Industrial equipment control method, device, system and medium based on edge-cloud cooperation

Abstract

The application discloses an industrial equipment control method, device and system based on edge cloud cooperation and a readable storage medium. The method comprises the following steps: collecting target data of different data types of an industrial equipment according to a sampling period; sending the target data collected in the current sampling period to a cloud end, so that the cloud end performs incremental learning on a preset control model according to the target data collected in the current sampling period, and obtains target control models of different data types; receiving the target control models sent by the cloud end; inputting the target data collected in the current sampling period into the target control models according to the data types, so as to obtain control parameters of actuators in the industrial equipment corresponding to different data types; and sending control instructions to the industrial equipment. Thus, the production data is continuously analyzed and diagnosed on the edge side, the production data continuously accumulated is continuously self-learned and self-optimized on the cloud end, and the functions of a smart industrial system of the whole industrial production process are realized.

Description

Technical Field

[0001] This application relates to the field of Internet of Things (IoT) technology, and in particular to an industrial equipment control method, apparatus, system, and medium based on edge-cloud collaboration. Background Technology

[0002] In the era of the Internet of Everything, the development of industrial intelligence faces the following new technological challenges. The application of traditional industrial target control models requires repeated manual debugging, is only applicable to specific tasks, and requires re-debugging and reprogramming when encountering similar tasks or different environments. It cannot draw on past operational experience, resulting in problems such as low efficiency, poor adaptability, and insufficient dexterity. Moreover, the application data of traditional industrial target control models is managed locally, making it difficult to provide big data support. Especially when industrial equipment generates a large amount of data, command issuance and fault alarms are often not timely enough. Summary of the Invention

[0003] In view of this, this application provides an industrial equipment control method, device, system and medium based on edge-cloud collaboration. By continuously analyzing and diagnosing target data at the edge and continuously learning and optimizing the ever-increasing target data in the cloud, the system realizes the function of a smart industrial system for self-management of industrial production.

[0004] According to an embodiment of the first aspect of this application, an industrial equipment control method based on edge-cloud collaboration is provided, suitable for edge devices, the method comprising:

[0005] Collect target data of different types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data.

[0006] The target data collected in the current sampling period is sent to the cloud so that the cloud can incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is the target control model trained based on the target data collected in the previous sampling period.

[0007] Receive the target control model sent from the cloud;

[0008] According to the data type, the target data collected in the current sampling period is input into the target control model to obtain the control parameters of the actuators in industrial equipment corresponding to different data types;

[0009] Control commands are sent to industrial equipment to control the actuators to move according to the control parameters in the control commands.

[0010] Optionally, after collecting target data of different data types from industrial equipment according to the sampling period, the edge-cloud collaborative industrial equipment control method also includes:

[0011] Filter the target data.

[0012] According to an embodiment of the second aspect of this application, an industrial equipment control method based on edge-cloud collaboration is provided, applicable to the cloud, the method comprising:

[0013] Receive target data of different data types collected in the current sampling period from the edge terminal, wherein the target data includes: monitoring data, production data and / or equipment operation data;

[0014] Incremental learning is performed on the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is the target control model trained based on the target data collected in the previous sampling period.

[0015] If the target control model meets the preset conditions, the target control model is sent to the edge terminal so that the edge terminal can determine the control parameters of the actuators in industrial equipment corresponding to different data types through the target control model and the target data collected in the current sampling period.

[0016] Optionally, incremental learning is performed on the preset control model based on the target data collected in the current sampling period, including:

[0017] The target data collected in the current sampling period is divided into training data and test data;

[0018] Based on the data type, the preset control model is trained according to the training data to determine the candidate model for different data types.

[0019] Obtain the theoretical control parameters corresponding to the test data;

[0020] The test data is input into the candidate model according to the data type to obtain the test control parameters;

[0021] If the difference between the test control parameters and the theoretical control parameters is within a preset range, the candidate module is determined as the target control model.

[0022] Optionally, after receiving target data of different data types collected in the current sampling period from the edge terminal, the edge-cloud collaborative industrial equipment control method further includes:

[0023] The target data is expanded according to a preset step size corresponding to different data types to obtain updated target data. The preset step size is determined based on the range distribution characteristics of different data types.

[0024] Optionally, after receiving target data of different data types collected in the current sampling period from the edge terminal, the edge-cloud collaborative industrial equipment control method further includes:

[0025] The target data is input into the generative adversarial network corresponding to different data types to obtain the updated target data.

[0026] Optionally, after receiving target data of different data types collected in the current sampling period from the edge terminal, the edge-cloud collaborative industrial equipment control method further includes:

[0027] The target data is preprocessed, including noise reduction, cleaning, data transformation and / or labeling.

[0028] According to an embodiment of the third aspect of this application, an industrial equipment control method based on edge-cloud collaboration is provided, comprising:

[0029] The edge device collects target data of different data types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data.

[0030] The edge device sends the target data collected in the current sampling period to the cloud;

[0031] The cloud receives target data collected during the current sampling period from the edge device;

[0032] The cloud performs incremental learning on the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is the target control model trained based on the target data collected in the previous sampling period.

[0033] The cloud sends the target control model to the edge;

[0034] The target control model receives data from the cloud at the edge.

[0035] The edge inputs the target data collected in the current sampling period into the target control model according to the data type, and obtains the control parameters of the actuators in the industrial equipment corresponding to different data types;

[0036] The edge device sends control commands to the industrial equipment to control the actuators to move according to the control parameters in the control commands.

[0037] Optionally, the edge-cloud collaborative industrial equipment control method also includes:

[0038] The target data is filtered at the edge.

[0039] Optionally, the cloud performs incremental learning on the preset control model based on the target data collected in the current sampling period, including:

[0040] The cloud divides the target data collected in the current sampling period into training data and test data;

[0041] The cloud-based system trains preset control models according to data type and training data to determine candidate models for different data types.

[0042] Obtain the theoretical control parameters corresponding to the test data from the cloud;

[0043] The cloud platform inputs test data into the candidate model according to the data type to obtain test control parameters;

[0044] If the difference between the test control parameters and the theoretical control parameters is within a preset range, the cloud will determine the candidate model as the target control model.

[0045] Optionally, the edge-cloud collaborative industrial equipment control method also includes:

[0046] The cloud expands the target data according to the preset step size corresponding to different data types to obtain updated target data. The preset step size is determined according to the range distribution characteristics of different data types.

[0047] Optionally, the edge-cloud collaborative industrial equipment control method also includes:

[0048] The cloud inputs the target data into the generative adversarial network corresponding to different data types to obtain the updated target data.

[0049] Optionally, the edge-cloud collaborative industrial equipment control method also includes:

[0050] The target data is preprocessed in the cloud, including noise reduction, cleaning, data transformation and / or tagging.

[0051] According to an embodiment of the fourth aspect of this application, an industrial equipment control device based on edge-cloud collaboration is provided, comprising:

[0052] The data acquisition module is used to collect target data of different types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data.

[0053] The communication module is used to send the target data collected in the current sampling period to the cloud, so that the cloud can incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period.

[0054] Receive the target control model sent from the cloud;

[0055] The control module is used to input the target data collected in the current sampling period into the target control model according to the data type, so as to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types;

[0056] The communication module is also used to send control commands to industrial equipment to control the actuators according to the control parameters in the control commands.

[0057] Optionally, the edge-cloud collaborative industrial equipment control device also includes:

[0058] The filtering module is used to filter the target data.

[0059] According to an embodiment of the fifth aspect of this application, an industrial equipment control device based on edge-cloud collaboration is provided, comprising:

[0060] The communication module is used to receive target data of different data types collected in the current sampling period from the edge terminal. The target data includes: monitoring data, production data and / or equipment operation data.

[0061] The training module is used to incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is the target control model trained based on the target data collected in the previous sampling period.

[0062] The communication module is also used to send the target control model to the edge device, so that the edge device can determine the control parameters of the actuators in the industrial equipment corresponding to different data types through the target control model and the target data collected in the current sampling period.

[0063] Optionally, the edge-cloud collaborative industrial equipment control device also includes:

[0064] The classification module is used to divide the target data collected in the current sampling period into training data and test data;

[0065] The training module is specifically used to train the preset control model according to the data type and the training data, and to determine the candidate model for different data types.

[0066] The testing module is used to obtain the theoretical control parameters corresponding to the test data; and to input the test data into the candidate model according to the data type to obtain the test control parameters; and if the difference between the test control parameters and the theoretical control parameters is within a preset range, the candidate model is determined as the target control model.

[0067] Optionally, the edge-cloud collaborative industrial equipment control device also includes:

[0068] An expansion module is used to expand the target data according to a preset step size corresponding to different data types to obtain updated target data. The preset step size is determined based on the range distribution characteristics of different data types. Alternatively, the target data can be input into a generative adversarial network corresponding to different data types to obtain updated target data.

[0069] Optionally, the edge-cloud collaborative industrial equipment control device also includes:

[0070] The preprocessing module is used to preprocess the target data, including noise reduction, cleaning, data transformation and / or labeling.

[0071] According to an embodiment of the sixth aspect of this application, an edge-cloud collaborative industrial equipment control system is provided, comprising: industrial equipment, an edge device, and a cloud device;

[0072] Industrial equipment used to generate target data;

[0073] At the edge, it is used to collect target data of different types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data; and to send the target data collected in the current sampling period to the cloud.

[0074] In the cloud, the system receives target data collected in the current sampling period from the edge device; and performs incremental learning on a preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period; and if the target control model meets preset conditions, the system sends the target control model to the edge device.

[0075] At the edge, it is also used to receive the target control model sent from the cloud; and to input the target data collected in the current sampling period into the target control model according to the data type to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types; and to send control commands to the industrial equipment to control the actuators to act according to the control parameters in the control commands.

[0076] According to an embodiment of the seventh aspect of this application, a read storage medium is provided, on which a program or instructions are stored, which, when executed by a processor, implement the steps of the above-described industrial equipment control method based on edge-cloud collaboration.

[0077] Using the above technical solution, industrial equipment continuously generates new target data such as monitoring data, production data, and equipment operation data during operation. Different types of target data generated by the industrial equipment are periodically collected at the edge and sent to the cloud for incremental learning of the target control model. The cloud then returns the incrementally learned target control model to the edge. Thus, the edge can input the target data collected in the current sampling period into the updated target control model according to different data types to predict the control parameters of the actuators in the industrial equipment corresponding to different data types. Finally, the edge sends control commands to the industrial equipment, causing the actuators to act according to the control parameters output by the model, thereby achieving precise control of the industrial equipment. Through the above technical solutions, on the one hand, highly complex computations such as data management, analysis, prediction, and model training are deployed in the cloud and at the edge, realizing a code development model of cloud compilation / edge execution. This significantly reduces the hardware cost of industrial equipment. Moreover, through collaborative work between the cloud and the edge, the edge can perform calculations and processing closer to the industrial equipment, reducing data transmission latency and enabling control commands to be issued to the industrial equipment in a timely manner. Simultaneously, the edge only transmits necessary results to the cloud server, reducing the amount of data transmitted over the network, providing more efficient, flexible, and reliable computing and services, and improving the scalability of industrial applications. On the other hand, it fully considers constantly changing data, updating and optimizing the target control model using new data without retraining the entire model. This eliminates the need for repeated manual debugging for specific tasks, allowing the target control model to adapt to new tasks or environments. This ensures the model's output control parameters closely match the actual scenario, enhancing the accuracy of control parameters and providing reliable data support for the subsequent integrated operation of industrial equipment to meet diverse user needs. On the other hand, by using target control models trained in the cloud to predict and analyze control parameters, equipment control can be automated, and a large amount of data analysis and processing can be completed in a short time, thereby greatly improving control efficiency and accuracy, reducing manual control workload, and significantly reducing operating control costs.

[0078] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0079] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0080] Figure 1 A flowchart illustrating the edge-cloud collaborative industrial equipment control method provided in an embodiment of this application is shown.

[0081] Figure 2 This illustration shows one of the structural block diagrams of an industrial equipment control device based on edge-cloud collaboration provided in an embodiment of this application;

[0082] Figure 3 This is a second structural block diagram of an industrial equipment control device based on edge-cloud collaboration provided in an embodiment of this application. Detailed Implementation

[0083] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0084] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0085] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this application means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “attached” to another element, it can be directly connected or attached to the other element, or there may be intermediate elements. Furthermore, “connected” or “attached” as used herein can include wireless connections or wireless interconnections. The term “and / or” as used herein includes all or any unit and all combinations of one or more associated listed items.

[0086] Exemplary embodiments according to this application will now be described in more detail with reference to the accompanying drawings. However, these exemplary embodiments may be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein. It should be understood that these embodiments are provided so that the disclosure of this application is thorough and complete, and that the concept of these exemplary embodiments is fully conveyed to those skilled in the art.

[0087] This embodiment provides an industrial equipment control method based on edge-cloud collaboration, such as... Figure 1 As shown, and illustrated for application in the cloud and on the server side, the method includes:

[0088] Step 101: The edge device collects target data of different data types from industrial equipment according to the sampling period;

[0089] The target data, categorized into different types, includes: monitoring data (e.g., monitoring images, monitoring videos), production data, and / or equipment operation data (e.g., equipment operating status, current operating parameters). These different types of target data can originate from the same or different industrial equipment. Monitoring data and equipment operation data can be used for rapid fault detection in industrial equipment, allowing for timely control of the equipment to perform fault-related operations when a fault occurs. Production data can be used to assess whether the operational status of industrial equipment meets requirements, allowing for timely adjustment of equipment parameters when requirements are not met. The sampling period can be reasonably set according to the equipment's control precision, for example, 10 minutes, 1 hour, or 1 day.

[0090] Specifically, industrial equipment can refer to various devices used in industrial fields such as machine tools, painting equipment, generators, intelligent instruments, automated control equipment (PLC / DCS), material handling equipment, and communication and data exchange equipment (OPC / DA) for production, processing, manufacturing, transportation, and storage. Industrial equipment is equipped with different functional sensors to collect various target data. To ensure the reliability and stability of equipment operation, industrial equipment can be configured with high-precision sensors.

[0091] It should be noted that the edge device can acquire target data from the industrial equipment by sending data acquisition requests to the sensors in the industrial equipment according to the sampling period. Communication connections between the edge device and the industrial equipment being monitored include, but are not limited to: 5G, WiFi, I / O, analog signals, Modbus communication, Ethernet communication, Profinet communication, Profibus communication, CAN communication, and HART communication. Internal modules of the edge device can achieve data transmission via the MQTT (Message Queuing Telemetry Transport) protocol.

[0092] Furthermore, in one embodiment, after step 101, the industrial equipment control method based on edge-cloud collaboration further includes: filtering the target data at the edge.

[0093] In this embodiment, target data received from industrial equipment is filtered according to set conditions and rules, retaining only data that meets the criteria and filtering out irrelevant or redundant data. This ensures data quality and accuracy while reducing data volume, making data processing more efficient and intelligent.

[0094] Step 102: The edge device sends the target data collected in the current sampling period to the cloud;

[0095] It should be noted that the cloud is equipped with open data interfaces for different industrial data formats to receive target data from different data sources, so as to receive or store the target data reported by the edge.

[0096] Step 103: The cloud receives the target data collected in the current sampling period sent by the edge device;

[0097] Step 104: The cloud performs incremental learning on the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types;

[0098] The preset control model is a target control model trained based on the target data collected in the previous sampling period. It can be understood that if no target control model exists that has been trained based on the target data collected in the previous sampling period, the initial model will be used as the preset control model for continuous training.

[0099] In this embodiment, after acquiring new target data, the edge device sends it to the cloud for storage. The cloud uses the target control model trained on the target data from the previous sampling period as a foundation, retrains the model through incremental learning, and stores the trained target control model. This fully considers constantly changing data, updating and optimizing the target control model using new data without retraining the entire model. It eliminates the need for repeated manual debugging for specific tasks, enabling the target control model to adapt to the operational requirements of different industrial environments. While ensuring the model's output control parameters closely match the actual scenario and enhancing the accuracy of the control parameters, it also saves users the maintenance costs of model upgrades and rollbacks, meeting diverse user needs.

[0100] Furthermore, in one embodiment, after step 104, the industrial equipment control method based on edge-cloud collaboration further includes: preprocessing the target data, wherein the preprocessing includes: noise reduction processing (such as moving average processing, median filtering processing), cleaning processing, data transformation processing and / or tagging processing.

[0101] In this embodiment, noise or outliers can be removed from the target data through denoising processing to improve data accuracy and consistency. The target data can be cleaned and normalized through cleaning processing to fill in missing values ​​or remove duplicate values, improving the quality and usability of the target data. The target data can be transformed through data transformation processing to meet specific needs or extract more useful data. The target data can be classified or labeled through labeling processing for supervised learning or classification tasks, enabling subsequent data analysis and machine learning tasks to be performed more accurately.

[0102] In practical applications, step 104, which involves the cloud incrementally learning the preset control model based on the target data collected in the current sampling period, specifically includes the following steps:

[0103] Step 104-1: The cloud divides the target data collected in the current sampling period into training data and test data.

[0104] In this embodiment, after receiving the target data, the cloud can divide the target data into training data and test data. The training data is used to train the model, and the test data is used to test the trained model, thereby ensuring the consistency of the sources of the training data and test data and making the test results more accurate.

[0105] Step 104-2: The cloud platform trains the preset control model according to the data type and the training data to determine the candidate models for different data types.

[0106] In this embodiment, a preset control model corresponding to each data type is trained for the training data, and a candidate model for each data type is obtained. This provides models with different management strategies for different data formats, enabling different types of tasks to be processed using suitable models, further ensuring the accuracy of data analysis and processing. Simultaneously, classifying and training different target data by data type reduces data complexity and helps improve model training efficiency.

[0107] Step 104-3: Obtain the theoretical control parameters corresponding to the test data from the cloud;

[0108] Step 104-4: The cloud inputs the test data into the candidate model according to the data type to obtain the test control parameters;

[0109] Step 104-5: If the difference between the test control parameters and the theoretical control parameters is within a preset range, the cloud will determine the candidate model as the target control model.

[0110] In this embodiment, the candidate model is tested using test data to determine if it meets the required prediction accuracy. If there is a significant difference between the test control parameters obtained from the candidate model and the theoretical control parameters corresponding to the test data, it indicates that the candidate model's accuracy is insufficient, and retraining is required. Conversely, if the difference between the test control parameters and the theoretical control parameters corresponding to the test data is within a preset range, i.e., the difference is small, it can be determined that the candidate model has reached the required accuracy. In this case, the candidate model can be used as the target control model for predicting control parameters. This ensures the accuracy of industrial equipment control when using the target control model.

[0111] Furthermore, in one embodiment, if the amount of collected data is small, it may not meet the data requirements for model training. In this case, the target data can be expanded in the following way:

[0112] Method 1: The cloud expands the target data according to the preset step size corresponding to different data types to obtain the updated target data.

[0113] The preset step size is determined based on the range distribution characteristics of different data types.

[0114] In this embodiment, the feasible range of data expansion for different data types is determined by utilizing the range distribution characteristics of different data types, and then a preset step size is determined based on the feasible range. Using the target data as the initial value, the data is offset according to the preset step size to obtain more data similar to the initial value. This expands the target data, providing more data support for subsequent model training.

[0115] Method 2: The cloud inputs the target data into the generative adversarial network corresponding to different data types to obtain the updated target data.

[0116] Generative Adversarial Networks (GANs) are used to generate realistic new samples that are similar to real data.

[0117] In this embodiment, the target data is used as the real sample. The distribution of the real sample is determined by the generative adversarial network, and a fake sample similar to the real sample is generated based on this. The real sample and the fake sample are used as the expanded target data to provide more data support for subsequent model training.

[0118] Step 105: The cloud sends the target control model to the edge device;

[0119] Step 106: The edge device receives the target control model sent from the cloud;

[0120] Step 107: The edge device inputs the target data collected in the current sampling period into the target control model according to the data type to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types;

[0121] In this embodiment, the target control model trained in the cloud is deployed at the edge. When industrial equipment needs to be controlled, the edge device can call the target control model to process, diagnose, predict, and analyze the real-time target data collected from the industrial equipment, and then output the control parameters of the actuators. This deploys the control parameter analysis and prediction tasks closer to the industrial equipment at the edge, reducing equipment data management and analysis latency, enabling control commands to be sent to the industrial equipment in a timely manner, and improving the user experience.

[0122] For example, if the robotic arm of a machine tool malfunctions and is in a stopped state, the edge control model determines new control parameters related to stopping the robotic arm and the corresponding conveyor belt.

[0123] Step 108: The edge device sends control commands to the industrial equipment.

[0124] The control instructions include control parameters for actuators in industrial equipment corresponding to different data types.

[0125] It is worth mentioning that the communication connection between the edge device and the actuators of industrial equipment includes, but is not limited to: 5G, WiFi, I / O, analog signals, Modbus communication, and Ethernet communication. Industrial equipment can subscribe to target data provided by the edge device via MQTT.

[0126] This application provides an edge-cloud collaborative industrial equipment control method. During operation, industrial equipment continuously generates new target data, including monitoring data, production data, and equipment operation data. The method utilizes the edge to periodically collect target data of different data types generated by the industrial equipment and sends it to the cloud for incremental learning of the target control model. The cloud then returns the incrementally learned target control model to the edge. Thus, the edge can input the target data collected in the current sampling period into the updated target control model according to different data types to predict the control parameters of the actuators in the industrial equipment corresponding to different data types. Finally, the edge sends control commands to the industrial equipment, causing the actuators to act according to the control parameters output by the model, thereby achieving precise control of the industrial equipment. Through the above technical solutions, on the one hand, highly complex computations such as data management, analysis, prediction, and model training are deployed in the cloud and at the edge, realizing a code development model of cloud compilation / edge execution. This significantly reduces the hardware cost of industrial equipment. Moreover, through collaborative work between the cloud and the edge, the edge can perform calculations and processing closer to the industrial equipment, reducing data transmission latency and enabling control commands to be issued to the industrial equipment in a timely manner. Simultaneously, the edge only transmits necessary results to the cloud server, reducing the amount of data transmitted over the network, providing more efficient, flexible, and reliable computing and services, and improving the scalability of industrial applications. On the other hand, it fully considers constantly changing data, updating and optimizing the target control model using new data without retraining the entire model. This eliminates the need for repeated manual debugging for specific tasks, allowing the target control model to adapt to new tasks or environments. This ensures the model's output control parameters closely match the actual scenario, enhancing the accuracy of control parameters and providing reliable data support for the subsequent integrated operation of industrial equipment to meet diverse user needs. On the other hand, by using target control models trained in the cloud to predict and analyze control parameters, equipment control can be automated, and a large amount of data analysis and processing can be completed in a short time, thereby greatly improving control efficiency and accuracy, reducing manual control workload, and significantly reducing operating control costs.

[0127] For specific examples, edge-cloud collaborative industrial equipment control methods include:

[0128] 1. Monitor the target data (such as industrial production data, equipment status data, and video surveillance data) generated by the controlled object of the industrial equipment through the sensors of the industrial equipment. The controlled object can be a structure (device) or a group thereof, which can be univariate or multivariate, have time delay or no time delay, or have strong interference, etc.

[0129] 2. Target data from industrial equipment is collected at the edge and then filtered;

[0130] 3. The edge device uploads the collected target data to the cloud database for storage;

[0131] 4. After preprocessing the target data in the database in the cloud, the candidate modules are obtained through incremental learning.

[0132] 5. The cloud uses an appropriate evaluation function to evaluate the candidate modules, and the candidate module with the best evaluation index is sent to the edge as the intelligent industrial model;

[0133] 6. The edge device calls the intelligent industrial model (target control model) to process, diagnose, predict and analyze the collected real-time target data, and then outputs control commands to the industrial equipment;

[0134] 7. Industrial equipment uses a D / A converter to convert analog signals to drive actuators, thereby controlling the actuators and achieving the purpose of controlling the controlled object. Step 1 is repeated. The actuators can be solenoid valves, servo motors, pneumatic control valves, etc.

[0135] Through the above embodiments, models used for application and inference are deployed at the edge, reducing application latency in device data management and analysis, and improving user experience. Furthermore, data is uploaded to the cloud for continuous training and optimization, enabling intelligent industrial systems with self-analysis, self-learning, self-diagnosis, self-recovery, self-management, and self-optimization capabilities. This improves the scalability of industrial systems to adapt to operational needs in different industrial environments, reduces system complexity, saves user costs for model upgrades and rollbacks, conserves service resources, and enhances the development efficiency and scalability of industrial model systems.

[0136] It should be noted that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0137] Furthermore, such as Figure 2 As shown, as a specific implementation of the above-mentioned industrial equipment control method based on edge-cloud collaboration, this application provides an industrial equipment control device 200 based on edge-cloud collaboration, which includes: a data acquisition module 201, a communication module 202, and a control module 203.

[0138] The data acquisition module 201 is used to collect target data of different types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data.

[0139] Communication module 202 is used to send the target data collected in the current sampling period to the cloud, so that the cloud can incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period; and to receive the target control model sent by the cloud.

[0140] Control module 203 is used to input the target data collected in the current sampling period into the target control model according to the data type, so as to obtain the control parameters of the actuator in the industrial equipment corresponding to different data types;

[0141] The communication module 202 is also used to send control commands to industrial equipment to control the actuators to move according to the control parameters in the control commands.

[0142] Furthermore, the edge-cloud collaborative industrial equipment control device 200 also includes a filtering module (not shown in the figure), which is used to filter the target data.

[0143] Furthermore, such as Figure 3 As shown, as a specific implementation of the above-mentioned industrial equipment control method based on edge-cloud collaboration, this application provides an industrial equipment control device 300 based on edge-cloud collaboration, which includes: a communication module 301 and a training module 302.

[0144] The communication module 301 is used to receive target data of different data types collected in the current sampling period sent by the edge terminal. The target data includes: monitoring data, production data and / or equipment operation data.

[0145] Training module 302 is used to incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is the target control model trained based on the target data collected in the previous sampling period.

[0146] The communication module 301 is also used to send the target control model to the edge terminal, so that the edge terminal can determine the control parameters of the actuators in the industrial equipment corresponding to different data types through the target control model and the target data collected in the current sampling period.

[0147] Furthermore, the edge-cloud collaborative industrial equipment control device 300 also includes: a classification module (not shown in the figure), which is used to divide the target data collected in the current sampling period into training data and test data; a training module 302, which is specifically used to train the preset control model according to the data type and the training data to determine the candidate models of different data types; a testing module (not shown in the figure), which is used to obtain the theoretical control parameters corresponding to the test data; and to input the test data into the candidate model according to the data type to obtain the test control parameters; and if the difference between the test control parameters and the theoretical control parameters is within a preset range, the candidate model is determined as the target control model.

[0148] Furthermore, the edge-cloud collaborative industrial equipment control device 300 also includes: an expansion module (not shown in the figure), which is used to expand the target data according to the preset step size corresponding to different data types to obtain updated target data, wherein the preset step size is determined according to the range distribution characteristics of different data types; or, inputting the target data into the generative adversarial network corresponding to different data types to obtain updated target data.

[0149] Furthermore, the edge-cloud collaborative industrial equipment control device 300 also includes a preprocessing module (not shown in the figure), which is used to preprocess the target data, including noise reduction, cleaning, data conversion and / or tagging.

[0150] Specific limitations regarding edge-cloud collaborative industrial equipment control devices can be found in the limitations of edge-cloud collaborative industrial equipment control methods described above, and will not be repeated here. Each module in the aforementioned edge-cloud collaborative industrial equipment control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0151] Based on the above, Figure 1 Accordingly, embodiments of this application also provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. Figure 1 The method for controlling industrial equipment based on edge-cloud collaboration is shown.

[0152] Based on this understanding, the technical solution of this application can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive), and includes several instructions to cause a computer device (such as a personal computer, server, or network device) to execute the methods described in the various implementation scenarios of this application.

[0153] Based on the above, Figure 1 The method shown, and Figure 2 and Figure 3 To achieve the above objectives, the virtual device embodiment shown in this application also provides an edge-cloud collaborative industrial equipment control system, which includes: industrial equipment, edge device, and cloud device.

[0154] Specifically, the industrial equipment generates target data; the edge device collects target data of different data types from the industrial equipment according to a sampling period, wherein the target data includes: monitoring data, production data, and / or equipment operation data; and sends the target data collected in the current sampling period to the cloud; the cloud receives the target data collected in the current sampling period sent by the edge device; and incrementally learns a preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period; and if the target control model meets preset conditions, it sends the target control model to the edge device; the edge device also receives the target control model sent by the cloud; and inputs the target data collected in the current sampling period into the target control model according to the data type to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types; and sends control commands to the industrial equipment to control the actuators to move according to the control parameters in the control commands.

[0155] Furthermore, the cloud specifically divides the target data collected in the current sampling period into training data and test data; according to the data type, it trains the preset control model based on the training data to determine candidate models of different data types; it obtains the theoretical control parameters corresponding to the test data; it inputs the test data into the candidate model according to the data type to obtain test control parameters; if the difference between the test control parameters and the theoretical control parameters is within a preset range, the candidate model is determined as the target control model.

[0156] Those skilled in the art will understand that the edge-cloud collaborative industrial equipment control system structure provided in this embodiment does not constitute a limitation on the edge-cloud collaborative industrial equipment control system, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0157] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platform, or by hardware to collect target data of different data types from industrial equipment according to the sampling period. The target data includes: monitoring data, production data and / or equipment operation data; sending the target data collected in the current sampling period to the cloud, so that the cloud can incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period; receiving the target control model sent by the cloud; inputting the target data collected in the current sampling period into the target control model according to the data type to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types; sending control commands to the industrial equipment to control the actuators to act according to the control parameters in the control commands. Alternatively, the system receives target data of different data types collected in the current sampling period from the edge terminal. The target data includes monitoring data, production data, and / or equipment operation data. Based on the target data collected in the current sampling period, the system incrementally learns a preset control model to obtain target control models for different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period. The target control model is then sent to the edge terminal, enabling the edge terminal to determine the control parameters of the actuators in the industrial equipment corresponding to different data types using the target control model and the target data collected in the current sampling period. In this embodiment, on the one hand, highly complex computations such as data management, analysis, prediction, and model training are deployed in the cloud and at the edge, realizing a code development model of cloud compilation / edge execution. This significantly reduces the hardware cost of industrial equipment. Furthermore, through collaborative work between the cloud and the edge terminal, the edge terminal can perform calculations and processing closer to the industrial equipment, reducing data transmission latency and enabling timely delivery of control commands to the industrial equipment. Simultaneously, the edge terminal only transmits necessary results to the cloud server, reducing the amount of data transmitted over the network, providing more efficient, flexible, and reliable computing and services, and improving the scalability of industrial applications. On the other hand, it fully considers the constantly changing data, and uses new data to update and optimize the target control model without retraining the entire model. It eliminates the need for repeated manual debugging for specific tasks, enabling the target control model to adapt to new tasks or environments. This ensures that the model's output control parameters are consistent with the actual scenario, enhances the accuracy of the control parameters, and provides reliable data support for the subsequent integrated operation of industrial equipment to meet different user needs.On the other hand, by using target control models trained in the cloud to predict and analyze control parameters, equipment control can be automated, and a large amount of data analysis and processing can be completed in a short time, thereby greatly improving control efficiency and accuracy, reducing manual control workload, and significantly reducing operating control costs.

[0158] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.

[0159] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.

Claims

1. An industrial equipment control method based on edge-cloud collaboration, characterized in that, Applicable to edge ends, the method includes: Target data of different data types from the industrial equipment are collected according to the sampling period, wherein the target data includes: monitoring data, production data and / or equipment operation data; The target data collected in the current sampling period is sent to the cloud so that the cloud can incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period. Receive the target control model sent from the cloud; According to the data type, the target data collected in the current sampling period is input into the target control model to obtain the control parameters of the actuators in the industrial equipment corresponding to different data types; Control commands are sent to the industrial equipment to control the actuator to operate according to the control parameters in the control commands.

2. The industrial equipment control method based on edge-cloud collaboration according to claim 1, characterized in that, After collecting target data of different data types from the industrial equipment according to the sampling period, the method further includes: The target data is then filtered.

3. An industrial equipment control method based on edge-cloud collaboration, characterized in that, Applicable to the cloud, the method includes: Receive target data of different data types collected in the current sampling period from the edge terminal, wherein the target data includes: monitoring data, production data and / or equipment operation data; Incremental learning is performed on the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period. The target control model is sent to the edge device so that the edge device can determine the control parameters of the actuators in the industrial equipment corresponding to different data types using the target control model and the target data collected in the current sampling period.

4. The industrial equipment control method based on edge-cloud collaboration according to claim 3, characterized in that, The incremental learning of the preset control model based on the target data collected in the current sampling period includes: The target data collected in the current sampling period is divided into training data and test data; According to the data type, the preset control model is trained according to the training data to determine candidate models for different data types; Obtain the theoretical control parameters corresponding to the test data; The test data is input into the candidate model according to the data type described above to obtain the test control parameters; If the difference between the test control parameters and the theoretical control parameters is within a preset range, the candidate model is determined as the target control model.

5. The industrial equipment control method based on edge-cloud collaboration according to claim 3 or 4, characterized in that, After receiving target data of different data types collected in the current sampling period from the receiving edge terminal, the method further includes: The target data is preprocessed, wherein the preprocessing includes: noise reduction, cleaning, data transformation and / or labeling.

6. An industrial equipment control method based on edge-cloud collaboration, characterized in that, The method includes: The edge device collects target data of different data types from the industrial equipment according to the sampling period, wherein the target data includes: monitoring data, production data and / or equipment operation data; The edge device sends the target data collected in the current sampling period to the cloud; The cloud receives the target data collected during the current sampling period from the edge terminal; The cloud platform performs incremental learning on the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types. The preset control model is a target control model trained based on the target data collected in the previous sampling period. The cloud sends the target control model to the edge device; The edge device receives the target control model sent from the cloud; The edge terminal inputs the target data collected in the current sampling period into the target control model according to the data type, so as to obtain the control parameters of the actuator in the industrial equipment corresponding to different data types; The edge device sends the control command to the industrial equipment to control the actuator to operate according to the control parameters in the control command.

7. An industrial equipment control device based on edge-cloud collaboration, characterized in that, The device includes: The data acquisition module is used to collect target data of different data types from the industrial equipment according to the sampling period, wherein the target data includes: monitoring data, production data and / or equipment operation data; A communication module is used to send the target data collected in the current sampling period to the cloud, so that the cloud can incrementally learn a preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period; and, Receive the target control model sent from the cloud; The control module is used to input the target data collected in the current sampling period into the target control model according to the data type, so as to obtain the control parameters of the actuator in the industrial equipment corresponding to different data types; The communication module is also used to send the control command to the industrial equipment to control the actuator to operate according to the control parameters in the control command.

8. An industrial equipment control device based on edge-cloud collaboration, characterized in that, The device includes: The communication module is used to receive target data of different data types collected in the current sampling period from the edge terminal, wherein the target data includes: monitoring data, production data and / or equipment operation data; The training module is used to incrementally learn the preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period. The communication module is further configured to send the target control model to the edge terminal, so that the edge terminal can determine the control parameters of the actuators in the industrial equipment corresponding to different data types through the target control model and the target data collected in the current sampling period.

9. An industrial equipment control system based on edge-cloud collaboration, comprising: Industrial equipment, edge computing, and the cloud; The industrial equipment is used to generate the target data; The edge device is used to collect target data of different data types from the industrial equipment according to a sampling period, wherein the target data includes: monitoring data, production data and / or equipment operation data; and to send the target data collected in the current sampling period to the cloud. The cloud is configured to receive the target data collected in the current sampling period sent by the edge terminal; and to perform incremental learning on a preset control model based on the target data collected in the current sampling period to obtain target control models of different data types, wherein the preset control model is a target control model trained based on the target data collected in the previous sampling period; and if the target control model meets preset conditions, to send the target control model to the edge terminal. The edge device is also used to receive the target control model sent by the cloud; and to input the target data collected in the current sampling period into the target control model according to the data type to obtain the control parameters of the actuator in the industrial equipment corresponding to different data types; and to send the control command to the industrial equipment to control the actuator to act according to the control parameters in the control command.

10. A readable storage medium having a program or instructions stored thereon, characterized in that, When the program or instructions are executed by the processor, they implement the steps of the industrial equipment control method based on edge-cloud collaboration as described in any one of claims 1 to 5.

Images