Power grid edge intelligent fire point detection container management method based on KubeEdge

Abstract

The invention provides a power grid edge intelligent fire point detection container management method based on KubeEdge, and the method comprises the steps: deploying an edge container management platform based on KubeEdge, and the edge container management platform comprises a control end and batch edge devices; the control end carries out storage, packaging and containerization processing on the trained fire point detection model to obtain a fire point detection container and a container mirror image of the fire point detection container; each edge device creates a configuration file for a fire point detection container so as to obtain the required fire point detection container from the container mirror image of the control end, and then fire point detection is performed by using the corresponding fire point detection container; and the control end monitors the real-time operation state of the fire point detection container in each edge device to obtain a real-time fire point detection result, and dynamically manages the fire point detection container in each edge device when the operation state of the edge device is changed.

Description

Technical Field

[0001] The present invention relates to the field of computer network technology, and in particular to a KubeEdge-based power grid edge intelligent fire point detection container management method. Background Art

[0002] The current fire point detection technology of power grid in remote and mountainous areas is not mature enough. It only relies on real-time observation by employees using surveillance cameras, which has high labor costs and low efficiency. At present, some pilot areas have carried out intelligent edge fire point detection. However, due to the large number and wide distribution of edge devices, it will cost a lot of manpower and material resources to deploy the fire point detection solution to each edge device based on manual deployment one by one, which is in urgent need of development and reform. In recent years, with the improvement of the intelligence level of edge devices, the AI ​​computing power of edge devices has become stronger and stronger, and it can support sufficient edge real-time intelligent detection services. For the deployment of edge intelligent fire point detection algorithms, the current more advanced solutions are deployed in the supervision center, relying on centralized large-scale computing servers for real-time fire point detection, which greatly improves the efficiency of the control center personnel. However, considering that in mountainous and marginal areas, network fluctuations and interruptions may occur frequently under certain special circumstances, the centralized solution greatly reduces the reliability of the entire business. At the same time, with the iterative update of different fire point algorithms, edge devices should be able to uniformly deploy new fire point algorithm containers to provide consistent fire point detection services. In the face of the needs of edge devices of the power grid, lightweight container software represented by Docker can better meet the needs of edge devices. At the same time, its containerized services can be stored and backed up in the content server in the form of images. By simplifying the container environment, it can achieve general and universal deployment on edge smart devices with different architectures, and support the deployment and update of real-time fire point detection algorithms for power grids.

[0003] There are many fire point detection technologies in the prior art, but these technologies all have some problems. For example, the Chinese invention patent with the authorization announcement number CN112580549B discloses a fire point detection method, device and storage medium, which includes: obtaining multi-channel historical remote sensing image data; constructing a fire point detection model based on a convolutional neural network, and using the historical remote sensing image data to train the model, inputting real-time image data into the trained fire point detection model, and extracting the feature information of each channel for fire point detection. However, this method lacks the management and distributed deployment capabilities for containerized resources in the single detection capability of the fire point detection model. This makes the scheduling and allocation efficiency of resources relatively low in a large-scale power grid environment, and the edge nodes cannot be fully utilized for distributed fire point detection. No distributed resource management platform has been introduced, so it is impossible to dynamically allocate resources according to business needs, and there is a lack of intelligence in adaptive resource regulation. At the same time, if the fire point detection model or node fails, it may affect the stability and continuity of the overall detection.

[0004] For example, the Chinese invention patent application with application publication number CN118608982A discloses a deep learning fire point identification algorithm and system based on multi-source satellite data, in which the control center connects the data acquisition, data processing, identification analysis and intelligent control modules. The satellite data is acquired through the data acquisition module, and the data processing module merges and verifies to obtain a standard band map; the identification and analysis module groups and divides the band map, extracts the mean point, and generates a comprehensive channel map; the intelligent control module extracts the edge of the comprehensive map, obtains the target remote sensing image, and performs security verification through the information query library to obtain the warning coordinates and warning time, thereby improving the accuracy of fire point identification and reducing the false alarm rate. However, this method mainly relies on a centralized control center for fire point identification. This model is difficult to achieve resource coordination and adaptive management of edge nodes in large-scale power grid scenarios, limiting the scalability and deployment flexibility of the system. This method is difficult to flexibly regulate according to the load and business needs of different nodes, limiting the adaptability of the fire point identification system and the accuracy of resource allocation.

[0005] For example, the Chinese invention patent with authorization announcement number CN117911887B discloses a forest fire smoke recognition and fire point detection method and system based on Sentinel-2. It proposes to extract smoke based on the constructed Sentinel-2 satellite image forest fire smoke segmentation database, and use the spectral characteristics and full-scale global-local perception mode of the smoke recognition model to realize smoke extraction; based on the spatiotemporal characteristics of the image area, the smoke recognition-assisted time-corrected spatial context algorithm is used to determine the real fire point. However, this method relies on smoke segmentation and spatiotemporal characteristics for fire point recognition. Although it has certain advantages in forest fire scenes, it may not be sufficient to accurately detect different types of fire points in dynamic and diverse power grid environments. In addition, due to the lack of edge collaboration support, the fire point recognition speed may be low, especially in data-intensive environments.

[0006] For example, a Chinese invention patent application with application publication number CN118802519A discloses a method for rapid deployment of edge computing services. The method relies on a cloud computing service center to establish a public service warehouse to achieve online management and download distribution of services. An edge cloud service deployment list is generated based on application requirements, and a priority model is established based on service activity and dependency, so that urgent services are deployed first. Multi-threading and double-buffer queue processing technology are used to accelerate the deployment process of edge cloud services, thereby simplifying operations, shortening deployment time, and improving speed and response efficiency. However, this method focuses on the rapid deployment of edge computing services, setting deployment priorities based on service activity and dependency, but lacks real-time resource management and dynamic adjustment based on business needs. Summary of the invention

[0007] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0008] To this end, the first purpose of the present invention is to propose a KubeEdge-based power grid edge intelligent fire point detection container management method to solve the complex deployment of edge intelligent fire point detection services and the difficult problem of edge fire point detection resource management.

[0009] The second object of the present invention is to propose a KubeEdge-based power grid edge intelligent fire point detection container management system.

[0010] A third objective of the present invention is to provide an electronic device.

[0011] A fourth object of the present invention is to provide a computer-readable storage medium.

[0012] To achieve the above-mentioned object, the first aspect of the present invention proposes a KubeEdge-based power grid edge intelligent fire point detection container management method, comprising:

[0013] Deploy a KubeEdge-based edge container management platform, which includes a control terminal and multiple edge devices;

[0014] The control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container;

[0015] Each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then uses the corresponding fire point detection container to perform fire point detection respectively;

[0016] The control end monitors the real-time operating status of the fire point detection container in each edge device, so as to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal.

[0017] In the method of the first aspect of the present invention, the deployment is based on the edge container management platform of KubeEdge, including: deploying Kubernetes on the control server in the control end, installing the Cloudcore module of KubeEdge in the control server, and deploying the Edgecore module of KubeEdge on each edge device, thereby obtaining the edge container management platform.

[0018] In the method of the first aspect of the present invention, the control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container, including: the control end obtains the trained fire point detection model, and saves and packages the trained fire point detection model; for the packaged fire point detection model, using Docker to build a fire point detection container and a container image of the fire point detection container.

[0019] In the method of the first aspect of the present invention, the saved fire point detection model is encapsulated using a Django server.

[0020] In the method of the first aspect of the present invention, the fire point detection model adopts a lightweight pruning fire point algorithm based on Yolo V8.

[0021] In the method of the first aspect of the present invention, the abnormal real-time operating status refers to a container failure or insufficient resources.

[0022] To achieve the above-mentioned object, the second aspect of the present invention proposes a KubeEdge-based grid edge intelligent fire point detection container management system, wherein the system is deployed with KubeEdge, and the system includes a control terminal and multiple edge devices;

[0023] The control end is used to save, encapsulate, and containerize the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container; it is also used to monitor the real-time operating status of the fire point detection container in each edge device to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal.

[0024] Each edge device is used to create a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then use the corresponding fire point detection container to perform fire point detection respectively.

[0025] In the system of the second aspect of the present invention, the control end uses the Django server to encapsulate the saved fire point detection model, and uses Docker to build the encapsulated fire point detection model to obtain a fire point detection container and a container image of the fire point detection container.

[0026] To achieve the above-mentioned purpose, the third aspect of the present invention proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method proposed in the first aspect of the present invention.

[0027] To achieve the above-mentioned purpose, the fourth aspect of the present invention proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method proposed in the first aspect of the present invention.

[0028] The invention provides a KubeEdge-based power grid edge intelligent fire point detection container management method, system, electronic device and storage medium. By deploying an edge container management platform based on KubeEdge, the edge container management platform includes a control end and multiple edge devices; the control end saves, encapsulates and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container; each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then uses the corresponding fire point detection container to perform fire point detection respectively; the control end monitors the real-time operating status of the fire point detection container in each edge device to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal. In this case, by deploying an edge container management platform based on KubeEdge, the control end is used to obtain the fire point detection container, and the real-time operating status of the fire point detection container in each edge device is monitored to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal. Therefore, KubeEdge is used to manage the computing resources of the edge devices of the power grid, which can efficiently schedule and manage the resources and status of the edge devices in real time. In addition, each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end for fire point detection. Therefore, the edge fire point detection application encapsulated in the container can better adapt to the heterogeneous characteristics of the edge fire point detection hardware. Between the fire point detection intelligent devices of the same operating system, the service can be seamlessly migrated and shared, avoiding the difficulties caused by multiple deployments of the basic environment and repeated deployments during algorithm upgrades. Therefore, the present invention solves the problem of complex deployment of edge intelligent fire point detection services and the difficulty of managing edge fire point detection resources.

[0029] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0031] Figure 1 A schematic diagram of a flow chart of a KubeEdge-based power grid edge intelligent fire point detection container management method provided in an embodiment of the present invention;

[0032] Figure 2 A diagram of the platform architecture provided by an embodiment of the present invention;

[0033] Figure 3 A schematic diagram of a specific process of a KubeEdge-based power grid edge intelligent fire point detection container management method provided in an embodiment of the present invention;

[0034] Figure 4 A block diagram of a KubeEdge-based grid edge intelligent fire point detection container management system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0036] The following describes a KubeEdge-based grid edge intelligent fire point detection container management method and system according to an embodiment of the present invention with reference to the accompanying drawings.

[0037] The embodiment of the present invention provides a KubeEdge-based power grid edge intelligent fire point detection container management method to solve the complex deployment of edge intelligent fire point detection services and the difficult problem of edge fire point detection resource management.

[0038] Figure 1 A flow chart of a KubeEdge-based grid edge intelligent fire point detection container management method provided in an embodiment of the present invention.

[0039] like Figure 1 As shown, the KubeEdge-based power grid edge intelligent fire point detection container management method includes the following steps:

[0040] Step S101, deploy an edge container management platform based on KubeEdge, the edge container management platform includes a control end and multiple edge devices.

[0041] In step S101, a KubeEdge-based edge container management platform is deployed, including: deploying Kubernetes on the control server in the control end, installing the Cloudcore module of KubeEdge in the control server, and deploying the Edgecore module of KubeEdge on each edge device, thereby obtaining an edge container management platform.

[0042] In step S101, the control end may be a control plane in the cloud, and the control plane includes a management and control server. The edge device is an edge intelligent fire point detection device. Each edge device may be, but is not limited to, a satellite edge intelligent fire point detection device.

[0043] In some embodiments, Figure 2The platform architecture diagram provided by the embodiment of the present invention. The platform architecture is a satellite edge intelligent resource management platform architecture. The satellite edge intelligent resource management platform architecture is an edge container management platform deployed with KubeEdge. Among them, the control plane is the control end, and the Cloudcore module of KubeEdge is installed in the management and control server. The resource plane is each edge intelligent fire point detection device. The Edgecore module of KubeEdge is deployed in each edge intelligent fire point detection device. Among them, there is mirror storage in the management and control server, and the management and control server can perform status monitoring, control management and resource management on each edge intelligent fire point detection device to realize centralized management of edge intelligent fire point detection devices.

[0044] Step S102: The control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container.

[0045] In step S102, the fire point detection model adopts a lightweight pruning fire point algorithm based on Yolo V8.

[0046] In step S102, the control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container, including: the control end obtains the trained fire point detection model, and saves and packages the trained fire point detection model; for the packaged fire point detection model, Docker is used to build a fire point detection container and a container image of the fire point detection container.

[0047] In step S102, the saved fire point detection model is encapsulated using the Django server.

[0048] In step S103, each edge device creates a configuration file for a fire point detection container to obtain a required fire point detection container from the container image of the control end, and then uses the corresponding fire point detection container to perform fire point detection respectively.

[0049] In step S103, each edge device creates a configuration file for the fire point detection container, and each edge device pulls the required fire point detection container from the container image of the fire point detection container on the control end based on the created configuration file. After each edge device obtains the required fire point detection container, it starts to perform fire point detection, wherein the fire point detection container outputs the detection result of whether there is a fire point or no fire point. At this time, the real-time operation status of the fire point detection container is normal, which is a regular event.

[0050] In step S104, the control end monitors the real-time operating status of the fire point detection container in each edge device, so as to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal.

[0051] In step S104, the abnormal real-time operation status refers to a container failure or insufficient resources.

[0052] Specifically, the control end monitors the real-time operating status of the fire point detection container in each edge device. When the pushed fault information is detected, it indicates that the corresponding fire point detection container has a container failure, and the control end re-pushes a new fire point detection container to the corresponding edge device; when insufficient resources are detected, the control end adjusts the resource capacity of the edge device.

[0053] Figure 3 A schematic diagram of a specific process of a KubeEdge-based grid edge intelligent fire point detection container management method provided in an embodiment of the present invention. Figure 3 As shown in the figure, the KubeEdge-based power grid edge intelligent fire point detection container management method specifically includes:

[0054] 1) Deploy KubeEdge and build a cluster management platform. The cluster management platform is the edge container management platform on which KubeEdge is deployed. Deploy Kubernetes on the control server of the control end of the edge container management platform, and install the Cloudcore module of KubeEdge in the control server. The Cloudcore module is the core component of KubeEdge, and serves as the control interface to manage the communication of edge intelligent fire point detection equipment. Deploy the Edgecore module of KubeEdge on the edge intelligent terminal (i.e., edge intelligent fire point detection equipment). The Edgecore module is responsible for the intelligent fire point detection container application on the edge node (i.e., edge intelligent fire point detection equipment). The Cloudcore module can be used to manage edge nodes and realize centralized management of edge intelligent fire point detection equipment.

[0055] 2) Store the weights and parameters of the fire point detection model. In the control end of the edge container management platform, design and train the fire point detection model. The fire point detection model can use the lightweight pruning fire point algorithm based on Yolo V8. After the training is completed, save the trained parameters and weights of the fire point detection model, and save the parameters and weights of the model in the original container in the form of files to save the fire point detection model. These parameter and weight files will be distributed to each edge device in the form of container files along with the fire point detection container to facilitate the subsequent deployment and reasoning of the fire point detection model.

[0056] 3) Use Django server to encapsulate the fire point detection model. Use Django server to encapsulate the saved fire point detection model as a service with an access interface, thereby realizing the encapsulation of the fire point detection model. Django server provides a web interface based on HTTP protocol. The subsequent control server can interact with the fire point detection model in each edge device through HTTP service to obtain the current fire point detection status and receive the fire point alarm signal.

[0057] 4) Use Docker tools to build containers and images. In order to facilitate the deployment and management of the power grid edge intelligent fire point detection service, the management and control server uses Docker to process the encapsulated fire point detection model, build a fire point detection container and create an image of the fire point detection container (i.e., container image). In the process of building the image, the environmental dependencies, basic files and weights required for the fire point detection model are placed in a unified file shelf, and containerized construction is performed using Docker. This ensures that the deployment and management of the fire point detection container are simple and consistent in different edge devices. After building the container image, back up the image to the management server so that each edge device can automatically pull the image of the fire point detection container and deploy it. The image uploaded to the management server can provide full-time backup for the edge intelligent terminal. When an error occurs in the operation of the edge intelligent terminal, the image can be pulled immediately.

[0058] 5) Create a configuration file for resource configuration. The edge intelligent terminal creates a configuration file (yaml script) for the fire point detection container, which specifies the required image and adaptively allocates memory and computing resources according to the capacity of the edge intelligent terminal based on the different precision resource requirements of the fire point detection service. Through the configuration file, you can define the deployment configuration of the fire point detection application, including resource restrictions, number of replicas, etc. This ensures that the fire point detection service can obtain appropriate resources to meet its operation requirements.

[0059] 6) Push the fire point container and start fire point detection. The edge intelligent terminal pulls the fire point detection container from the image of the control end based on the configuration file. The edge intelligent terminal starts fire point detection based on the fire point detection model in the obtained fire point detection container.

[0060] 7) Edge intelligent terminals continuously perform fire point detection.

[0061] 8) Determine the edge node event trigger. Specifically, use Cloudcore of the control server as the control middleware to manage the edge nodes. Through the various functions of Kubernetes, the real-time running status of the fire point detection container in the edge device can be monitored. If a regular event is detected (such as the detection result of a fire point or no fire point), no processing is performed, and the edge intelligent terminal continues to perform fire point detection; if insufficient resources are detected, the container resource amount (ie, resource capacity) can be dynamically adjusted according to the real-time load of the fire point detection model, thereby realizing dynamic management of the edge fire point detection container.

[0062] 9) When a fire point detection container in the edge device is monitored to fail, the management and control server can restart or reconfigure the fire point detection container based on the detected container failure information. This ensures the continuous availability of the service, reduces the fire point detection interruption time caused by failure and disrepair, and reduces fire point events caused by neglect of maintenance.

[0063] The method of the present invention is applicable to technical fields such as computer networks, graphics and image processing, computer performance improvement, encryption and decryption, and Chinese character input. Specifically, Docker is used to process edge fire point detection applications (i.e., fire point detection models), and edge fire point related software (such as models and files) are all deployed through Docker services, and all Linux system edge intelligent devices under the general Arm architecture can run. At the same time, KubeEdge is used to manage the computing resources of power grid edge intelligent devices, which can efficiently schedule and manage the resources and status of edge fire point detection devices in real time, thereby improving the continuous availability of services. Edge fire point detection applications encapsulated in containers can better adapt to the heterogeneous characteristics of edge fire point detection hardware. Between fire point detection intelligent devices with the same operating system, services can be seamlessly migrated and shared, avoiding the difficulties caused by multiple deployments of the basic environment and repeated deployments during algorithm upgrades. The fire point detection container management platform built with KubeEdge is used for distributed resource management, and the distributed edge intelligent fire point detection devices are organized to form a distributed detection network, which is convenient for adapting technologies such as federated learning, thereby enabling the ability of edge fire point detection to be mutually enhanced. The method of the present invention uses an edge fire point detection model, based on the Yolo V8 algorithm, to design a lightweight pruning fire point algorithm and encapsulate it into a fire point detection container and test it. The results show that the system can effectively complete the deployment of the edge fire point detection container, and the management and control nodes (i.e., management and control servers) can also effectively detect and schedule the application resources of the edge fire point detection container. At the same time, when an edge node fails, the management and control node can detect and report it automatically in time. From the test results, after adopting this method, the problem of complex deployment of edge intelligent fire point detection services is solved, and the problem of difficult management of edge fire point detection resources is also solved. Under this mechanism, the service and management efficiency of edge intelligent fire point detection equipment is significantly improved.

[0064] In order to implement the above embodiment, the present invention also proposes a KubeEdge-based grid edge intelligent fire point detection container management system. The system is deployed with KubeEdge.

[0065] Figure 4 A block diagram of a KubeEdge-based grid edge intelligent fire point detection container management system provided in an embodiment of the present invention.

[0066] like Figure 4 As shown, the KubeEdge-based grid edge intelligent fire point detection container management system includes a control end and multiple edge devices, including:

[0067] The control end is used to save, encapsulate, and containerize the trained fire point detection model to obtain the fire point detection container and the container image of the fire point detection container; it is also used to monitor the real-time operating status of the fire point detection container in each edge device to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal.

[0068] Each edge device is used to create a configuration file for a fire point detection container, so as to obtain the required fire point detection container from the container image of the control end, and then use the corresponding fire point detection container to perform fire point detection respectively.

[0069] Furthermore, in a possible implementation of the embodiment of the present invention, the control end uses a Django server to encapsulate the saved fire point detection model.

[0070] Furthermore, in a possible implementation of the embodiment of the present invention, the control end uses Docker to build the encapsulated fire point detection model to obtain a fire point detection container and a container image of the fire point detection container.

[0071] Furthermore, in a possible implementation of an embodiment of the present invention, the fire point detection model adopts a lightweight pruning fire point algorithm based on Yolo V8.

[0072] Furthermore, in a possible implementation method of an embodiment of the present invention, Kubernetes is deployed on the control server in the control end, and the Cloudcore module of KubeEdge is installed in the control server, and the Edgecore module of KubeEdge is deployed on each edge device, thereby obtaining an edge container management platform.

[0073] Furthermore, in a possible implementation manner of the embodiment of the present invention, the abnormal real-time operating status refers to a container failure or insufficient resources.

[0074] It should be noted that the aforementioned explanation of the embodiment of the KubeEdge-based grid edge intelligent fire point detection container management method is also applicable to the KubeEdge-based grid edge intelligent fire point detection container management system of this embodiment, and will not be repeated here.

[0075] In an embodiment of the present invention, by deploying an edge container management platform based on KubeEdge, the edge container management platform includes a control end and multiple edge devices; the control end saves, encapsulates, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container; each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then uses the corresponding fire point detection container to perform fire point detection respectively; the control end monitors the real-time operating status of the fire point detection container in each edge device to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal. In this case, by deploying an edge container management platform based on KubeEdge, the control end is used to obtain the fire point detection container, and the real-time operating status of the fire point detection container in each edge device is monitored to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal. Therefore, KubeEdge is used to manage the computing resources of the edge devices of the power grid, which can efficiently schedule and manage the resources and status of the edge devices in real time. In addition, each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end for fire point detection. Therefore, the edge fire point detection application encapsulated in the container can better adapt to the heterogeneous characteristics of the edge fire point detection hardware. Between the fire point detection intelligent devices of the same operating system, the service can be seamlessly migrated and shared, avoiding the difficulties caused by multiple deployments of the basic environment and repeated deployments during algorithm upgrades. Therefore, the present invention solves the problem of complex deployment of edge intelligent fire point detection services and the difficulty of managing edge fire point detection resources.

[0076] In order to effectively manage the containerized resources of power grid fire point detection, the method and system of the present invention construct a distributed resource management platform, which deploys different components on the central node (i.e., the control server) and the edge node to realize centralized management and allocation of resources. By deploying the KubeEdge platform, Cloudcore and Edgecore components can be deployed on the central node and the edge node respectively, the resource status can be monitored in real time, and the container resources can be dynamically managed according to business needs. At the same time, the function of KubeEdge is used to manage, monitor and run containerized business services for node resources, and the intelligent model of deep learning fire point identification technology and fire point prediction technology is combined to realize adaptive resource allocation between intelligent terminal nodes (i.e., edge devices) at the edge of the power grid, thereby improving resource utilization efficiency. In addition, the system can also monitor the status of each node in real time, correct the faults of the model or node in time, and ensure the stability and reliability of the power grid fire point detection system. At the same time, Kubernetes will be deployed on the central node, and its various functions will be used to realize the monitoring of node resources, and for different business types, the adaptive resource allocation between satellite nodes (i.e., edge nodes) will be completed using intelligent models, so as to realize the coordinated utilization of resources and improve resource utilization efficiency. And when a model or node fails, the system can correct the error in time. The purpose of the present invention is to provide a fire point detection container deployment and management solution for edge intelligent devices in the power network scenario, which can be uniformly orchestrated and managed, and iteratively upgraded. By encapsulating the fire point detection algorithm into a container that can provide a unified calling interface, the container is managed uniformly on the KubeEdge platform, simplifying the application process, overcoming the high cost and low efficiency defects of the current power grid edge intelligent devices that require manual maintenance and operation management, simplifying the fire point detection application deployment and management process, realizing the application of the fire point detection algorithm in the power grid edge intelligent environment, and reducing the use threshold and operation and maintenance cost of multi-point and multi-region distributed multi-fire point detection equipment.

[0077] In order to implement the above embodiments, the present invention also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0078] In order to implement the above embodiments, the present invention further proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0079] In order to implement the above embodiments, the present invention further provides a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0080] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0081] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0082] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

[0083] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0084] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0085] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0086] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0087] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A KubeEdge-based power grid edge intelligent fire point detection container management method, characterized in that: include: Deploy a KubeEdge-based edge container management platform, which includes a control terminal and multiple edge devices; The control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container; Each edge device creates a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then uses the corresponding fire point detection container to perform fire point detection respectively; The control end monitors the real-time operating status of the fire point detection container in each edge device, so as to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal.

2. The KubeEdge-based grid edge intelligent fire point detection container management method according to claim 1 is characterized in that: The deployment is based on the KubeEdge edge container management platform, including: Kubernetes is deployed on the control server in the control end, and the Cloudcore module of KubeEdge is installed in the control server, and the Edgecore module of KubeEdge is deployed on each edge device to obtain an edge container management platform.

3. The KubeEdge-based grid edge intelligent fire point detection container management method according to claim 1 is characterized in that: The control end saves, packages, and containerizes the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container, including: The control end obtains the trained fire point detection model, and saves and encapsulates the trained fire point detection model; For the encapsulated fire point detection model, a fire point detection container and a container image of the fire point detection container are obtained by using Docker to build.

4. The KubeEdge-based grid edge intelligent fire point detection container management method according to claim 1 or 3, characterized in that: Use the Django server to encapsulate the saved fire point detection model.

5. The KubeEdge-based power grid edge intelligent fire point detection container management method according to claim 1 is characterized in that: The fire point detection model adopts a lightweight pruning fire point algorithm based on Yolo V8.

6. The KubeEdge-based power grid edge intelligent fire point detection container management method according to claim 1 is characterized in that: The abnormal real-time operating status refers to a container failure or insufficient resources.

7. A KubeEdge-based grid edge intelligent fire point detection container management system, characterized in that: The system is deployed with KubeEdge, and the system includes a control terminal and multiple edge devices; The control end is used to save, encapsulate, and containerize the trained fire point detection model to obtain a fire point detection container and a container image of the fire point detection container; it is also used to monitor the real-time operating status of the fire point detection container in each edge device to dynamically manage the fire point detection container in each edge device when the real-time operating status is abnormal. Each edge device is used to create a configuration file for the fire point detection container to obtain the required fire point detection container from the container image of the control end, and then use the corresponding fire point detection container to perform fire point detection respectively.

8. The KubeEdge-based grid edge intelligent fire point detection container management system according to claim 7, characterized in that: The control end uses the Django server to encapsulate the saved fire point detection model, and uses Docker to build the encapsulated fire point detection model to obtain a fire point detection container and a container image of the fire point detection container.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 6 when executed by a processor.

Images