Industrial internet of things for intelligent control of filling production, control methods and media thereof
The present disclosure discloses an industrial Internet of Things for intelligent control of filling production, including a user platform, a service platform, a management platform, a sensor network platform, and an object platform interacting in sequence, wherein the service platform is provided with a plurality of independent service sub-platforms; the management platform is provided with a general management platform and a plurality of independent management sub-platforms, the management sub-platforms interacting with the general management platform, and the management sub-platforms interacting with the corresponding service sub-platforms; and the sensor network platform is provided with a general sensor network platform and a plurality of independent sensor network sub-platforms, the sensor network sub-platforms interacting with the general sensor network platform, and the general sensor network platform interacting with the general management platform. The present disclosure determines the target production device through processing a relationship diagram using a fault prediction model, the relationship diagram indicating a relationship between the devices corresponding to the production processes of the mixed filling production line, which can effectively reduce the complexity of the detection device, reduce defect detection costs, and improve the yield rate.
1. An industrial Internet of Things for intelligent control of filling production, comprising a user platform, a service platform, a management platform, a sensor network platform, and an object platform interacting in sequence, wherein
the service platform is provided with a plurality of independent service sub-platforms; the management platform is provided with a general management platform and a plurality of independent management sub-platforms, the management sub-platforms interacting with the general management platform, and the management sub-platforms interacting with corresponding service sub-platforms; the sensor network platform is provided with a general sensor network platform and a plurality of independent sensor network sub-platforms, the sensor network sub-platforms interacting with the general sensor network platform; wherein
the plurality of sensor network sub-platforms are configured to:
receive defect data of products detected by different defect detection devices in the object platform;
the general sensor network platform is configured to:
summarize all the defect data and send the defect data to the general management platform uniformly;
the general management platform is configured to:
revise production parameters of a target production device in the object platform according to the plurality of defect data, and send the revised production parameters and the corresponding defect data as reference data to the target production device and the management sub-platform corresponding to the target production device, wherein the target production device comes from a continuous production line, the continuous production line is a mixed filling production line, and determining the target production device includes:
determining a relationship diagram based on a production process of the mixed filling production line and devices related to the production process, wherein the relationship diagram indicates a relationship between the devices corresponding to the production processes of the mixed filling production line;
determining the target production device by processing the relationship diagram based on a fault prediction model, the fault prediction model including a graph neural network model;
store the reference data and send the reference data to the user platform through the corresponding service sub-platform; and
the user platform is configured to display the reference data to a user.
2. The industrial Internet of Things for intelligent control of filling production of claim 1 , wherein the general management platform is configured with a defect analysis model and a defect revising model; and
the general management platform is configured to:
input the defect data into the defect analysis model and receive analysis data outputted by the defect analysis model; and
input the analysis data into the defect revising model, and determine the revised production parameters based on data outputted by the defect revising model.
3. The industrial Internet of Things for intelligent control of filling production of claim 2 , wherein the defect revising model is a neural network model, an input of the defect revising model includes the defect data, preset candidate production parameters corresponding to each target production device, and an output of the defect revising model includes prediction defects corresponding to the candidate production parameters; and
the general management platform is configured to determine candidate production parameters that the prediction defects meet preset conditions as the revised production parameters.
4. The industrial Internet of Things for intelligent control of filling production of claim 1 , wherein the relationship diagram includes at least one node and at least one edge, the devices include a disinfection device, a peeling and cleaning device, a segmentation device, a filling device, a sterilization device, and a cooling device; wherein
at least one node of the relationship diagram corresponds to the devices related to the production process, respectively, and the at least one node includes a disinfection device node, a peeling and cleaning device node, a segmentation device node, a filling device node, a sterilization device node, and a cooling device node;
the at least one edge of the relationship diagram is a directed edge, the direction of the directed edge includes pointing the disinfection device node to the peeling and cleaning device node, pointing the peeling and cleaning device node to the segmentation device node, pointing the segmentation device node to the filling device node, pointing the filling device node to the sterilization device node, and pointing the sterilization device node to the cooling device node.
5. The industrial Internet for intelligent control of filling production of claim 4 , wherein features of the at least one node include the production parameters or monitoring data of each device.
6. The industrial Internet of Things for intelligent control of filling production of claim 4 , wherein the fault prediction model includes a feature extraction layer, the feature extraction layer is used to process image data of each node to obtain an image feature, and the image data is obtained by an image acquisition device; and
the general management platform is configured to determine the image feature as one of the features of the node.
7. The industrial Internet of Things for intelligent control of filling production of claim 2 , wherein the defect analysis model is configured with first data and second data; the first data is benchmark data corresponding to the defect data; the second data is a contribution degree of the production parameters of each production device corresponding to the defect data; and
the general management platform is configured that:
when the defect data is input into the defect analysis model, the defect analysis model generates differential data by comparing the defect data and the corresponding first data and calculates the contribution degree of the production parameters of each production device of all the differential data as the analysis data based on the second data.
8. The industrial Internet of Things for intelligent control of filling production of claim 7 , wherein the defect analysis model obtains the contribution degree of the production parameters of each production device of all the differential data based on the second data when calculating the analysis data; and
the defect analysis model generates the analysis data by performing normalized superposition on the contribution degree corresponding to the same production parameters of the differential data.
9. The industrial Internet of Things for intelligent control of filling production of claim 2 , wherein the defect revising model is configured with fourth data; the fourth data is a corresponding relationship between the analysis data and the revising range of the production parameters; and
the general management platform is configured that:
when the analysis data is input into the defect revising model, the defect revising model generates the revised production parameters based on the analysis data and the corresponding fourth data.
10. A control method for an industrial Internet of Things for intelligent control of filling production, which is applied to a user platform, a service platform, a management platform, a sensor network platform, and an object platform interacting in sequence, wherein
the service platform is provided with a plurality of independent service sub-platforms; the management platform is provided with a general management platform and a plurality of independent management sub-platforms, the management sub-platforms interacting with the general management platform, and the management sub-platforms interacting with corresponding service sub-platforms; the sensor network platform is provided with a general sensor network platform and a plurality of independent sensor network sub-platforms, the sensor network sub-platforms interacting with the general sensor network platform; and
the control method comprises:
receiving, by the plurality of sensor network sub-platforms, defect data of products detected by different defect detection devices in the object platform respectively; summarizing all the defect data and sending the defect data to the general management platform uniformly by the general sensor network platform;
revising production parameters of a target production device in the object platform according to the plurality of defect data, and sending the revised production parameters and the corresponding defect data as reference data to the target production device and the management sub-platform corresponding to the target production device by the general management platform, wherein the target production device comes from a continuous production line, the continuous production line is a mixed filling production line, and determining the target production device includes:
determining a relationship diagram based on a production process of the mixed filling production line and devices related to the production process, wherein the relationship diagram indicates a relationship between the devices corresponding to the production processes of the mixed filling production line;
determining the target production device by processing the relationship diagram based on a fault prediction model, the fault prediction model including a graph neural network model;
storing the reference data and sending the reference data to the user platform through the corresponding service sub-platform by the management sub-platform; and
displaying the reference data to a user by the user platform.
11. The control method for an industrial Internet of Things for intelligent control of filling production of claim 10 , wherein the general management platform is configured with a defect analysis model and a defect revising model;
the general management platform inputs the defect data into the defect analysis model and receives analysis data outputted by the defect analysis model; and
the general management platform inputs the analysis data into the defect revising model and determines the revised production parameters based on data outputted by the defect revising model.
12. The control method for an industrial Internet of Things for intelligent control of filling production of claim 11 , wherein the defect revising model is a neural network model, an input of the defect revising model includes the defect data, preset candidate production parameters corresponding to each target production device, and an output of the defect revising model includes prediction defects corresponding to the candidate production parameters; and
the general management platform is configured to determine candidate production parameters that the prediction defects meet preset conditions as the revised production parameters.
13. The control method for an industrial Internet of Things for intelligent control of filling production of claim 12 , wherein the input of the defect revising model also includes a credibility degree of each target production device with a fault.
14. The control method for an industrial Internet of Things for intelligent control of filling production of claim 10 , wherein the relationship diagram includes at least one node and at least one edge, the devices include a disinfection device, a peeling and cleaning device, a segmentation device, a filling device, a sterilization device, and a cooling device; wherein
at least one node of the relationship diagram corresponds to the devices related to the production process, respectively, and the at least one node includes a disinfection device node, a peeling and cleaning device node, a segmentation device node, a filling device node, a sterilization device node, and a cooling device node;
the at least one edge of the relationship diagram is a directed edge, the direction of the directed edge includes pointing the disinfection device node to the peeling and cleaning device node, pointing the peeling and cleaning device node to the segmentation device node, pointing the segmentation device node to the filling device node, pointing the filling device node to the sterilization device node, and pointing the sterilization device node to the cooling device node.
15. The control method for an industrial Internet of Things for intelligent control of filling production of claim 14 , wherein features of the at least one node include the production parameters or monitoring data of each device.
16. The control method for an industrial Internet of Things for intelligent control of filling production of claim 14 , wherein the fault prediction model includes a feature extraction layer, the feature extraction layer is used to process image data of each node to obtain an image feature, and the image data is obtained by an image acquisition device; and
the general management platform is configured to determine the image feature as one of the features of the node.
17. The control method for an industrial Internet of Things for intelligent control of filling production of claim 11 , wherein the defect analysis model is configured with first data and second data; the first data is benchmark data corresponding to the defect data; the second data is a contribution degree of the production parameters of each production device corresponding to the defect data; and
when the general management platform inputs the defect data into the defect analysis model, the defect analysis model generates differential data by comparing the defect data and the corresponding first data and calculates the contribution degree of the production parameters of each production device of all the differential data as the analysis data based on the second data.
18. The control method for an industrial Internet of Things for intelligent control of filling production of claim 17 , wherein the defect analysis model obtains the contribution degree of the production parameters of each production device of all the differential data based on the second data when calculating the analysis data; and
the defect analysis model generates the analysis data by performing normalized superposition on the contribution degree corresponding to the same production parameters of the differential data.
19. The control method for an industrial Internet of Things for intelligent control of filling production of claim 11 , wherein the defect revising model is configured with fourth data; the fourth data is a corresponding relationship between the analysis data and the revising range of the production parameters; and
when the general management platform inputs the analysis data into the defect revising model, the defect revising model generates the revised production parameters based on the analysis data and the corresponding fourth data.
20. A non-transitory computer-readable medium, comprising at least one set of instructions, wherein when executed by one or more processors of a computing device, the at least one set of instructions direct the computing device to perform the control method for an industrial Internet of Things for intelligent control of filling production according to claim 10 .