layer sends the generated product abnormal production record or equipment abnormal status record to the result output layer, and then the result output layer sends the product abnormal production record or equipment abnormal status record to the cloud computing platform, and the cloud computing platform sends the product abnormal production record or equipment abnormal status record to the corresponding maintenance personnel, and after the maintenance personnel repair the product or production equipment, the cloud computing platform re-collects the corresponding product status data and equipment operation data; When it is determined that the product production is completed, the production supervision layer generates a product production record based on all state change data, real-time product image model and real-time equipment image model received during the generation process and sends it to the cloud computing platform. Example

[0028] This embodiment is a further limitation of Embodiment 1, and step S7 is implemented by the following process: Whenever the cloud computing platform receives a new product production record, the production supervision model is retrained using the methods of step S2, step S3, and step S4 to generate a new production supervision model.

[0029] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A data model construction method, characterized in that: The data model construction method comprises the following steps: S1. Collect the production data information required for the product, as well as the product status data of the product at each moment on each production line during the normal production process, and collect the equipment operation data of each equipment on the production line at each moment; S2, setting up a cloud computing platform, and then the cloud computing platform establishes a product image model according to a plurality of pre-stored product production records, and then establishes a product production supervision model, and at the same time establishes a real-time product image model according to the product status data generated in step S1; S3, the cloud computing platform establishes an equipment image model based on the product production record, and then establishes an equipment supervision model, and at the same time establishes a real-time equipment image model based on the equipment operation data generated in step S1; S4. Establish a production supervision model, wherein the production supervision model is provided with a data input layer, a production supervision layer, and a result output layer, and the product production supervision model and the equipment supervision model generated in step S2 and step S3 are input into the production supervision layer; S5. Overlap the product production supervision model and the equipment supervision model corresponding to the same production line, thereby generating a product supervision pair; S6. Input the real-time product image model, the real-time equipment image model and the production data information into the input layer, and then match the real-time product image model and the real-time equipment image model with the product supervision pair according to the production data information, and then generate the corresponding product abnormal production record and equipment abnormal status record, and generate the product production record after the product production is completed; S7. The cloud computing platform updates the production supervision model according to the newly generated product production records.

2. A data model construction method according to claim 1, characterized in that: The process of collecting the product status data includes: Set numbers for each production line and each type of product, set up a cloud computing platform, and then collect the production data information required for the production of each type of product, and set numbers for the production data information according to the product numbers; The production data information includes the types and quantities of raw materials required for production, the production line numbers of each type of raw material for processing, and the order in which the product components generated by each production line are summarized; Set up exclusive sensors and infrared sensing devices for each production equipment on each production line; When the production raw materials are placed on the corresponding production line, the infrared sensing device of the first production equipment on the production line collects infrared image data of the production raw materials, and sets several processing times according to the processing time of each production equipment for the production raw materials; During the processing time, each infrared sensing device collects product infrared image data according to the processing time length of the production equipment where it is located and integrates it to generate product status data.

3. A data model construction method according to claim 2, characterized in that: The process of collecting the equipment operation data includes: Various sensors and infrared sensing devices located on the production line collect state change data of production equipment and equipment infrared image data, and integrate the state change data and equipment infrared image data to generate equipment operation data.

4. A data model construction method according to claim 3, characterized in that: The process of establishing the real-time product image model and the real-time equipment image model includes: When the cloud computing platform receives the product status data, it generates a corresponding real-time product image model based on the product infrared image data in the product status data, and marks the corresponding product and production line number; When the cloud computing platform receives the equipment status data, it generates a corresponding real-time equipment image model based on the equipment infrared image data in the equipment status data, and marks the corresponding product and production line numbers.

5. A data model construction method according to claim 2, characterized in that: The process of establishing the product production supervision model includes: The cloud computing platform is preset with Num product production records, which include historical product status data and historical equipment status data; Establish a corresponding product image model according to the product infrared image data, and overlap Num product image models at the same position of the same production line, where Num is a natural number greater than 50; Obtain the position range of each model line in the product image model, distribute the position range normally, select the central area relative to 50% of the entire position range as the normal position range of the product in the corresponding production line, and mark it on the corresponding product image model; The k product image models on the same production line are connected in sequence according to the corresponding processing time sequence, and then the product production supervision model of the corresponding product on each production line is obtained, and the corresponding product and production line numbers are marked. At the same time, the production data information is bound to each product production supervision model.

6. A data model construction method according to claim 5, characterized in that: The process of establishing the equipment supervision model includes: The cloud computing platform extracts historical equipment operation data from product production records, and extracts Num pieces of equipment infrared image data from the historical equipment operation data, and then generates equipment image models of each production equipment and the corresponding normal position range; Extract state change data from historical equipment operation data, and establish the same number of rectangular coordinate systems according to the number of types of state change data, and map Num state change data of the same production equipment on the same production line of the same type of product into the rectangular coordinate system; The state change data with the largest number of overlaps is taken as the standard state change data, and w time intervals are divided on the x-axis of the rectangular coordinate system, where w is a natural number greater than 10; Then, the calculus absolute value between each state change data and the standard state change data in each time interval is obtained; Set an abnormal threshold. If the absolute value of the integral in a time interval is greater than or equal to the abnormal threshold, it is determined that the corresponding state change data in the time interval has an abnormal change interval. Then, based on the judgment result, the abnormal change interval of the state change data is deleted or directly eliminated; Integrate all the retained state change data to obtain the normal state change data range of the production equipment of the corresponding production line when producing the corresponding product, and mark the corresponding product and production line numbers; According to the front and rear position relationship of each production equipment on the corresponding production line, each equipment image model is connected in series, and the corresponding normal state change data range interval is associated with the corresponding equipment image model, thereby obtaining the equipment supervision model.

7. A data model construction method according to claim 6, characterized in that: The functions of each layer in the production supervision model include: The cloud computing platform inputs the product production supervision model and the equipment supervision model into the production supervision layer in the production supervision model; The data input layer is used to receive product production records, production data data, real-time product image models, equipment operation data and real-time equipment image models, and then generate corresponding model matching instructions; The production supervision layer is used to call the corresponding product production supervision model and equipment supervision model according to the model matching instruction, and then generate the corresponding product judgment result and equipment judgment result by judging the real-time equipment operation data and product status data; The result output layer is used to receive product judgment results and equipment judgment results from the production supervision layer, and then generate maintenance prompts and send them to corresponding maintenance personnel.

8. A data model construction method according to claim 7, characterized in that: The process of generating the product regulatory pair includes: Inputting production material data, real-time product image model and real-time equipment image model into the data input layer of the production supervision model; The data input layer generates a model matching instruction according to the product number and production line number and production data information carried by the real-time product image model and the real-time equipment image model; The production supervision layer searches for the corresponding product production supervision model and equipment supervision model according to the model matching instructions, and then overlaps the product production supervision model and the equipment supervision model according to the product number and production line number they carry to generate a product supervision pair. The products generated by each production line in the model matching instructions are summarized in order, and each product supervision pair is spliced ​​in sequence.

9. A data model construction method according to claim 8, characterized in that: The process of judging the status of products and production equipment according to the production supervision model includes: Match the corresponding product supervision pair according to the product number and production line number of the real-time product image model and the real-time equipment image model; Overlap and compare the real-time product image model with the product image model in the product supervision pair, and generate product abnormal production records based on the comparison results; Overlap and compare the real-time equipment image model with the equipment image model in the product supervision pair, and map the state change data to the corresponding normal state change data range for comparison, and generate equipment abnormal production records based on the comparison results; The result output layer outputs abnormal production records of products or abnormal status records of equipment, and then maintenance personnel repair the corresponding products or production equipment; When it is determined that the product production is completed, the production supervision layer generates a product production record based on all state change data, real-time product image model and real-time equipment image model received during the generation process and sends it to the cloud computing platform.

10. A data model construction method according to claim 9, characterized in that: Whenever the cloud computing platform receives a new product production record, the production supervision model is retrained using the methods of step S2, step S3, and step S4 to generate a new production supervision model.

Images