Data processing method and device, electronic equipment and storage medium

Abstract

The invention relates to a data processing method and device, electronic equipment and a storage medium. The method applied to edge equipment comprises the steps of obtaining production line data; preprocessing the production line data, and determining a first processing result including an actual production state of an associated production line based on a preprocessing result; sending the preprocessing result to cloud equipment; wherein the preprocessing result is used for the cloud equipment to determine a second processing result for predicting the production state of the production line. Through the method, on one hand, the real-time simulation and response efficiency of the automatic production line is improved; and on the other hand, the efficiency of predicting the production state of the automatic production line is improved, and the efficiency and safety of data transmission are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent production technology, and in particular to a data processing method and device, electronic equipment, and storage medium for an intelligent production system. Background Art

[0002] With the development of science and technology, intelligent production technology has become one of the important development directions of the global manufacturing industry, and has also become the core force leading the transformation of manufacturing industry into service and the transformation of production methods. Although the field of intelligent production technology has made great progress and formed an intelligent production equipment industry system represented by industrial robots, there are still certain shortcomings in the direction of common technologies and key technologies. For example, the current strength of industrial control software is relatively weak. How to improve the software control of intelligent production is one of the technical problems that need to be solved urgently. Summary of the invention

[0003] The present disclosure provides a data processing method and device, an electronic device, and a storage medium.

[0004] According to a first aspect of an embodiment of the present disclosure, a data processing method is provided, which is applied to an edge device, including:

[0005] Obtain production line data;

[0006] Preprocessing the production line data, and determining a first processing result including an actual production status of the associated production line based on the preprocessing result;

[0007] The preprocessing result is sent to a cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

[0008] In some embodiments, the preprocessing of the production line data and determining, based on the preprocessing result, a first processing result including an actual production status of the associated production line includes:

[0009] Preprocessing the production line data using a preset algorithm module to obtain a preprocessing result of each algorithm module;

[0010] Based on each simulation task associated with the actual production status of the production line, simulation processing is performed on the preprocessing results of the algorithm module associated with the simulation task to obtain the first processing results corresponding to each simulation task.

[0011] In some embodiments, the edge device is deployed with a virtual machine, a container is deployed in the virtual machine, and each algorithm module is distributed in a different container of the virtual machine.

[0012] In some embodiments, the method further comprises:

[0013] Using a preset production line model to train historical production line data to obtain initial parameters of the preset production line model; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result;

[0014] The initial parameters of the preset production line model are sent to the cloud device; wherein the initial parameters are used by the cloud device to train and update the preset production line model in combination with the preprocessing results of the historical production line data.

[0015] In some embodiments, the method further comprises:

[0016] Obtaining a control parameter for controlling the production line sent by the cloud device based on the second processing result;

[0017] The production status of the production line is controlled based on the control parameters sent by the cloud device.

[0018] In some embodiments, obtaining production line data includes:

[0019] Acquire real-time data of the production line; wherein the real-time data includes at least one of the following: operation data of equipment in the production line, and environmental data of equipment in the production line;

[0020] and / or,

[0021] Acquire non-real-time data of the production line; wherein the non-real-time data includes at least one of the following: layout data of the production line, equipment data in the production line.

[0022] According to a second aspect of an embodiment of the present disclosure, there is provided a data processing method, which is applied in a cloud device, and includes:

[0023] Receive a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line;

[0024] Based on the preprocessing result, a second processing result including predicting a production status of the production line is determined.

[0025] In some embodiments, the method further comprises:

[0026] Receiving initial parameters of a preset production line model sent by the edge device; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result, and the initial parameters are obtained by the edge device training historical production line data using the preset production line model;

[0027] Based on the initial parameters and the preprocessing results of the historical production line data, the preset production line model is trained and updated.

[0028] In some embodiments, the method further comprises:

[0029] determining a control parameter for controlling the production line based on the second processing result;

[0030] The control parameter is sent to the edge device; wherein the control parameter is used by the edge device to control the production status of the production line.

[0031] According to a third aspect of an embodiment of the present disclosure, a data processing apparatus is provided, which is applied to an edge device, and the apparatus includes:

[0032] Acquisition module, used to obtain production line data;

[0033] A first determination module, configured to preprocess the production line data and determine a first processing result including an actual production status of the associated production line based on the preprocessing result;

[0034] A sending module is used to send the preprocessing result to a cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

[0035] In some embodiments, the first determination module is also used to preprocess the production line data using a preset algorithm module to obtain preprocessing results of each algorithm module; based on each simulation task related to the actual production status of the production line, the preprocessing results of the algorithm module associated with the simulation task are simulated to obtain the first processing results corresponding to each simulation task.

[0036] In some embodiments, the edge device is deployed with a virtual machine, a container is deployed in the virtual machine, and each algorithm module in the first determination module is distributed in different containers of the virtual machine.

[0037] In some embodiments, the apparatus further comprises:

[0038] The first model processing module is used to train the historical production line data with the preset production line model to obtain the initial parameters of the preset production line model; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result; the initial parameters of the preset production line model are sent to the cloud device; wherein the initial parameters are used by the cloud device to train and update the preset production line model in combination with the preprocessing results of the historical production line data.

[0039] In some embodiments, the apparatus further comprises:

[0040] The first control module is used to obtain the control parameters for controlling the production line sent by the cloud device based on the second processing result; and control the production status of the production line based on the control parameters sent by the cloud device.

[0041] In some embodiments, the acquisition module is also used to acquire real-time data of the production line; wherein the real-time data includes at least one of the following: operation data of the equipment in the production line, environmental data of the equipment in the production line; and / or, to acquire non-real-time data of the production line; wherein the non-real-time data includes at least one of the following: layout data of the production line, equipment data in the production line.

[0042] According to a fourth aspect of an embodiment of the present disclosure, there is provided a data processing apparatus, which is applied to a cloud device, and the apparatus includes:

[0043] A receiving module, configured to receive a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line;

[0044] The second determination module is used to determine a second processing result including predicting the production status of the production line based on the preprocessing result.

[0045] In some embodiments, the apparatus further comprises:

[0046] A second model processing module is used to receive the initial parameters of the preset production line model sent by the edge device; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result, and the initial parameters are obtained by the edge device using the preset production line model to train historical production line data; based on the initial parameters and the preprocessing results of the historical production line data, the preset production line model is trained and updated.

[0047] In some embodiments, the apparatus further comprises:

[0048] A second control module is used to determine control parameters for controlling the production line based on the second processing result; and send the control parameters to the edge device; wherein the control parameters are used by the edge device to control the production status of the production line.

[0049] According to a fifth aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0050] a processor; a memory for storing instructions executable by the processor;

[0051] Wherein, the processor is configured to execute the data processing method as described in the first aspect or the second aspect above.

[0052] According to a sixth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, including:

[0053] When the instructions in the storage medium are executed by the processor of the edge device, the edge device is enabled to execute the data processing method as described in the first aspect; or, when the instructions in the storage medium are executed by the processor of the cloud device, the cloud device is enabled to execute the data processing method as described in the second aspect above.

[0054] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0055] In the disclosed embodiment, an edge-cloud fusion design architecture is adopted, and the processing of different types of data of the automated production line is deployed on different device ends respectively, and the edge and cloud ends cooperate to process the production line data of the automated production line. On the one hand, the edge device obtains the production line data and performs preprocessing, and determines the first processing result of the actual production status of the associated production line based on the preprocessing result (belonging to the processing of real-time production line data), and the processing of the real-time production line data is set in the edge device that can directly interact with the automated production line, which is conducive to real-time simulation and feedback of the automated production line, and improves the efficiency of real-time simulation and response of the automated production line; on the other hand, the second processing result (belonging to the processing of non-real-time production line data) of predicting the actual production status of the production line requires processing of massive production line data. Since the edge device first preprocesses the production line data and then sends the preprocessing result to the cloud device for production line status prediction, it is conducive to reducing the direct interaction of massive data between the edge device and the cloud device, and also conducive to reducing the repeated preprocessing of production line data by multiple devices, improving the efficiency and security of data transmission, and improving the efficiency of predicting the production status of the automated production line.

[0056] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0058] Figure 1 It is a framework structure diagram of an intelligent production system shown in related technology 1.

[0059] Figure 2 This is a functional architecture of a production line digital twin system shown in related technology 2.

[0060] Figure 3 This is an example of a data processing method flow shown in the embodiment of the present disclosure Figure 1 .

[0061] Figure 4 It is an example diagram of a system architecture of a data processing method shown in an embodiment of the present disclosure.

[0062] Figure 5 This is an example diagram of virtual machine container deployment of a data processing method shown in an embodiment of the present disclosure.

[0063] Figure 6 This is an example of a data processing method flow shown in the embodiment of the present disclosure Figure 2 .

[0064] Figure 7 It is a flow interaction diagram of a data processing method shown in an embodiment of the present disclosure.

[0065] Figure 8 1 is a diagram of a data processing device applied to an edge device shown in an embodiment of the present disclosure.

[0066] Fig. 9 1 is a diagram of a data processing device applied to a cloud device shown in an embodiment of the present disclosure.

[0067] Fig.10 It is a block diagram of an electronic device shown in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0068] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0069] Intelligent production systems usually include intelligent equipment, CNC equipment, and mechanical equipment, etc., and refer to the use of advanced information technology and intelligent technology to achieve automation, intelligence, and efficiency in the production process. As an important part of the field of intelligent production technology, intelligent production systems are also subject to some problems due to the influence of intelligent production technology, such as the high cost of intelligent production equipment, the high complexity of intelligent production processes, the high technical difficulty of intelligent production management, and the lack of real-time feedback systems. Therefore, how to provide a data processing method that can be applied to intelligent production systems is one of the technical problems that need to be solved urgently.

[0070] Figure 1 It is a framework structure diagram of an intelligent production system shown in the related art 1, such as Figure 1 As shown, the intelligent production system includes: a physical system, an edge digital twin node, a digital twin management and control system, a production and manufacturing simulation system, an order system, and an artificial intelligence (AI) algorithm model library. In related technology 1, the intelligent sensing device in the physical system is used to obtain physical production line information in real time and transmit it to the edge digital twin node; the edge digital twin node is used to build equipment models, predict equipment failures and life, and perform visual displays; the digital twin management and control system is used to interact with the order system and the artificial intelligence algorithm model library and generate simulation analysis jobs for production equipment scheduling, and the production and manufacturing simulation system is used to generate simulation results based on the received simulation analysis jobs; the digital twin management and control system is also used to optimize the production scheduling strategy based on the received simulation results, and send production scheduling instructions to the physical system to achieve equipment failure monitoring, prediction and maintenance on the edge side.

[0071] Figure 2 This is a functional architecture of a production line digital twin system shown in related technology 2, such as Figure 2 As shown, the production line digital twin system includes: real-time data acquisition module, industrial information model module, human-computer interaction module, production line performance evaluation module, and production line control service module. In related technology 2, the production line digital twin system instantiates the digital twin production line through the industrial information model module, and writes communication, motion control and other programs based on the physical engine (Unity3D) to build the human-computer interaction module of the production line twin system; the data acquisition module collects and analyzes the actual state data of the production line in real time, and synchronizes the twin system with the physical production line data based on the communication program; at the same time, the state data of the production line is managed and forwarded to the production line performance evaluation module; the production line performance evaluation module analyzes the production line operation data and calculates the performance indicators of the production line in real time; the production line control service module sends numerical control instructions to the equipment in the system to control the workshop production line equipment, so as to simplify the simulation process of production line manufacturing execution and improve the design efficiency of the production line simulation system.

[0072] However, in related technology 1, the AI ​​algorithm model library that can process real-time and non-real-time data is deployed entirely in the cloud, which is not conducive to the calculation of data with high real-time requirements, and it is difficult to achieve the effect of line-level real-time simulation and response; directly uploading the data collected and initially processed on the edge side to the cloud may have a certain risk of data leakage. In related technology 2, the data processing algorithm is deployed on the data acquisition module, resulting in slow algorithm operation and difficulty in achieving real-time response, and only simulation and feedback are performed on the production process, lacking long-term non-real-time big data calculation and prediction.

[0073] In this regard, the present disclosure provides a data processing method that can be applied to an intelligent production system. Figure 3 This is an example of a data processing method flow shown in the embodiment of the present disclosure Figure 1 ,like Figure 3 As shown, the data processing method applied to the edge device includes the following steps:

[0074] S11, obtaining production line data;

[0075] S12, preprocessing the production line data, and determining a first processing result including an actual production status of the associated production line based on the preprocessing result;

[0076] S13, sending the preprocessing result to a cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

[0077] In the embodiments of the present disclosure, the edge device can be a transit device between the physical world and the network world, and can perform various tasks such as data processing, data storage, parameter control, communication management, etc.; the edge device can be a terminal type such as a smart phone, tablet computer, notebook, personal digital assistant (PDA), handheld device, computing device, vehicle-mounted device and other terminal devices; it can also be other electronic devices, such as servers, storage devices, databases, virtual machines and other types of electronic devices, or a cluster composed of the above devices.

[0078] In the disclosed embodiments, the cloud device is generally an electronic device located in the cloud of the network, which can provide services such as data storage, data processing and communication management through the network, and can be an electronic device of the type of server, storage device, database, virtual machine, etc., or a cluster composed of the above devices, such as a server cluster, database cluster, etc. In the disclosed embodiments, a communication connection is established between the edge device and the cloud device in a wireless or wired manner, and the wireless communication connection mode involved includes, for example, Bluetooth or mobile hotspot (Wireless Fidelity, WiFi), etc.

[0079] It should be noted that the intelligent production system is a comprehensive system, which usually includes an automatic production line module, a sensing and control module, an industrial communication module, a data analysis module, an artificial intelligence module, etc. In the disclosed embodiment, the data processing method can be deployed to the edge device and the cloud device respectively, and the edge device and the cloud device can collaboratively process the production line data in the intelligent production system, perform production line monitoring and feedback, fault prediction, life prediction, etc. on the intelligent production system, so as to guide the operation of the intelligent production system and optimize the production process.

[0080] In step S11, the edge device obtains production line data, which can be real-time acquisition of production line data or acquisition of production line data at a certain sampling interval, wherein the production line data refers to various types of data in the automated production line, such as the operation data of the production line equipment, the environmental data of the environment in which the production line equipment is located, and the local data of the production line equipment.

[0081] Figure 4 is a system architecture example diagram of a data processing method shown in an embodiment of the present disclosure, such as Figure 4 As shown in the figure, the system architecture includes the physical layer, automation layer, edge layer and cloud platform. The physical layer can deploy production, logistics, environmental monitoring and other equipment in the automated production line; the automation layer can deploy collection, control, management and other equipment in the automated production line; the edge layer can deploy the aforementioned edge devices; and the cloud platform can deploy the aforementioned cloud devices.

[0082] Exemplarily, the physical layer can deploy various types of equipment for automated production lines, including: hardware equipment for automated production lines, such as programmable logic controllers (PLCs), motors, robots, sensors, inverters, solenoid valves, industrial cameras, instruments, etc.; logistics and transportation equipment for automated production lines, such as automated guided vehicles (AGVs), etc.; communication equipment for automated production lines, such as communication interfaces, network cables, etc.; physical environment monitoring equipment for automated production lines, such as vibration sensors, light sensors, etc., which are used to monitor physical quantities (vibration, electromagnetism, radiation, light, heat, etc.) in the environment that affect the equipment generation activities in the automated production line.

[0083] For example, as mentioned above Figure 4As shown, the automation layer can deploy various types of equipment for automated production lines, including: data acquisition equipment, production line control equipment, line control equipment, visual positioning equipment, communication management equipment (such as message clients), etc. The equipment in the automation layer is a functional device that can directly act on the automated production line. It should be noted that the data acquisition equipment is used to collect the status data, equipment parameters, etc. of the equipment in the automated production line, as well as the data of various terminals and various sensors of the equipment; the production line control equipment is used to manage and control the operation of various equipment on the production line, and can automatically perform the set operations and tasks, coordinate and synchronize the operation of multiple devices, and optimize the efficiency and production capacity of the entire production line. At the same time, it collects data during the operation of the equipment and feeds it back to the data acquisition module; the line control equipment is used to control and manage the transportation system, conveyor belts, transportation equipment and other line equipment on the automated production line, which can realize the coordinated optimization of transportation activities between different equipment, adjust the production rhythm according to demand and actual conditions, and adjust the production rhythm according to product types. According to the production plan, material allocation is carried out, the position and status of each product on the production line are detected to realize progress tracking, and the operation data of the line equipment in each link can be collected and fed back to the data acquisition module; the visual positioning equipment uses computer vision technology to locate and identify items on the production line, and realizes automated assembly, quality inspection, item classification, etc. through the positioning and identification of parts, and also feeds back the data to the data acquisition module; the communication management equipment is used to communicate with the physical layer equipment to collect various types of data at the physical layer, and communicate with the edge layer equipment to feed back the data collected by the data acquisition module to the edge device of the edge layer, among which the message client is used to process message transmission, which can be to receive or send production line data.

[0084] In some embodiments, the edge device can directly obtain the production line data such as production, logistics, and environmental monitoring of the associated automated production line from each device in the physical layer; and obtain the control and management line data of the associated automated production line from each device in the automation layer. In other embodiments, the acquisition device of the automation layer can collect the production line data such as production, logistics, and environmental monitoring of the associated automated production line from each device in the physical layer, and the control and management device of the automation layer can obtain the control and management line data of the associated automated production line, and all the above production line data can be sent to the edge device together, so that the edge device can obtain all the production line data by communicating and interacting with the devices in the automation layer.

[0085] In step S12, the edge device preprocesses the production line data, and determines a first processing result including the actual production status of the associated production line based on the preprocessing result, wherein the preprocessing method may be data cleaning, data storage, data analysis, etc. of the production line data; the actual production status of the production line represents the actual operating state of the production line, which may be a normal state, a fault state, etc. It should be noted that the production line data may include real-time data and non-real-time data, and the preprocessing result obtained based on the production line data may be a real-time preprocessing result and / or a non-real-time preprocessing result, and the first processing result may be a control parameter for controlling the automated production line determined by the edge device based on the real-time preprocessing result and / or the non-real-time preprocessing result, such as a scheduling instruction, a decision instruction, etc. In addition, the edge device may also determine the fault prediction result, life prediction result, etc. of the automated production line based on the non-real-time preprocessing result.

[0086] In the embodiments of the present disclosure, there are multiple ways for the edge device to determine the first processing result. In some embodiments, the edge device can preset functional modules that implement various functions of the automated production line, use each functional module to pre-process the production line data and determine the first processing result of the actual production status of the associated production line, wherein the functional modules at least include: a visualization module, a production line decision module, a data interaction management module, a line level processing module, a communication module, etc.

[0087] In other embodiments, the edge device may preset an algorithm module for preprocessing production line data and preset simulation tasks that can realize various functions of the automated production line. The simulation tasks may be composed of different algorithm modules. The edge device uses the preprocessing results obtained by the preset algorithm module to obtain the first processing results corresponding to each simulation task based on the preprocessing results of the algorithm module associated with the simulation task.

[0088] In step S13, the edge device sends the preprocessing result to the cloud device, which can be sent to the cloud device in real time or at a certain time interval, so that the cloud device determines the second processing result including the prediction of the production status of the production line based on the preprocessing result. The second processing result can be the fault prediction result, life prediction result, etc. of the automated production line determined by the cloud device based on the real-time preprocessing result and / or the non-real-time preprocessing result. In addition, the cloud device can also determine the control parameters (such as scheduling instructions, decision instructions, etc.) for controlling the automated production line based on the real-time preprocessing result of the production line, and send the control parameters to the edge device to guide the operation of the automated production line.

[0089] In the disclosed embodiment, the edge device and the cloud device can also use big data analysis technology and artificial intelligence technology to analyze and process the production line data to obtain the first processing result and / or the second processing result of the actual production status of the associated production line. For example, the preset production line model and historical production line data are used for training to obtain the preset production line model parameters of the automated production line, and the currently acquired production line data and the preset production line model are used for prediction to obtain the control parameters, faults and / or life prediction results of the automated production line.

[0090] Normally, local and short-periodic data (real-time production line data) in the automated production line can be used to simulate the automated production line, and control data for controlling the operation of the automated production line can be generated based on the simulation results (first processing result); cloud data and long-periodic data (non-real-time production line data) in the automated production line can be used to predict the production status of the automated production line (second processing result). It should be noted that big data analysis technology is usually used to predict the production status, resulting in non-real-time production line data usually being long-term historical production line data with a large amount of data.

[0091] In the disclosed embodiment, an edge-cloud fusion design architecture is adopted, and the processing of different types of data of the automated production line is deployed on different device ends respectively, and the edge and cloud ends cooperate to process the production line data of the automated production line. On the one hand, the edge device obtains the production line data and performs preprocessing, and determines the first processing result of the actual production status of the associated production line based on the preprocessing result (belonging to the processing of real-time production line data), and the processing of the real-time production line data is set in the edge device that can directly interact with the automated production line, which is conducive to real-time simulation and feedback of the automated production line, and improves the efficiency of real-time simulation and response of the automated production line; on the other hand, the second processing result (belonging to the processing of non-real-time production line data) of predicting the actual production status of the production line requires processing of massive production line data. Since the edge device first preprocesses the production line data and then sends the preprocessing result to the cloud device for production line status prediction, it is conducive to reducing the direct interaction of massive data between the edge device and the cloud device, and also conducive to reducing the repeated preprocessing of production line data by multiple devices, improving the efficiency and security of data transmission, and improving the efficiency of predicting the production status of the automated production line.

[0092] In some embodiments, the preprocessing of the production line data and determining, based on the preprocessing result, a first processing result including an actual production status of the associated production line includes:

[0093] Preprocessing the production line data using a preset algorithm module to obtain a preprocessing result of each algorithm module;

[0094] Based on each simulation task associated with the actual production status of the production line, simulation processing is performed on the preprocessing results of the algorithm module associated with the simulation task to obtain the first processing results corresponding to each simulation task.

[0095] In the embodiment of the present disclosure, the edge device may be configured to include a simulation task and an algorithm module architecture, each simulation task may be composed of one or more algorithm modules, and the simulation task may call one or more algorithm modules to implement the function corresponding to the simulation task. Figure 4 As shown, the algorithm modules preset by the edge device in the computing cluster may include at least one of the following: model encapsulation algorithm, computing engine algorithm, data cleaning algorithm, data storage algorithm, computing power management algorithm, and communication management algorithm; the simulation tasks preset by the edge device in the simulation application may include at least one of the following: production line visualization task of the production line, real-time decision-making task of production problems, data interaction and management task, line-level simulation response task, and communication management task.

[0096] In the disclosed embodiments, the edge device uses a preset algorithm module to preprocess the production line data to obtain the preprocessing results of each algorithm module. For example, the edge device can use a data cleaning algorithm to clean the production line data to obtain cleaned production line data; for another example, the edge device can use a computing engine algorithm to calculate and analyze the production line data to obtain control data for controlling the automated production line.

[0097] In the disclosed embodiment, the edge device performs simulation processing on the preprocessing results of the algorithm module associated with the simulation task based on each simulation task of the actual production status of the associated production line, and obtains the first processing result corresponding to each simulation task. Among them, the simulation task can call one or more associated algorithm modules. When there are multiple associated algorithm modules, the simulation task can call multiple algorithm modules in a serial or parallel manner.

[0098] Exemplarily, the real-time decision-making task for production problems in edge devices can analyze and process local, real-time, short-cycle production line data by calling data cleaning algorithms, data storage algorithms, computing power management algorithms, and communication management algorithms, so as to achieve real-time monitoring and feedback of production status, timely predict problems that are likely to occur in the production process, and give predictive prompts and feedback. In the disclosed embodiment, the real-time decision-making task for production problems that processes real-time data is set on the edge device. Since the edge device can interact directly with the automated production line, it belongs to the short-distance data transmission and feedback type, which is conducive to rapid analysis and safe feedback of production line data, and is conducive to better supporting real-time intelligent decision-making and execution of local business.

[0099] For example, the production line visualization task in the edge device can build a digital mirror of the automated production line in the virtual space by calling the model encapsulation algorithm, computing engine algorithm, computing power management algorithm, data storage algorithm, etc., to achieve mapping from the physical world to the digital space and real-time synchronization between the virtual and physical worlds.

[0100] Exemplarily, the line-level simulation response task in the edge device can virtually simulate a series of processes such as the layout of the production line, the process path of production and manufacturing, and the logistics and transportation plan by calling data cleaning algorithms, data storage algorithms, communication management algorithms, and computing engine algorithms, so as to realize real-time closed-loop monitoring and feedback functions of the entire production process of the automated production line. It can realize real-time simulation and response at the line level, make up for the lack of real-time closed-loop monitoring and feedback functions in the field of intelligent production technology, so that production managers can understand a series of production states such as production efficiency, quality, energy consumption, etc. in real time during the production process, and can also promptly discover various problems that may arise in the production process and take corresponding measures to adjust and optimize the production line, thereby achieving improved production efficiency and product quality, reduced production energy consumption, and improved production safety.

[0101] In the embodiments of the present disclosure, the aforementioned various simulation tasks deployed in the edge device can perform corresponding processing on the production line data so as to realize functions such as the aforementioned production line visualization, real-time decision-making on production problems, data interaction and management, line-level simulation response, communication management, etc., which is conducive to the realization of rich and diverse functions of the automated production line, reduces the situation of line-level real-time simulation and scheduling functions in the production process, improves the integrity of information transmission, improves the real-time feedback processing efficiency of production line equipment failures, improves the timeliness of the issuance of production line control strategies, and realizes the effect of real-time feedback of data throughout the entire life cycle of production.

[0102] As mentioned above Figure 4In the embodiment of the present disclosure, the simulation task in the edge device and the preset algorithm module can communicate based on the industrial communication protocol to build a cross-platform industrial control network. Among them, the industrial communication protocol can be the protocol standard of the open automation system (OpenPlatform Communications Unified Architecture, OPC UA), which is a new generation of technical standards for data transmission in industrial networks. It not only integrates all the functions of the traditional industrial standard specification (OLE for Process Control, OPC), but also has a safe and reliable cross-platform architecture. It can be based on the programming interface of the transport layer (for example, TCP IPsocket), allowing devices with different operating systems to communicate with each other, and realizing unified communication between devices from different manufacturers.

[0103] It should be noted that in the aforementioned related technologies 1 and 2, the corresponding functional modules are directly used to simulate the response of the automated production line and / or predict the production status, and each functional module may reuse the same algorithm. This leads to redundant code design of the functional modules and waste of computer resources; if new functional modules need to be added later, the new functional modules need to be redesigned, and the new functional modules need to interact with each end for data, resulting in complex functional module design and insufficient compatibility and scalability.

[0104] In the disclosed embodiment, the edge device uses a preset algorithm module to preprocess the production line data, so that each simulation task can use the preprocessing result of the associated preset algorithm module to obtain the first processing result corresponding to the simulation task. On the one hand, each simulation task can reuse the preset algorithm module, which is conducive to saving computer resources; on the other hand, if new functions need to be added later, the new simulation task can be configured by the preset algorithm module, and the new function (simulation task) of the edge device is simple and easy to implement, and has good compatibility and scalability.

[0105] In some embodiments, the edge device is deployed with a virtual machine, a container is deployed in the virtual machine, and each algorithm module is distributed in a different container of the virtual machine.

[0106] In the disclosed embodiments, one or more virtual machines may be deployed in an edge device, each virtual machine having an independent hard disk and operating system. The virtual machine may use software to simulate a complete computer system to replace a physical computer to implement various tasks in a computer operating system. The virtual machine may be a Linux virtual machine, a Microsoft virtual machine, a Mac virtual machine, etc.

[0107] In the disclosed embodiment, one or more containers are deployed in the virtual machine, and the algorithm modules in the edge device are distributed in different containers of the virtual machine. For example, the preset algorithm module is encapsulated and deployed in the container. It should be noted that the container has excellent characteristics such as lightweight, security, portability, and second-level startup. The preset algorithm module is deployed in the container in the edge device. So that the edge device can achieve more efficient system resource utilization, second-level startup speed, consistent operating environment, deployment at any location, and more convenient migration.

[0108] Figure 5 is a diagram showing an example of virtual machine container deployment of a data processing method according to an embodiment of the present disclosure. Figure 5 As shown, a host operating system (Host Operation System) is deployed on the infrastructure (Infrastructure), that is, the edge device of the present disclosure, and multiple virtual machine hypervisors (Hypervisor) are installed on the host operating system. The virtual machine hypervisor is used to create a client operating system (Guest OS), for example, a Linux or Windows virtual operating system is installed on the client, and a container is deployed on each client operating system, and the container is used to encapsulate one or more applications (APP), that is, the algorithm module in the present disclosure.

[0109] It should be noted that there are certain limitations when deploying containers directly on physical computers. For example, since containers share a common kernel and namespace, if an attacker can access the operating system of the physical computer, they can access all applications on the physical computer, which poses certain security risks. It is extremely difficult to break into the virtual machine from the virtual machine to its hypervisor layer. For another example, when multiple containers are running at the same time, an application may occupy more operating system resources during operation, thereby depriving other applications of the minimum operating system resources required to continue running, resulting in forced shutdown of the operating system and unstable system operation caused by system resource preemption.

[0110] In the disclosed embodiment, the edge device adopts a combination of virtual machines and containers for deployment, deploys a virtual machine including a container on the edge device, and deploys a preset algorithm module into the container of the virtual machine. On the one hand, by deploying a virtual machine on the edge device, the system operation efficiency can be improved, the operation and maintenance cost can be reduced, the work deployment can be convenient, and the application performance can be higher; on the other hand, the container is deployed on the virtual machine, and the preset algorithm module is accommodated by the container, which is conducive to improving isolation, security, and stability.

[0111] In some embodiments, the method further comprises:

[0112] Using a preset production line model to train historical production line data to obtain initial parameters of the preset production line model; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result;

[0113] The initial parameters of the preset production line model are sent to the cloud device; wherein the initial parameters are used by the cloud device to train and update the preset production line model in combination with the preprocessing results of the historical production line data.

[0114] In the disclosed embodiment, the edge device and / or the cloud device may preset a production line model, and the edge device and the cloud device may preset the same or different preset production line models. The preset production line model may be a machine learning model, and the preset production line model may be trained independently; or the preset production line model may be collaboratively trained using data from multiple devices, for example, using federated learning to achieve collaborative training, which has the feature of supporting multiple participants to collaborate in building a shared machine learning model.

[0115] In some embodiments, when the edge device and the cloud device deploy different preset production line models, the edge device and the cloud device can respectively train the preset production line models. For example, the edge device uses the preset production line model to train the historical production line data to obtain the initial parameters of the preset production line model, and uses the preset production line model corresponding to the initial parameters as a model for determining the first processing result, such as a production line state control model (real-time prediction model).

[0116] In other embodiments, when the edge device and the cloud device deploy the same preset production line model, the edge device and the cloud device can jointly train the preset production line model, as described above. Figure 4 As shown, an edge model of a preset production line model is deployed on an edge device, and a cloud model of a preset production line model is deployed on a cloud device. Federated learning can be used to send the initial parameters trained by the edge device to the cloud device, so that the cloud device can combine the initial parameters and the preprocessing results of the cloud historical production line data to further train to obtain the target parameters (more accurate preset model parameters), and use the target parameters to update the model parameters of the preset production line model. Among them, the preset production line model of the cloud device can be updated with the target parameters so that the cloud device can determine the second processing result, such as the production line status prediction model (non-real-time prediction model); the target parameters can also be sent down to the edge device to update the preset production line model of the edge device so that the edge device can determine the first processing result.

[0117] Exemplarily, the edge device sets a real-time business model and a non-real-time business model, and the cloud device only sets the non-real-time business. The real-time business model is only trained on the edge device and outputs the first processing result. The non-real-time business model obtains initial parameters after preprocessing on the edge device, and uploads the initial parameters to the cloud device through the federated learning framework. The historical production line data of the cloud device is combined for further training to obtain more accurate model parameters (target parameters), and the second processing result is predicted by the cloud device.

[0118] It should be noted that the construction of the preset production line model requires the calculation and training of massive historical production line data. On the one hand, due to the limited storage capacity of edge devices and the global, non-real-time, long-cycle big data analysis and processing problems such as life prediction, direct processing in the edge device will result in high data calculation pressure and high data transmission pressure. On the other hand, if the edge device directly uploads the historical production line data to the cloud device, and uses the cloud device to predict each link in property insurance, it will not meet the requirements of business models with high real-time requirements, and will also face the risk of data leakage.

[0119] In the disclosed embodiment, based on the edge-cloud fusion architecture, the edge and cloud collaborate to train the preset production line model. On the one hand, the edge device performs preliminary training on the preset production line model to obtain initial parameters, which can reduce the amount of data and computing required for training, and help reduce the computing pressure of the edge device; on the other hand, the edge device sends the initial parameters to the cloud device instead of directly sending a large amount of production line data, which reduces the amount of data transmitted between devices, reduces the transmission pressure between devices, reduces the risk of data leakage, and helps improve data security. In addition, the edge device and the cloud device collaborate to build a shared preset production line model, and the benefits obtained by the collaborative model at each end can be reasonably distributed through the value matching mechanism, which is conducive to increasing the training volume of the preset production line model and improving the accuracy of the preset production line model.

[0120] In some embodiments, the method further comprises:

[0121] Obtaining a control parameter for controlling the production line sent by the cloud device based on the second processing result;

[0122] The production status of the production line is controlled based on the control parameters sent by the cloud device.

[0123] In the disclosed embodiment, the edge device obtains the control parameters for controlling the production line sent by the cloud device based on the second processing result, wherein the control parameters may be business rules, model parameters, decision instructions, etc. In the disclosed embodiment, the cloud device may send business rules for controlling the production line, such as production scheduling optimization instructions, when the second processing result indicates that the production status of the production line is normal; or may send decision results for controlling the production line, such as decision instructions for whether the production line continues to operate, when the second processing result indicates that the production status of the production line is abnormal.

[0124] In the disclosed embodiment, the edge device controls the production status of the production line based on the control parameters sent by the cloud device. The control parameters can be directly sent to the production line equipment in the aforementioned physical layer to control the production status of the production line equipment; or the control parameters can be first sent to each device in the aforementioned automation layer, and the production status of the production line equipment in the physical layer is controlled by the equipment in the automation layer. Among them, the production status of the production line equipment is controlled, for example, according to the production scheduling optimization instructions, the operation of various equipment and machines on the production line is controlled to optimize the efficiency and production capacity of the entire production line; for another example, according to the decision instructions, the start and stop of the corresponding equipment and machines on the production line are controlled.

[0125] Exemplarily, both the edge device and the cloud device are set with real-time business models and non-real-time business models. The real-time business model uploads initial parameters to the cloud device after being processed by the edge device. The cloud device obtains the target parameters (more accurate model parameters) and sends them to the edge device. The edge device predicts and obtains the first processing result based on the target parameters.

[0126] In the disclosed embodiment, the edge device controls the production status of the production line based on the control parameters sent by the cloud device based on the second processing result. On the one hand, since the control parameters are obtained after further processing by the cloud device, and the cloud device can usually store and process more production line data, the control parameters generated by the cloud device are more accurate, which is beneficial to improving the control accuracy of the automated production line. On the other hand, the control instruction received by the edge device is the timing for the cloud device to control the sending of the control parameters based on the second processing result, which increases the flexibility of the automated production line control.

[0127] In some embodiments, obtaining production line data includes:

[0128] Acquire real-time data of the production line; wherein the real-time data includes at least one of the following: operation data of equipment in the production line, and environmental data of equipment in the production line;

[0129] and / or,

[0130] Acquire non-real-time data of the production line; wherein the non-real-time data includes at least one of the following: layout data of the production line, equipment data in the production line.

[0131] In the disclosed embodiments, the edge device can obtain real-time production line data and / or non-real-time production line data, and can adopt appropriate data acquisition methods according to the characteristics of the device and / or the data, thereby improving the efficiency of data collection and saving the device resources of the edge device. In addition, the classified acquisition of production line data is also conducive to the subsequent data analysis and processing of the real-time and non-real-time production line services, which is convenient for improving the efficiency of edge device data processing.

[0132] Figure 6 This is an example of a data processing method flow shown in the embodiment of the present disclosure Figure 2 ,like Figure 6 As shown, the data processing method applied to the cloud device includes the following steps:

[0133] S21, receiving a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line;

[0134] S22. Based on the preprocessing result, determine a second processing result including predicting the production status of the production line.

[0135] In the disclosed embodiments, the cloud device is generally an electronic device located in the cloud of the network, which can provide services such as data storage, data processing and communication management through the network, and can be an electronic device of the type of server, storage device, database, virtual machine, etc., or a cluster composed of the above devices, such as a server cluster, database cluster, etc. In the disclosed embodiments, a communication connection is established between the cloud device and the edge device in a wireless or wired manner, and the wireless communication connection method involved includes, for example, Bluetooth or mobile hotspot.

[0136] As mentioned above, the data processing method can be deployed in edge devices and cloud devices respectively. The edge devices and cloud devices can jointly process the production line data in the intelligent production system, perform production line monitoring and feedback, fault prediction, life prediction, etc. on the intelligent production system, so as to guide the operation of the intelligent production system and optimize the production process.

[0137] In step S21, the cloud device receives the preprocessing result sent by the edge device, which can be received in real time or at a certain time interval. As mentioned above, the preprocessing result is obtained by the edge device based on the production line data. The preprocessing method can be that the edge device performs data cleaning, data storage, data analysis, etc. on the production line data. The preprocessing result can be a real-time preprocessing result and / or a non-real-time preprocessing result.

[0138] In step S22, the cloud device determines a second processing result including a prediction of the production status of the production line based on the preprocessing result. The cloud device determines the second processing result based on the real-time preprocessing result and / or non-real-time preprocessing result, for example, a fault prediction result or a life prediction result of the automated production line. In addition, the cloud device can also determine the control parameters (such as scheduling instructions, decision instructions, etc.) for controlling the automated production line based on the real-time preprocessing result of the production line, and send the control parameters to the edge device to guide the operation of the automated production line.

[0139] In the disclosed embodiment, the cloud device can also use big data analysis technology and artificial intelligence technology to analyze and process the production line data to obtain a second processing result of predicting the actual production status of the production line. For example, the cloud device uses the preset production line model and historical production line data for training to obtain the preset production line model parameters of the automated production line, and uses the currently acquired production line data and the preset production line model for prediction to obtain the control parameters, faults and / or life prediction results of the automated production line.

[0140] It should be noted that the second processing result of the cloud device predicting the actual production status of the production line (which belongs to the processing of non-real-time production line data) requires processing massive production line data, which is prone to risks such as high data transmission pressure and production line data leakage. In addition, if the cloud device directly uses the production line data for predictive processing, it is not conducive to the reuse of intermediate data, wastes equipment computing resources, and has low data processing efficiency.

[0141] In the disclosed embodiment, an edge-cloud fusion design architecture is adopted, and the processing of different types of data of the automated production line is deployed on different device ends respectively, and the edge and cloud ends cooperate to process the production line data of the automated production line. On the one hand, the cloud device predicts the production line status based on the preprocessing results of the received edge device after preprocessing the production line data, which helps to reduce the direct exchange of massive data between the edge device and the cloud device, and improves the efficiency and security of data transmission; on the other hand, the edge device first preprocesses the production line data, so that the cloud device and the edge device can reuse the preprocessing results, which helps to reduce the situation where multiple devices repeatedly preprocess the production line data, and improves the efficiency of predicting the production status of the automated production line.

[0142] In some embodiments, the method further comprises:

[0143] Receiving initial parameters of a preset production line model sent by the edge device; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result, and the initial parameters are obtained by the edge device training historical production line data using the preset production line model;

[0144] Based on the initial parameters and the preprocessing results of the historical production line data, the preset production line model is trained and updated.

[0145] In the disclosed embodiment, the cloud device and / or the edge device may preset a production line model, and the cloud device and the edge device may preset the same or different preset production line models. The preset production line model may be a machine learning model, and the preset production line model may be trained independently; or the preset production line model may be collaboratively trained using data from multiple devices, for example, using federated learning to achieve collaborative training, which has the characteristic of supporting multiple participants to collaborate in building a shared machine learning model.

[0146] In some embodiments, when the cloud device and the edge device deploy the same preset production line model, the cloud device and the edge device can jointly train the preset production line model, as described above. Figure 4 As shown, an edge model of a preset production line model is deployed on an edge device, and a cloud model of a preset production line model is deployed on a cloud device. The edge device trains to obtain the initial parameters of the preset production line model. After receiving the initial parameters, the cloud device combines the initial parameters and the preprocessing results of the cloud historical production line data to further train to obtain the target parameters (more accurate preset model parameters), and uses the target parameters to update the model parameters of the preset production line model. The preset production line model of the cloud device can be updated with the target parameters so that the cloud device can determine the second processing result, such as a production line status prediction model (non-real-time prediction model); the target parameters can also be sent to the edge device to update the preset production line model of the edge device so that the edge device can determine the first processing result.

[0147] It should be noted that the construction of the preset production line model requires the calculation and training of massive historical production line data. On the one hand, due to the limited storage capacity of edge devices and the global, non-real-time, long-cycle big data analysis and processing problems such as life prediction, direct processing in the edge device will result in high data calculation pressure and high data transmission pressure. On the other hand, if the edge device directly uploads the historical production line data to the cloud device, and uses the cloud device to predict each link in property insurance, it will not meet the requirements of business models with high real-time requirements, and will also face the risk of data leakage.

[0148] In the disclosed embodiment, based on the edge-cloud fusion architecture, the edge and cloud collaborate to train the preset production line model. On the one hand, the edge device performs preliminary training on the preset production line model to obtain initial parameters, which can reduce the amount of data and computing required for training, and help reduce the computing pressure of the edge device; on the other hand, the cloud device receives the initial parameters instead of directly receiving a large amount of production line data, which reduces the amount of data transmitted between devices, reduces the transmission pressure between devices, reduces the risk of data leakage, and helps improve data security. In addition, the cloud device and the edge device collaborate to build a shared preset production line model, and the benefits obtained by the collaborative model at each end can be reasonably distributed through the value matching mechanism, which is conducive to increasing the training volume of the preset production line model and improving the accuracy of the preset production line model.

[0149] In some embodiments, the method further comprises:

[0150] determining a control parameter for controlling the production line based on the second processing result;

[0151] The control parameter is sent to the edge device; wherein the control parameter is used by the edge device to control the production status of the production line.

[0152] In the disclosed embodiment, the cloud device determines the control parameters for controlling the production line based on the second processing result; wherein the control parameters may be business rules, model parameters, decision instructions, etc. In the disclosed embodiment, the cloud device may send the control parameters to the edge device in real time, or send the control parameters when the second processing result predicts a predetermined state of the production line. The predetermined state may be a production line failure, a normal production line, etc.

[0153] Illustratively, the cloud device may send business rules for controlling the production line, such as production scheduling optimization instructions, when the second processing result indicates that the production status of the production line is normal; or may send decision results for controlling the production line, such as decision instructions on whether to continue operating the production line, when the second processing result indicates that the production status of the production line is abnormal.

[0154] In the disclosed embodiment, the cloud device sends control parameters to the edge device so that the edge device controls the production status of the production line based on the control parameters. The edge device can directly send the control parameters to the production line equipment in the aforementioned physical layer to control the production status of the production line equipment; or it can first send the control parameters to the various devices in the aforementioned automation layer, and control the production status of the production line equipment in the physical layer through the devices in the automation layer. Among them, the production status of the production line equipment is controlled, for example, according to the production scheduling optimization instructions, the operation of various equipment and machines on the production line is controlled to optimize the efficiency and production capacity of the entire production line; for another example, according to the decision instructions, the start and stop of the corresponding equipment and machines on the production line are controlled.

[0155] In the disclosed embodiment, the cloud device sends control parameters based on the second processing result so that the edge device can control the production status of the production line. On the one hand, since the control parameters are obtained after further processing by the cloud device, and the cloud device can usually store and process more production line data, the accuracy of the control parameters generated by the cloud device is higher, which is conducive to improving the control accuracy of the automated production line; on the other hand, the cloud device can control the timing of sending the control parameters based on the second processing result, which increases the flexibility of automated production line control.

[0156] Figure 7 is a flow chart of a data processing method shown in an embodiment of the present disclosure, such as Figure 7 As shown, the data processing method is applied to edge devices and cloud devices, and includes the following steps:

[0157] S31, edge devices obtain production line data;

[0158] S32. The edge device preprocesses the production line data, and determines a first processing result including an actual production status of the associated production line based on the preprocessing result;

[0159] S33, the edge device sends the preprocessing result to the cloud device;

[0160] S34, the cloud device receives the preprocessing result sent by the edge device;

[0161] S35. The cloud device determines a second processing result including predicting the production status of the production line based on the preprocessing result.

[0162] In step S31, the edge device obtains production line data, which can be real-time acquisition of production line data or acquisition of production line data at a certain sampling interval, wherein the production line data refers to various types of data in the automated production line, such as the operation data of the production line equipment, the environmental data of the environment in which the production line equipment is located, and the local data of the production line equipment.

[0163] In step S32, the edge device preprocesses the production line data, and determines a first processing result including the actual production status of the associated production line based on the preprocessing result, wherein the preprocessing method may be data cleaning, data storage, data analysis, etc. of the production line data; the actual production status of the production line represents the actual operating status of the production line, which may be a normal state, a faulty state, etc.

[0164] In step S33, the edge device sends the preprocessing result to the cloud device. The preprocessing result can be sent to the cloud device in real time or at certain time intervals, so that the cloud device can determine a second processing result including a prediction of the production status of the production line based on the preprocessing result.

[0165] In step S34, the cloud device receives the preprocessing result sent by the edge device, and may receive the preprocessing result sent by the edge device in real time or at certain time intervals.

[0166] In step S35, the cloud device determines a second processing result including a prediction of the production status of the production line based on the preprocessing result, and the cloud device determines the second processing result based on the real-time preprocessing result and / or the non-real-time preprocessing result. In addition, the cloud device can also determine the control parameters (such as scheduling instructions, decision instructions, etc.) for controlling the automated production line based on the real-time preprocessing results of the production line, and send the control parameters to the edge device to guide the operation of the automated production line.

[0167] In the disclosed embodiment, an edge-cloud fusion design architecture is adopted, and the processing of different types of data of the automated production line is deployed on different device ends respectively, and the edge and cloud ends cooperate to process the production line data of the automated production line. On the one hand, the edge device obtains the production line data and performs preprocessing, and determines the first processing result of the actual production status of the associated production line based on the preprocessing result (belonging to the processing of real-time production line data), and the processing of the real-time production line data is set in the edge device that can directly interact with the automated production line, which is conducive to real-time simulation and feedback of the automated production line, and improves the efficiency of real-time simulation and response of the automated production line; on the other hand, the second processing result (belonging to the processing of non-real-time production line data) of predicting the actual production status of the production line requires processing of massive production line data. Since the edge device first preprocesses the production line data and then sends the preprocessing result to the cloud device for production line status prediction, it is conducive to reducing the direct interaction of massive data between the edge device and the cloud device, and also conducive to reducing the repeated preprocessing of production line data by multiple devices, improving the efficiency and security of data transmission, and improving the efficiency of predicting the production status of the automated production line.

[0168] As mentioned above Figure 4As shown, this embodiment is a data processing method applied to an intelligent production system, and the system architecture of this embodiment includes a physical layer, an automation layer, an edge layer, and a cloud platform. Among them, the physical layer deploys the automated production line equipment (production equipment, logistics equipment, physical environment monitoring equipment, etc.) of the intelligent production system; the automation layer deploys various types of equipment that control the automated production line (data acquisition equipment, production line control equipment, line control equipment, visual positioning equipment, communication management equipment). The edge layer (i.e., the edge device in the embodiment of the present disclosure) includes simulation applications and computing clusters; the cloud platform (i.e., the cloud device in the embodiment of the present disclosure) deploys the prediction model.

[0169] In this embodiment, the edge layer can obtain real-time production line data and non-real-time production line data in the automated production line through the physical layer and the automation layer, and perform data sentiment and storage. That is, the edge device in the embodiment of the present disclosure obtains the production line data and pre-processes the production line data.

[0170] In this embodiment, a virtualization deployment method is adopted in the computing cluster of the edge layer. By combining the virtual machine and the container, each algorithm module in the computer cluster is deployed in the virtual machine container. That is, in the embodiment of the present disclosure, the edge device is deployed with a virtual machine, and the virtual machine is deployed with a container. Each algorithm module is distributed in different containers of the virtual machine. Among them, the algorithm modules include: model encapsulation algorithm, computing engine algorithm, data cleaning algorithm, data storage algorithm, computing power management algorithm, communication management algorithm, etc.

[0171] In this embodiment, the simulation application of the edge layer can call one or more algorithm modules to implement the functions corresponding to the simulation application, that is, in the embodiment of the present disclosure, the simulation task is associated with one or more algorithm modules, wherein the simulation application includes: production line visualization, real-time decision-making on production problems, data interaction and management, line-level simulation response, communication management, etc. In this embodiment, the computing cluster and simulation application of the edge layer are both provided with a communication management module, which can exchange information model data based on an industrial communication protocol (for example, a protocol standard of an open automation system).

[0172] In this embodiment, various production line models (e.g., fault prediction model, life prediction model, production line control model, etc.) are deployed on both the edge layer and the cloud platform. In this embodiment, an edge model of a preset production line model is deployed on the edge device, and a cloud model of the preset production line model is deployed on the cloud device. Federated learning can be used to send the initial parameters trained by the edge device to the cloud device, so that the cloud device can further train the target parameters of the preset production line model (more accurate preset model parameters) by combining the initial parameters and the historical production line data on the cloud, and update the cloud model and edge model of the preset production line model using the target parameters.

[0173] In the disclosed embodiment, an edge-cloud fusion design architecture is adopted, and the real-time data of the automated production line is deployed at the edge layer for processing. Real-time decisions and responses can be made for local, real-time, and short-cycle data, thereby improving the real-time and security of the intelligent production system. A large amount of non-real-time data of the automated production line is deployed on the cloud platform for processing, which helps to reduce the computing pressure of the edge layer and improve the prediction accuracy of the production line status of the intelligent production system. Different types of processing tasks are reasonably allocated to different device ends to achieve balanced allocation of computing resources and improve the operating efficiency of the intelligent production system. Based on the edge-cloud fusion architecture and the application of federated learning algorithms, multi-terminal device data sharing is realized, thereby improving data processing security. Containers and virtual machines are deployed at the edge layer, and the algorithm is modularly encapsulated in the container, thereby improving the flexibility, portability, and security of algorithm deployment. By encapsulating the algorithm, different simulation applications can call different algorithm combinations to achieve the operation of the target function, thereby enhancing the scalability of the intelligent production system and making it easy to access other functions.

[0174] Figure 8 is a diagram of a data processing device applied to an edge device shown in an embodiment of the present disclosure, comprising Figure 8 It is known that, including:

[0175] An acquisition module 101 is used to acquire production line data;

[0176] A first determination module 102, configured to preprocess the production line data and determine a first processing result including an actual production status of the associated production line based on the preprocessing result;

[0177] The sending module 103 is used to send the preprocessing result to the cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

[0178] In some embodiments, the first determination module 102 is also used to preprocess the production line data using a preset algorithm module to obtain preprocessing results of each algorithm module; based on each simulation task of the actual production status of the associated production line, the preprocessing results of the algorithm module associated with the simulation task are simulated to obtain the first processing results corresponding to each simulation task.

[0179] In some embodiments, a virtual machine is deployed on the edge device, a container is deployed in the virtual machine, and each algorithm module in the first determination module 102 is distributed in different containers of the virtual machine.

[0180] In some embodiments, the apparatus further comprises:

[0181] The first model processing module is used to train the historical production line data with the preset production line model to obtain the initial parameters of the preset production line model; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result; the initial parameters of the preset production line model are sent to the cloud device; wherein the initial parameters are used by the cloud device to train and update the preset production line model in combination with the preprocessing results of the historical production line data.

[0182] In some embodiments, the apparatus further comprises:

[0183] The first control module is used to obtain the control parameters for controlling the production line sent by the cloud device based on the second processing result; and control the production status of the production line based on the control parameters sent by the cloud device.

[0184] In some embodiments, the acquisition module 101 is also used to acquire real-time data of the production line; wherein the real-time data includes at least one of the following: operation data of the equipment in the production line, environmental data of the equipment in the production line; and / or, to acquire non-real-time data of the production line; wherein the non-real-time data includes at least one of the following: layout data of the production line, equipment data in the production line.

[0185] Fig. 9 is a diagram of a data processing device applied to a cloud device shown in an embodiment of the present disclosure, comprising Fig. 9 It is known that, including:

[0186] A receiving module 201 is used to receive a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line;

[0187] The second determination module 202 is used to determine a second processing result including predicting the production status of the production line based on the preprocessing result.

[0188] In some embodiments, the apparatus further comprises:

[0189] A second model processing module is used to receive the initial parameters of the preset production line model sent by the edge device; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result, and the initial parameters are obtained by the edge device using the preset production line model to train historical production line data; based on the initial parameters and the preprocessing results of the historical production line data, the preset production line model is trained and updated.

[0190] In some embodiments, the apparatus further comprises:

[0191] A second control module is used to determine control parameters for controlling the production line based on the second processing result; and send the control parameters to the edge device; wherein the control parameters are used by the edge device to control the production status of the production line.

[0192] Fig.10 is a block diagram of an electronic device shown in an embodiment of the present disclosure. The electronic device may be an edge device or a cloud device. Fig.10 The electronic device 900 includes a processing component 922, which further includes one or more processors, and a memory resource represented by a memory 932 for storing instructions executable by the processing component 922, such as an application. The application stored in the memory 932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 922 is configured to execute instructions to perform the above-mentioned network configuration method.

[0193] The device 900 may also include a power supply component 926 configured to perform power management of the device 900, a wired or wireless network interface 950 configured to connect the device 900 to a network, and an input / output (I / O) interface 958. The device 900 may operate an operating system stored in the memory 932, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or the like.

[0194] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the aforementioned data processing method.

[0195] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims above.

[0196] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A data processing method, characterized in that: Applied to an edge device, the method comprises: Obtain production line data; Preprocessing the production line data, and determining a first processing result including an actual production status of the associated production line based on the preprocessing result; The preprocessing result is sent to a cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

2. The method according to claim 1, characterized in that The preprocessing of the production line data and determining a first processing result including an actual production status of the associated production line based on the preprocessing result includes: Preprocessing the production line data using a preset algorithm module to obtain a preprocessing result of each algorithm module; Based on each simulation task associated with the actual production status of the production line, simulation processing is performed on the preprocessing results of the algorithm module associated with the simulation task to obtain the first processing results corresponding to each simulation task.

3. The method according to claim 2, characterized in that The edge device is deployed with a virtual machine, a container is deployed in the virtual machine, and each algorithm module is distributed in different containers of the virtual machine.

4. The method according to claim 1, characterized in that: The method further comprises: Using a preset production line model to train historical production line data to obtain initial parameters of the preset production line model; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result; The initial parameters of the preset production line model are sent to the cloud device; wherein the initial parameters are used by the cloud device to train and update the preset production line model in combination with the preprocessing results of the historical production line data.

5. The method according to claim 1, characterized in that The method further comprises: Obtaining a control parameter for controlling the production line sent by the cloud device based on the second processing result; The production status of the production line is controlled based on the control parameters sent by the cloud device.

6. The method according to claim 1, characterized in that The obtaining of production line data includes: Acquire real-time data of the production line; wherein the real-time data includes at least one of the following: operation data of equipment in the production line, and environmental data of equipment in the production line; and / or, Acquire non-real-time data of the production line; wherein the non-real-time data includes at least one of the following: layout data of the production line, equipment data in the production line.

7. A data processing method, characterized in that: Applied to a cloud device, the method includes: Receive a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line; Based on the preprocessing result, a second processing result including predicting a production status of the production line is determined.

8. The method according to claim 7, characterized in that The method further comprises: Receiving initial parameters of a preset production line model sent by the edge device; wherein the preset production line model includes a model for determining the first processing result and / or the second processing result, and the initial parameters are obtained by the edge device training historical production line data using the preset production line model; Based on the initial parameters and the preprocessing results of the historical production line data, the preset production line model is trained and updated.

9. The method according to claim 7, characterized in that: The method further comprises: determining a control parameter for controlling the production line based on the second processing result; The control parameter is sent to the edge device; wherein the control parameter is used by the edge device to control the production status of the production line.

10. A data processing device, characterized in that: Applied to edge devices, the device comprises: Acquisition module, used to obtain production line data; A first determination module, configured to preprocess the production line data and determine a first processing result including an actual production status of the associated production line based on the preprocessing result; A sending module is used to send the preprocessing result to a cloud device; wherein the preprocessing result is used by the cloud device to determine a second processing result including predicting the production status of the production line.

11. A data processing device, characterized in that: Applied to cloud devices, the device comprises: A receiving module, configured to receive a preprocessing result sent by an edge device; wherein the preprocessing result is obtained by the edge device preprocessing the production line data, and the preprocessing result is used by the edge device to determine a first processing result including an actual production status of the associated production line; The second determination module is used to determine a second processing result including predicting the production status of the production line based on the preprocessing result.

12. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to execute the data processing method according to any one of claims 1 to 6; or, is configured to execute the data processing method according to any one of claims 7 to 9.

13. A computer-readable storage medium, characterized in that: When the instructions in the storage medium are executed by the processor of the edge device, the edge device is enabled to execute the data processing method as described in any one of claims 1 to 6; or, when the instructions in the storage medium are executed by the processor of the cloud device, the cloud device is enabled to execute the data processing method as described in any one of claims 7 to 9.

Images