A method for collaborative deployment of a hybrid cloud and edge computing of an integrated energy platform

Abstract

This invention discloses a hybrid cloud and edge computing collaborative deployment method for an integrated energy platform, belonging to the field of integrated energy system information technology. The invention relies on a collaborative orchestration layer to construct a global resource view, dynamically planning the deployment locations of services in private clouds, public clouds, and edge nodes based on business type, data sensitivity, and load characteristics. Edge nodes complete on-site data preprocessing and real-time local control, selecting key data for uploading to the private cloud for cross-domain data analysis. The data synchronization layer establishes a cloud-edge bidirectional synchronization link; in the event of network anomalies, the edge operates offline autonomously and caches data; after the link recovers, it automatically synchronizes and verifies data consistency. The adaptive disaster recovery layer provides redundant storage of edge core configurations in the cloud, enabling services to migrate to the nearest edge or cloud disaster recovery node in the event of node failure. This invention rationally coordinates heterogeneous resources, strengthens network outage self-healing and fault transfer capabilities, and improves the stability of the cloud-edge integrated system.

Description

Technical Field

[0001] This invention relates to the field of information technology for integrated energy systems, and in particular to a method and apparatus for analyzing the flow rate, efficiency, and water consumption rate of hydropower plant generator units. Background Technology

[0002] Integrated energy systems, as crucial carriers for the synergistic complementarity of various energy forms such as electricity, heat, cooling, water, gas, and steam, have their information platforms widely applied in scenarios such as industrial parks, buildings, and factories. Among related technologies, a technical system supporting massive device data acquisition, real-time monitoring, and optimized scheduling has been constructed through the collaborative operation of cloud computing, edge computing, and hybrid cloud architecture. Specifically, this system covers the entire process from centralized computing and storage in the cloud to millisecond-level response at the edge, including key aspects such as the hybrid deployment of private and public clouds, local autonomous control of edge nodes, and data interaction between the cloud and the edge. It aims to meet the stringent requirements of integrated energy platforms for computing power, network, storage, and real-time performance.

[0003] However, existing hybrid cloud and edge computing collaborative deployment solutions directly adopt a static private cloud and public cloud partitioning strategy, lacking dynamic scheduling capabilities based on business load. This may lead to low utilization of computing resources or degraded service quality during peak spot trading periods. Simultaneously, edge nodes are typically deployed and run independently, lacking a unified containerized orchestration and collaborative scheduling mechanism. When nodes fail or experience excessive load, dynamic migration of business tasks is impossible. Furthermore, existing technologies lack sufficient data consistency guarantees for cloud-edge collaboration. In the event of network outages or cloud platform failures, edge nodes cannot obtain the latest policy parameters, resulting in decreased local control accuracy and limited overall system fault tolerance, making it difficult to meet the high reliability requirements of integrated energy systems for continuous operation. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the first objective of this invention is to propose a method for the collaborative deployment of hybrid cloud and edge computing in an integrated energy platform.

[0006] Another objective of this invention is to propose a hybrid cloud and edge computing collaborative deployment device for an integrated energy platform.

[0007] The third objective of this invention is to provide a computer device.

[0008] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, a first aspect of the present invention proposes a method for the collaborative deployment of hybrid cloud and edge computing in an integrated energy platform, comprising:

[0010] S1 maintains a global resource view through a collaborative orchestration layer and dynamically decides the deployment location of each business module in a private cloud, public cloud, or edge node based on business type, data sensitivity, and computing load characteristics. S2 performs local preprocessing and real-time control of the collected field equipment data through edge nodes, and uploads the preprocessed key data to the private cloud for cross-regional analysis. S3 establishes a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. S4 uses an adaptive disaster recovery layer to redundantly back up the core configuration data of edge nodes in the cloud, and when a node failure is detected, it dynamically switches the business tasks of the failed node to nearby edge nodes or cloud disaster recovery nodes for execution.

[0011] In one embodiment of the present invention, S1 includes: The collaborative orchestration layer collects the CPU, memory, storage, network bandwidth, and GPU computing power resource status of private clouds, public clouds, and various edge nodes in real time, forming a unified panoramic view of resources; Based on a real-time priority strategy, services requiring millisecond-level response are forced to be deployed on edge nodes or private cloud nodes close to the data source. Based on a data security strategy, data processing services involving user privacy and trade secrets are prioritized for deployment on the private cloud. Based on a cost optimization strategy, the distribution ratio of non-core business computing tasks between the public cloud and the private cloud is dynamically adjusted. Based on a load balancing strategy, computing tasks are distributed to nodes with lower loads for execution.

[0012] In one embodiment of the present invention, it further includes: When an edge node receives a computing task request, it evaluates the task's real-time requirements, data volume, and computational complexity. Query the remaining local resources on the edge node to determine if the local resources can meet the task execution conditions; if the local resources are sufficient and the real-time requirements are high, then execute the task locally on the edge node. If local resources are insufficient or the complexity of the task exceeds the capabilities of the edge node, the task will be uploaded to the private cloud for execution via the hybrid cloud gateway. If the private cloud has a high load and the data is allowed to leave the domain, the task will be further scheduled to be executed on the public cloud; The execution result is returned to the edge node that initiated the request through the data synchronization layer.

[0013] In one embodiment of the present invention, S2 includes: Edge nodes calculate the Hamming distance between the received raw byte stream and each protocol template using the protocol feature library they maintain. They select the protocol with the smallest Hamming distance as the parsing protocol. If the smallest Hamming distance exceeds a preset threshold, it is marked as an unknown protocol and an alarm is triggered. The parsed data is then normalized to obtain data in a uniform format. The data in a unified format is filtered, cleaned, compressed, and subjected to preliminary statistical analysis to generate aggregated key indicators; the aggregated key indicators are then uploaded to the private cloud through a data synchronization layer.

[0014] In one embodiment of the present invention, it further includes: Based on formula The original data values ​​are converted into normalized values ​​within a target unified interval, where Max and Min are the dynamic maximum and minimum values ​​of the measurement point in the historical data. To unify the target range.

[0015] In one embodiment of the present invention, S3 includes: The data synchronization layer uses a combination of publish-subscribe message queues and state version vectors to serialize and compress telemetry data using differential coding combined with efficient compression algorithms. It determines partitions based on the consistent hashing of data tags to ensure that data from the same device is written sequentially to the same partition. The strategy parameters sent from the cloud are parsed by the converter to extract the strategy description and target value. The control quantity is calculated according to the device characteristic model and encapsulated into a fieldbus message and sent to the edge node. When resuming transmission from a breakpoint, the edge node and the cloud calculate the checksum of the data block to be synchronized. If they match, the transmission is confirmed; otherwise, the data is retransmitted block by block.

[0016] In one embodiment of the present invention, S4 includes: The adaptive disaster recovery layer redundantly backs up the core configuration data of the edge nodes in the cloud and persistently stores the key policy parameters issued by the cloud locally on the edge nodes. When an edge node failure is detected, the device connection and data acquisition tasks within the jurisdiction of the failed node are dynamically switched to the neighboring edge node, and the business containers of the failed node are automatically launched on the neighboring node. When a cloud-edge network outage is detected, the edge node automatically switches to offline autonomous mode and continues to run core business based on locally cached policy parameters and models. Once the network is restored, it will automatically perform data synchronization and state alignment. When a failure of the private cloud master node is detected, the system automatically switches to a backup private cloud node or a public cloud disaster recovery node. Critical services are quickly started on the disaster recovery node, and the status is automatically synchronized after the failure is recovered.

[0017] To achieve the above objectives, a second aspect of the present invention provides a hybrid cloud and edge computing collaborative deployment device for an integrated energy platform, comprising: The deployment scheduling and management module is used to maintain a global resource view through the collaborative orchestration layer, and dynamically decide the deployment location of each business module in the private cloud, public cloud or edge node based on business type, data sensitivity and computing load characteristics. The edge local processing module is used to perform local preprocessing and real-time control of the collected field equipment data through edge nodes, and upload the preprocessed key data to the private cloud for cross-regional analysis. The cloud-edge data synchronization module is used to establish a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. The adaptive disaster recovery switching module is used to perform redundant backup of the core configuration data of edge nodes in the cloud through the adaptive disaster recovery layer, and dynamically switch the business tasks of the faulty node to the nearby edge node or the cloud disaster recovery node when a node failure is detected.

[0018] The present invention discloses a method and apparatus for the collaborative deployment of hybrid cloud and edge computing in an integrated energy platform, which can realize business-aware dynamic resource scheduling, reduce the operating cost of the integrated energy platform, and improve the system's real-time response capability, data consistency and fault tolerance.

[0019] To achieve the above objectives, a third aspect of this application provides a computer device, including a processor and a memory; wherein the processor runs a program corresponding to the executable program code by reading executable program code stored in the memory, for implementing the method described in the first aspect embodiment.

[0020] To achieve the above objectives, a fourth aspect of this application provides a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method described in the first aspect.

[0021] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0022] Figure 1 This is a flowchart of a hybrid cloud and edge computing collaborative deployment method for an integrated energy platform according to an embodiment of the present invention; Figure 2 This is a structural diagram of a hybrid cloud and edge computing collaborative deployment device for an integrated energy platform according to an embodiment of the present invention; Figure 3 It is a computer device according to an embodiment of the present invention. Detailed Implementation

[0023] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0025] The following description, with reference to the accompanying drawings, describes a method and apparatus for the collaborative deployment of hybrid cloud and edge computing in an integrated energy platform according to an embodiment of the present invention.

[0026] Figure 1 This is a flowchart of a hybrid cloud and edge computing collaborative deployment method for an integrated energy platform according to an embodiment of the present invention, such as... Figure 1 As shown, it includes: S1 maintains a global resource view through a collaborative orchestration layer and dynamically decides the deployment location of each business module in a private cloud, public cloud, or edge node based on business type, data sensitivity, and computing load characteristics. Specifically, the edge node initialization and registration process involves the newly deployed edge node automatically initializing its container runtime environment via a lightweight Kubernetes distribution (such as K3s) after power-on. The edge node sends a registration request to the cloud-based orchestration layer, reporting node information (node ​​ID, geographical location, network conditions, computing resources, etc.). After verifying the node's identity, the orchestration layer adds the node to the unified resource view and issues a list of initial business container images to be deployed on that node. The edge node pulls the business container images from the cloud image repository and starts the container service. The cloud uses edge container cloud image processing methods to whitelist, detect features, and harden and encrypt the images, ensuring their operational stability and security.

[0027] Furthermore, cloud deployment and policy initialization involve deploying core services (optimized scheduling engine, device model library, pricing model library, etc.) in the private cloud and auxiliary services (big data analysis, carbon emission accounting, report generation, etc.) in the public cloud. The hybrid cloud gateway configures a secure communication link between the private and public clouds, setting data synchronization policies and traffic control parameters. The collaborative orchestration layer initializes deployment policy rules based on business needs, specifying default deployment locations (private cloud / public cloud / edge node) and scheduling policies for various services. The cloud trains an initial load prediction model, device performance model, and optimization policy parameters based on global data, and distributes them to each edge node through the data synchronization layer.

[0028] S2 performs local preprocessing and real-time control of the collected field equipment data through edge nodes, and uploads the preprocessed key data to the private cloud for cross-regional analysis. Specifically, during normal operation, the cloud-edge collaboration involves edge nodes continuously collecting data from field equipment, performing high-frequency data aggregation and local preprocessing. For control commands with high real-time requirements, edge nodes execute them directly locally, achieving millisecond-level response. Edge nodes then upload aggregated key metrics (such as equipment energy efficiency, load trends, and abnormal events) to the private cloud via a data synchronization layer. The private cloud performs cross-regional analysis and model training. The private cloud and public cloud exchange non-sensitive data through a hybrid cloud gateway, with the public cloud executing large-scale data analysis tasks and generating reports. The collaborative orchestration layer continuously monitors the resource status of each node and dynamically adjusts deployment strategies based on changes in business load. For example, during peak spot trading periods, optimized computing tasks are migrated from edge nodes to the private cloud for execution in advance.

[0029] Furthermore, in the dynamic scheduling of cloud-edge collaborative tasks, when an edge node receives a complex computing task request, the scheduling algorithm of the collaborative orchestration layer performs the following decisions: assessing the task's real-time requirements, data volume, and computational complexity; querying the edge node's remaining local resources (CPU, memory, network bandwidth) to determine if local resources are sufficient; if local resources are sufficient and the task has high real-time requirements (such as millisecond-level response requirements), the task is executed locally on the edge node; if local resources are insufficient or the task's computational complexity exceeds the edge node's capabilities, the task is uploaded to the private cloud for execution through the hybrid cloud gateway. The private cloud utilizes its more abundant computing resources to process complex tasks and returns the results to the edge node; if the private cloud has a high load and the data is allowed to leave the domain (non-sensitive data), the task is further scheduled to be executed on the public cloud, utilizing the public cloud's elastic computing capabilities to handle peak loads; after the execution results are returned to the edge node, the edge node completes the business loop based on the results.

[0030] S3 establishes a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. Specifically, the cloud-edge collaborative data synchronization involves a data synchronization layer that executes data synchronization according to a preset synchronization strategy: From cloud to edge synchronization: Load prediction model parameters and optimization strategy parameters trained in the cloud are pushed to each edge node through the data synchronization layer. Upon receiving this data, the edge nodes update their local models and strategy libraries, ensuring that the strategies executed locally by the edge nodes are consistent with the global optimization goals in the cloud. From edge to cloud synchronization: Device operation data collected by the edge nodes and local optimization execution results are synchronized using different strategies based on data classification: Data with high real-time requirements (such as device alarms) is immediately pushed to the cloud; data with medium real-time requirements uses short-cycle polling synchronization; historical statistical data uses a timed batch upload method to reduce network overhead. When the cloud-edge network connection is interrupted, the edge node initiates an offline autonomous mode, caching the data to be uploaded in local persistent storage. Simultaneously, the strategy parameters issued by the cloud are persistently stored locally on the edge node, and synchronization verification is automatically completed after the network is restored to ensure eventual consistency. While some patents propose that nodes can operate autonomously based on the latest model when cloud platform communication is interrupted or fails, this invention further expands on this idea by automatically triggering data synchronization and state alignment after network recovery.

[0031] Furthermore, regarding adaptive disaster recovery and failover, when a node failure occurs, the adaptive disaster recovery layer executes the following failover process: Edge node failure: After the node health monitoring module detects a failure, it dynamically switches the device connections and data acquisition tasks within the jurisdiction of the failed node to a nearby edge node. The business containers of the failed node are automatically launched on the nearby node according to the configuration backed up in the cloud. Containerized deployment supports the rapid distribution and startup of application images. After the failure is recovered, the failed node rejoins the cluster and automatically synchronizes the data increments during the failure period from the cloud. Cloud-edge network interruption: The edge node switches to offline autonomous mode and continues to run core businesses based on locally cached policy parameters and models, ensuring power safety and basic operation, and improving system resilience. After the network is restored, data synchronization and state alignment are automatically triggered, and cached data during the offline period is uploaded to the cloud through a breakpoint resume mechanism. Cloud platform failure: When the private cloud master node fails, the standby private cloud node automatically takes over the core services. If the standby node is unavailable and business data is allowed to leave the domain, critical services are quickly launched on the public cloud disaster recovery node, and the state is automatically synchronized after the failure is recovered.

[0032] S4 uses an adaptive disaster recovery layer to redundantly back up the core configuration data of edge nodes in the cloud, and when a node failure is detected, it dynamically switches the business tasks of the failed node to nearby edge nodes or cloud disaster recovery nodes for execution.

[0033] Specifically, the deployment strategy is dynamically adjusted and optimized. The orchestration layer continuously monitors indicators such as resource utilization, task execution latency, and network bandwidth usage of each node, and periodically (e.g., every 5 minutes) evaluates the effectiveness of the deployment strategy. Deployment strategy adjustments are automatically triggered when the following conditions are detected: when the load on a private cloud node continuously exceeds a threshold, some non-core computing tasks are dynamically migrated to the public cloud; when the processing capacity of an edge node is insufficient, some of its computing tasks are allocated to nearby edge nodes or uploaded to the private cloud using a task scheduling algorithm; before the peak period of spot trading, optimized computing tasks are migrated from edge nodes to the private cloud in advance based on historical load pattern predictions, reserving edge node resources to cope with sudden local control needs; tasks with high data security levels are still forcibly deployed to the private cloud for execution, even if public cloud resources are more abundant. The adjustment results are distributed to relevant nodes through the data synchronization layer to ensure that the dynamic optimization of the deployment strategy does not affect the normal operation of the business.

[0034] This invention proposes a hybrid cloud and edge computing collaborative deployment architecture for an integrated energy platform. The system consists of five layers: a hybrid cloud infrastructure layer, an edge computing layer, a collaborative orchestration layer, a data synchronization layer, and an adaptive disaster recovery layer. Through intelligent scheduling algorithms and containerized deployment technology, it achieves an organic integration of global optimization in the cloud and local response at the edge. This architecture is based on the concept of integrated edge-cloud layered processing. The edge side focuses on data acquisition and basic control, the edge side undertakes filtering, aggregation, and local real-time computing, and the cloud platform is responsible for unified cross-regional analysis and indicator management.

[0035] Furthermore, the hybrid cloud infrastructure layer is the computing and storage foundation of the deployment architecture of this invention, comprising two main parts: private cloud and public cloud, and supporting flexible dynamic deployment strategy switching.

[0036] This invention proposes a layered dynamic routing and adaptive traffic scheduling network, rather than a simple combination of public and private clouds, to improve the network architecture. The core is a service-label-based intelligent gateway matrix, with each gateway node having a built-in lightweight traffic prediction model. The specific network structure is as follows: on the private cloud side, a primary control gateway node (Active) and a backup control gateway node (Standby) are deployed, using the VRRP protocol for high availability; on the public cloud side, regional gateway nodes are deployed, each maintaining an encrypted tunnel to the private cloud; gateways learn each other's service-label routing tables through dynamic routing protocols (OSPF+BGP), forming a service-aware virtual private network (BS-VPN).

[0037] Further, the data processing and training process involves the gateway node collecting link latency (L), packet loss rate (P), and bandwidth utilization (B) in real time, and using exponentially weighted moving average (EWMA) to predict the link quality for the next time step.

[0038] in The current measured quality (overall score) is represented by α, which is a smoothing factor (dynamically configured based on historical link data). When the predicted quality falls below the threshold, link switching or traffic shaping is automatically triggered.

[0039] Furthermore, the private cloud is deployed in the user's data center or the power plant / campus intranet, carrying core sensitive data and high-frequency real-time services. Private cloud nodes utilize enterprise-grade server clusters, interconnected via a private network, and deployed using a Kubernetes container management platform. The services carried by the private cloud include: IoT access services (handling access to devices within the campus), the core module of the optimized scheduling engine (relying on low-latency computing), policy execution services (requiring millisecond-level instruction issuance), device model libraries, pricing model libraries, and other core data. In Datang Group's smart power plant solution, the private cloud is deployed within the power plant's intranet to process sensitive data and critical business operations.

[0040] Furthermore, the public cloud utilizes infrastructure services provided by mainstream public cloud service providers (such as Alibaba Cloud, Huawei Cloud, Tencent Cloud, etc.) to support non-sensitive data and large-scale computing tasks. This invention provides a typical computing example on the public cloud: the incremental training of the load forecasting model employs a federated averaging algorithm (FedAvg), aggregating local model updates uploaded by each edge node to avoid transmitting raw data. The global model update formula is:

[0041] in The weights of the k-th edge node after local training are given. Let n be the number of samples at this node, and n be the total number of samples. After each round of aggregation, the public cloud will update the model parameters. Distribute to edge nodes.

[0042] Furthermore, the deployment region of the public cloud can be selected based on user distribution and compliance requirements, supporting multi-region deployment for proximity-based access. The services supported by the public cloud include: historical data storage and big data analysis, carbon emission accounting and report generation, cross-regional integrated energy optimization and scheduling (without relying on low latency), machine learning model training and updates, and non-real-time reporting and visualization. By 2025, the core characteristic of hybrid cloud has upgraded from "simple connection between cloud and on-premises resources" to "deep integration of cloud and on-premises environments," pursuing seamless scheduling of computing, storage, and network resources.

[0043] Furthermore, the hybrid cloud gateway acts as a bridge between the private and public clouds, responsible for secure communication and data exchange between them. The hybrid cloud gateway provides the following functions: Network interconnection: Achieving high-speed and stable connections between the private and public clouds via dedicated lines, VPNs, or encrypted tunnels, with communication latency controlled within a reasonable range; Data synchronization: Enabling real-time or near real-time synchronization of critical data (such as device models, user configurations, and policy parameters) between the private and public clouds; Security isolation: Ensuring that sensitive data (such as raw device data, user information, and transaction policies) does not leave the private cloud environment, while only non-sensitive data (such as statistical analysis results and trained model parameters) can be synchronized to the public cloud; Traffic control: Dynamically adjusting the bandwidth and frequency of inter-cloud data transmission based on business priorities and network conditions to prevent inter-cloud data synchronization from affecting critical business operations.

[0044] Furthermore, the edge computing layer, which serves as the "peripheral nerves" and "fast reflex arc" of the deployment architecture of this invention, consists of various edge computing nodes (edge ​​gateways, smart terminals, or edge servers) deployed close to the data source. Edge nodes are primarily deployed at integrated energy project sites, such as parks, buildings, factories, charging stations, and microgrids, responsible for achieving localized, low-latency data processing and decision execution.

[0045] The core functions of edge nodes include: data access and protocol parsing, connecting with a wide variety of energy devices, sensors, and smart meters, and enabling real-time parsing and data normalization of various industrial protocols such as Modbus, BACnet, IEC 104, and OPC UA, providing data in a unified format for the upper layer.

[0046] Furthermore, the edge nodes maintain a protocol signature library, containing frame header signatures for each protocol (such as 0x0001 for Modbus and 0x6801 for IEC104). For the received raw byte stream... Calculate the Hamming distance H between the target and each protocol template in the feature library. j :

[0047] choose The simplest protocol is used as the parsing protocol. If the simplest... If the threshold is exceeded, it will be marked as an unknown protocol and an alarm will be triggered.

[0048] Furthermore, the parsed data is in the format specific to each protocol, and the edge nodes perform normalization transformation:

[0049] in This represents the dynamic maximum and minimum values ​​of the measuring point in the historical data. [] represents the target uniform range (e.g., 0-100). Non-numerical state variables are mapped to an enumerated value dictionary.

[0050] Furthermore, high-frequency data aggregation and edge preprocessing involve filtering, cleaning, compressing, and performing preliminary statistical analysis on the raw high-frequency collected data. For example, second-level power consumption data is aggregated at the edge into minute-level effective values, greatly reducing network transmission pressure and cloud storage burden.

[0051] Furthermore, millisecond-level real-time control and rapid closed-loop operation: Executing local control strategies with high real-time requirements. Examples include demand control based on real-time power to prevent over-limit penalties; islanding detection and rapid switching in microgrids; and rapid power regulation of distributed power sources and energy storage to smooth out fluctuations. These control commands must be completed within tens to hundreds of milliseconds, and only edge computing can meet these latency requirements.

[0052] Furthermore, local business logic and autonomous operation: When the platform's main system or network is temporarily interrupted, edge nodes can maintain the autonomous operation of critical local businesses according to a preset rule base, ensuring energy security and basic operation, and improving system resilience.

[0053] Furthermore, the containerized deployment and orchestration of edge nodes: In this invention, all edge nodes utilize containerization technology (Docker) to encapsulate business services, and are uniformly orchestrated and managed through lightweight Kubernetes distributions (such as K3s, KubeEdge, etc.). Containerized deployment of edge nodes offers the following advantages: Lightweight and Standardized: Container technology provides a lightweight operating environment, enabling edge nodes to run standardized business services with lower resource overhead, supporting rapid deployment and elastic scaling of edge nodes. In January 2025, China Southern Power Grid Digital Platform Technology applied for a patent for edge container cloud image processing based on cloud-edge collaboration scenarios. Through image whitelisting, image detection models, and repair and hardening processes, the operational stability of images has been significantly improved. Unified Orchestration and Management: The cloud uses container orchestration platforms (such as KubeEdge and K3s) to uniformly manage edge nodes distributed across various parks, enabling one-click deployment, remote upgrades, and elastic scaling of applications, significantly reducing the complexity of on-site operations and maintenance. Differentiated Container Resource Pool: Referring to the hierarchical container scheduling concept, containers are divided into differentiated levels based on business type, such as S-level (core real-time control), A-level (important monitoring), and B-level (general data reporting), allocating different resource quotas and scheduling priorities to services of different levels. Cloud-based training and edge inference: AI models trained in the cloud (such as load forecasting models and equipment fault diagnosis models) can be packaged into container images and pushed down to edge nodes for execution, achieving a collaborative "cloud-based training and edge-based inference" model. A specific model example is given: The load forecasting model uses an LSTM-Attention structure. Input features: historical load sequences (hourly values ​​over the past 24 hours), temperature, humidity, and date type. Network structure: two LSTM layers (64 units per layer) + attention layer + fully connected layer. During cloud training, data normalization is performed. The loss function uses the mean absolute percentage error (MAPE).

[0054] The optimizer employs an adaptive learning rate algorithm (such as Adam). Edge nodes load model weights distributed from the cloud and use a lightweight inference engine to perform single-step predictions, meeting real-time requirements.

[0055] Furthermore, the equipment fault diagnosis model employs an autoencoder and an isolated forest approach: features include equipment vibration frequency, temperature, and current harmonic distortion rate. The autoencoder is trained in the cloud to reconstruct normal samples, and the reconstruction error is defined as... A dynamic threshold (such as the mean plus three standard deviations) is set based on the error distribution of normal samples. Edge nodes calculate the reconstruction error of the current sample in real time, and trigger an alarm if the threshold is exceeded.

[0056] Furthermore, the deployment modes of edge nodes are as follows: Lightweight deployment: suitable for resource-constrained field devices, running core acquisition and control services, and uploading data to the cloud after compression; Standard deployment: suitable for industrial parks and commercial buildings, running a complete edge service suite, including data acquisition, local optimization, autonomous control and other functions; Cluster deployment: suitable for large parks or multi-project centralized management scenarios, where multiple edge nodes form an edge cluster, and load balancing and failover are achieved through collaborative scheduling.

[0057] Furthermore, the collaborative orchestration layer, the core management module of this invention's hybrid cloud and edge computing collaborative deployment architecture, is responsible for the unified scheduling of global resources, intelligent allocation of business tasks, and collaborative control between the cloud and the edge. A unified resource view: The collaborative orchestration layer maintains a global resource view, collecting real-time resource status (CPU, memory, storage, network bandwidth, GPU computing power, etc.) from private clouds, public clouds, and each edge node, forming a unified panoramic view of resources and providing a data foundation for task scheduling decisions.

[0058] Furthermore, the business-aware deployment strategy engine: Based on the characteristics of integrated energy services, this invention designs a dynamic deployment strategy engine that supports the following deployment strategies: Real-time priority strategy: Services requiring millisecond-level response (such as device control command issuance and protection action execution) are forcibly deployed on edge nodes or private cloud nodes close to the data source to ensure end-to-end latency is within an acceptable range; Data security strategy: Data processing services involving user privacy and trade secrets are prioritized for deployment on the private cloud, while non-sensitive data analysis tasks can be dynamically scheduled to the public cloud; Cost optimization strategy: Taking into account both public cloud service fees and private cloud operation and maintenance costs, the allocation ratio of non-core business computing tasks between the public and private clouds is dynamically adjusted; Load balancing strategy: Based on the real-time load status of each node, computing tasks are allocated to nodes with lower loads to avoid local overload. Cloud-edge collaborative task scheduling algorithm: The collaborative orchestration layer uses an intelligent scheduling algorithm to achieve dynamic allocation of computing tasks.

[0059] Furthermore, when an edge node receives a computing task request, the algorithm executes the following decision-making process: It assesses the task's real-time requirements (expected completion time, allowable delay), data volume, and computational complexity; it queries the edge node's remaining local resources to determine if the local resources can meet the task execution conditions; if local resources are sufficient and real-time requirements are high, the task is executed locally on the edge node; if local resources are insufficient or the task complexity exceeds the edge node's capabilities, the task is uploaded to the private cloud for execution via the hybrid cloud gateway; if the private cloud has a high load or requires utilizing the powerful computing power of the public cloud, and data is allowed to leave the domain, the task is further scheduled to the public cloud for execution; the execution result is returned to the requesting edge node through the data synchronization layer, completing the closed loop. Cloud-edge collaborative scheduling method reference: The cloud-edge collaborative scheduling mechanism of this invention integrates multiple approaches from existing technologies. Referring to the task allocation and state synchronization method in the edge cloud collaborative computing system applied for by State Grid Zhejiang Electric Power Co., Ltd., the optimal allocation action is selected based on multi-domain context vectors through a context decision model, and the reward value is used as a dynamic learning objective to train and update the model in real time. Meanwhile, referencing the edge computing-based smart microgrid control method applied for by the State Grid (Suzhou) Urban Energy Research Institute, the control commands are simulated based on the edge digital twin model when the local edge aggregation node performs computing tasks, and the computing tasks are sent to the target edge aggregation node to alleviate the slow control response problem caused by the centralized architecture.

[0060] Further, the data synchronization layer is responsible for ensuring data consistency and synchronizing policy parameters between the cloud and the edge, and between private and public clouds. It ensures that key data such as models, policies, and configurations used by various modules of the integrated energy platform remain consistent, serving as a crucial support for the cloud-edge collaborative architecture. The data synchronization layer architecture employs a combination of publish-subscribe message queues and state version vectors. Inputs: Device telemetry data (structured format) generated by edge nodes, local model update parameters, and policy instructions issued from the cloud. Outputs: Edge nodes receive converted binary instructions to drive the actuators; the cloud receives compressed telemetry data, decompresses it, and stores it in a time-series database. Step 1: Data serialization and compression. Telemetry data uses differential coding combined with efficient compression algorithms (such as Zstd). For continuous value sequences, the difference is calculated first, and then variable-length encoding is applied to the differential sequence, achieving a compression rate of over 70%. Step 2: Topic allocation and partitioning strategy. Partitions are determined based on consistent hashing of data tags (device ID, data type), ensuring that data from the same device is written sequentially to the same partition. Step 3: Data transformation from cloud to edge. The strategy parameters (such as PID coefficients) delivered from the cloud are converted into a sequence of instructions executable by the edge nodes. The converter parses the strategy description, extracts the target value, and calculates the control quantity (such as PWM duty cycle) based on the device characteristic model. The calculation formula uses incremental PID:

[0061] Where 'e' represents the deviation (target value minus current value). Finally, it is encapsulated into a fieldbus message (e.g., Modbus / TCP). Step 4: Data consistency verification during breakpoint resumption. Calculate the checksum (e.g., CRC32) of the data block to be synchronized. The edge node and the cloud calculate separately; if they match, confirmation is confirmed; otherwise, retransmission is performed block by block.

[0062] Furthermore, a two-way data synchronization mechanism is implemented: The data synchronization layer supports the following two-way synchronizations: From cloud to edge: Device model parameters, optimization strategy parameters, load prediction models, carbon accounting coefficients, etc., trained in the cloud are distributed to edge nodes to ensure that the strategies executed locally on the edge nodes are consistent with the global optimization goals in the cloud; From edge to cloud: Device operation data, local optimization execution results, alarm events, etc., collected by the edge nodes are uploaded to the cloud for cloud model training and strategy optimization updates. Synchronization strategy configuration: Differentiated synchronization strategies are set according to the different data characteristics of integrated energy business: Real-time synchronization: For data with high timeliness requirements such as strategy commands and control parameters, real-time push synchronization is used, with latency controlled in milliseconds; Near real-time synchronization: For data that needs to be updated quickly, such as device telemetry data and operating status, short-cycle polling or incremental synchronization is used, with latency controlled in seconds to minutes; Batch synchronization: For data with low real-time requirements, such as historical statistical data, report data, and logs, timed batch upload is used to reduce network overhead.

[0063] Furthermore, data caching and breakpoint resumption during network interruptions: When the cloud-edge network connection is interrupted, edge nodes temporarily store the data to be uploaded in a local cache. Once the network is restored, the breakpoint resumption mechanism is automatically activated to ensure no data loss. Simultaneously, policy parameters issued by the cloud are persistently stored locally on the edge nodes, and automatic synchronization verification is performed after network recovery to ensure eventual consistency. A patent proposes that when cloud platform communication is interrupted or fails, each node can guide autonomous operation within its jurisdiction based on the latest regional model. Even when the fog node level fails, the edge nodes can still independently perform local optimization, enhancing the system's robustness and reliability under abnormal operating conditions.

[0064] Furthermore, distributed data consistency is guaranteed: when multiple edge nodes simultaneously modify the same type of data (e.g., multiple edge nodes simultaneously report device status updates), the data synchronization layer employs a distributed consensus mechanism (such as timestamp-based version control or the Paxos protocol) to ensure eventual consistency of the data received by the cloud, avoiding data conflicts. Drawing inspiration from the task allocation and status synchronization method applied for by State Grid Zhejiang Electric Power, which ensures eventual consistency of distributed states through priority-aware merging logic, this invention sets different priorities for different business data. In the event of a conflict, conflicts are resolved based on priority, ensuring that data consistency meets business expectations.

[0065] Furthermore, the adaptive disaster recovery layer provides bidirectional redundant backup and automatic failover capabilities between the cloud and edge to ensure the business continuity of the integrated energy platform in single-point failure scenarios. Bidirectional redundant backup between the cloud and edge: The adaptive disaster recovery layer implements the following bidirectional backup mechanisms: Cloud backup of edge nodes: Core configuration data of edge nodes (device models, pricing models, policy rules, etc.) are redundantly backed up in the cloud. When an edge node fails and is rebuilt, the configuration can be quickly restored from the cloud. Edge node backup of the cloud: Key policy parameters issued by the cloud are persistently stored locally on the edge nodes. When the cloud-edge network is interrupted or the cloud platform fails, the edge nodes can continue to operate autonomously according to the latest policies stored locally, avoiding complete business interruption. Adaptive Disaster Recovery Switchover: When a node failure is detected, the adaptive disaster recovery layer executes the following switchover process: Edge Node Failure Switchover: When an edge node fails, the device connections and data acquisition tasks within its jurisdiction are dynamically switched to the nearest edge node, and the business containers of the failed node are automatically started on the nearest node; Cloud-Edge Network Interruption Switchover: When the cloud-edge network is interrupted, the edge node automatically switches to "offline autonomous mode" and continues to run core businesses based on locally cached policy parameters and models. After the network is restored, data synchronization and state alignment are automatically performed; Cloud Platform Failure Switchover: When the private cloud master node fails, it automatically switches to the backup private cloud node or public cloud disaster recovery node. Critical services (such as optimization scheduling engine and policy distribution service) are quickly started on the disaster recovery node, and the state is automatically synchronized after the fault is recovered.

[0066] Furthermore, disaster recovery is tiered: differentiated disaster recovery levels are set based on business importance. S-level core businesses (IoT access, optimized scheduling, policy execution) are configured with the highest level of disaster recovery protection, with an RTO (Recovery Time Objective) ≤ 30 seconds and an RPO (Recovery Point Objective) ≤ 0 seconds; A-level important businesses (Predictive Services, Service Bus) are configured with standard disaster recovery, with an RTO ≤ 2 minutes; B-level general businesses (Reporting Services, Historical Data Query) are configured with basic disaster recovery, with an RTO that can be appropriately extended. This tiered protection approach is consistent with the tiered deployment strategy of the high-availability cluster of the integrated energy software platform, ensuring that critical core businesses receive optimal recovery protection in fault scenarios.

[0067] The embodiments of the present invention also have the following technical effects: 1. Significantly reduce the overall operating costs of the integrated energy platform. By migrating non-core computing tasks (such as historical data analysis and report generation) to the public cloud through a hybrid cloud architecture, private cloud hardware investment and maintenance costs are reduced. Local preprocessing at edge nodes reduces the amount of data uploaded to the cloud, lowering network bandwidth costs. Compared to a pure private cloud deployment, total infrastructure costs can be reduced by approximately 30%-40%; compared to a pure public cloud deployment, data outbound costs can be reduced by approximately 50%-70%. Differentiated container resource pools allocate appropriate resources to different levels of business, further improving resource utilization efficiency.

[0068] 2. Significantly improves the system's real-time response capability. Edge nodes are deployed near the data source, processing real-time control commands without going through the cloud platform, reducing end-to-end latency from seconds to milliseconds. The millisecond-level real-time control of edge nodes meets the needs of high-time-sensitivity scenarios such as microgrid island switching and rapid power regulation of distributed power sources. The autonomous operation capability of edge nodes ensures that core services are not interrupted during network outages.

[0069] 3. Achieve business-aware dynamic resource scheduling and deployment optimization. The collaborative orchestration layer dynamically decides the deployment location (private cloud / public cloud / edge node) of each business module based on characteristics such as business type (real-time / non-real-time), data sensitivity (sensitive / non-sensitive), and computing load. During peak spot trading periods, optimized computing tasks can be migrated to the cloud in advance to ensure service quality during peak times; in scenarios with high data security requirements, deployment on the private cloud is mandatory. Dynamic adjustment of deployment strategies avoids the resource waste and service quality degradation problems of static configuration.

[0070] 4. Achieve unified containerized orchestration and management of edge nodes. Distributed edge nodes are managed uniformly through lightweight Kubernetes, supporting one-click deployment, remote upgrades, and elastic scaling of applications. This significantly reduces on-site work complexity and construction time, enabling a business model that productizes the edge and cloud platforms, supporting large-scale deployment. Differentiated container resource pools and a cloud-based training-edge inference model provide flexible and efficient edge computing capabilities for integrated energy services.

[0071] 5. Ensure data consistency and business continuity in cloud-edge collaboration. The data synchronization layer ensures eventual consistency of cloud and edge data through a two-way synchronization mechanism and a retransmission design for network outages. When the network is interrupted, edge nodes can operate autonomously offline based on local caching, and automatically synchronize after network recovery to avoid data loss. The adaptive disaster recovery layer provides bidirectional redundant backup between cloud and edge and automatic failover to ensure business continuity of the integrated energy platform in single-point failure scenarios. S-level core services have an RTO ≤ 30 seconds and an RPO ≤ 0 seconds, achieving an overall system availability of 99.99%.

[0072] 6. Enhance system scalability. The hybrid cloud architecture supports elastic scaling of computing resources. When business load increases, public cloud computing nodes or edge nodes can be dynamically added, expanding the platform's processing capacity without downtime. Edge nodes adopt a standardized containerized deployment solution. When adding a new project, only edge node devices need to be deployed on-site, and applications are uniformly orchestrated and deployed in the cloud, significantly shortening the launch cycle of new projects.

[0073] 7. Balancing data security and computing elasticity. The private cloud ensures that core sensitive data (device models, user information, transaction strategies) does not leave the park, meeting data security compliance requirements; the public cloud provides elastic computing capabilities to handle peak loads and complex computing tasks. The two are securely interconnected through encrypted dedicated lines or VPNs, achieving an optimal balance between security and elasticity.

[0074] 8. Reduce operational complexity and improve operational efficiency. Application deployment and upgrades for all edge nodes are completed through unified cloud orchestration, eliminating the need for on-site manual operation. Operations personnel can view the real-time operating status, resource utilization, and business load of all edge nodes through a centralized cloud monitoring platform, receive fault alarms, and remotely execute recovery operations, significantly reducing the workload of operations personnel and the frequency of on-site visits.

[0075] To achieve the above embodiments, such as Figure 2 As shown, this embodiment also provides a hybrid cloud and edge computing collaborative deployment device 10 for an integrated energy platform, including: The deployment scheduling management module 100 is used to maintain a global resource view through the collaborative orchestration layer, and dynamically decide the deployment location of each business module in the private cloud, public cloud or edge node based on business type, data sensitivity and computing load characteristics. The edge local processing module 200 is used to perform local preprocessing and real-time control of the collected field equipment data through edge nodes, and upload the preprocessed key data to the private cloud for cross-regional analysis. The cloud-edge data synchronization module 300 is used to establish a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. The adaptive disaster recovery switching module 400 is used to perform redundant backup of the core configuration data of edge nodes in the cloud through the adaptive disaster recovery layer, and dynamically switch the business tasks of the faulty node to the nearby edge node or the cloud disaster recovery node when a node failure is detected.

[0076] The hybrid cloud and edge computing collaborative deployment device for an integrated energy platform according to an embodiment of the present invention can realize business-aware dynamic resource scheduling, reduce the operating cost of the integrated energy platform, and improve the system's real-time response capability, data consistency and fault tolerance.

[0077] To implement the methods of the above embodiments, the present invention also provides a computer device, such as... Figure 3As shown, the computer device 600 includes a memory 601 and a processor 602; wherein, the processor 602 reads executable program code stored in the memory 601 to run a program corresponding to the executable program code, so as to implement the various steps of the method described above.

[0078] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing embodiments.

[0079] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0080] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for the collaborative deployment of hybrid cloud and edge computing in an integrated energy platform, characterized in that, include: S1 maintains a global resource view through a collaborative orchestration layer and dynamically decides the deployment location of each business module in a private cloud, public cloud, or edge node based on business type, data sensitivity, and computing load characteristics. S2 performs local preprocessing and real-time control of the collected field equipment data through edge nodes, and uploads the preprocessed key data to the private cloud for cross-regional analysis. S3 establishes a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. S4 uses an adaptive disaster recovery layer to redundantly back up the core configuration data of edge nodes in the cloud, and when a node failure is detected, it dynamically switches the business tasks of the failed node to nearby edge nodes or cloud disaster recovery nodes for execution.

2. The method as described in claim 1, characterized in that, S1 includes: The collaborative orchestration layer collects the CPU, memory, storage, network bandwidth, and GPU computing power resource status of private clouds, public clouds, and various edge nodes in real time, forming a unified panoramic view of resources; Based on a real-time priority strategy, services requiring millisecond-level response are forced to be deployed on edge nodes or private cloud nodes close to the data source. Based on a data security strategy, data processing services involving user privacy and trade secrets are prioritized for deployment on the private cloud. Based on a cost optimization strategy, the distribution ratio of non-core business computing tasks between the public cloud and the private cloud is dynamically adjusted. Based on a load balancing strategy, computing tasks are distributed to nodes with lower loads for execution.

3. The method as described in claim 2, characterized in that, The method further includes: When an edge node receives a computing task request, it evaluates the task's real-time requirements, data volume, and computational complexity. Query the remaining local resources on the edge node to determine if the local resources can meet the task execution conditions; if the local resources are sufficient and the real-time requirements are high, then execute the task locally on the edge node. If local resources are insufficient or the complexity of the task exceeds the capabilities of the edge node, the task will be uploaded to the private cloud for execution via the hybrid cloud gateway. If the private cloud has a high load and the data is allowed to leave the domain, the task will be further scheduled to be executed on the public cloud; The execution result is returned to the edge node that initiated the request through the data synchronization layer.

4. The method as described in claim 1, characterized in that, S2 includes: Edge nodes calculate the Hamming distance between the received raw byte stream and each protocol template using the protocol feature library they maintain. They select the protocol with the smallest Hamming distance as the parsing protocol. If the smallest Hamming distance exceeds a preset threshold, it is marked as an unknown protocol and an alarm is triggered. The parsed data is then normalized to obtain data in a uniform format. The data in a unified format is filtered, cleaned, compressed, and subjected to preliminary statistical analysis to generate aggregated key indicators; the aggregated key indicators are then uploaded to the private cloud through a data synchronization layer.

5. The method as described in claim 4, characterized in that, The method further includes: Based on formula The original data values ​​are converted into normalized values ​​within a target unified interval, where Max and Min are the dynamic maximum and minimum values ​​of the measurement point in the historical data. To unify the target range.

6. The method as described in claim 1, characterized in that, The S3 includes: The data synchronization layer uses a combination of publish-subscribe message queues and state version vectors to serialize and compress telemetry data using differential coding combined with efficient compression algorithms. It determines partitions based on the consistent hashing of data tags to ensure that data from the same device is written sequentially to the same partition. The strategy parameters sent from the cloud are parsed by the converter to extract the strategy description and target value. The control quantity is calculated according to the device characteristic model and encapsulated into a fieldbus message and sent to the edge node. When resuming transmission from a breakpoint, the edge node and the cloud calculate the checksum of the data block to be synchronized. If they match, the transmission is confirmed; otherwise, the data is retransmitted block by block.

7. The method as described in claim 1, characterized in that, The S4 includes: The adaptive disaster recovery layer redundantly backs up the core configuration data of the edge nodes in the cloud and persistently stores the key policy parameters issued by the cloud locally on the edge nodes. When an edge node failure is detected, the device connection and data acquisition tasks within the jurisdiction of the failed node are dynamically switched to the neighboring edge node, and the business containers of the failed node are automatically launched on the neighboring node. When a cloud-edge network outage is detected, the edge node automatically switches to offline autonomous mode and continues to run core business based on locally cached policy parameters and models. Once the network is restored, it will automatically perform data synchronization and state alignment. When a failure of the private cloud master node is detected, the system automatically switches to a backup private cloud node or a public cloud disaster recovery node. Critical services are quickly started on the disaster recovery node, and the status is automatically synchronized after the failure is recovered.

8. A hybrid cloud and edge computing collaborative deployment device for an integrated energy platform, characterized in that, include: The deployment scheduling and management module is used to maintain a global resource view through the collaborative orchestration layer, and dynamically decide the deployment location of each business module in the private cloud, public cloud or edge node based on business type, data sensitivity and computing load characteristics. The edge local processing module is used to perform local preprocessing and real-time control of the collected field equipment data through edge nodes, and upload the preprocessed key data to the private cloud for cross-regional analysis. The cloud-edge data synchronization module is used to establish a two-way data synchronization mechanism between the cloud and edge nodes through the data synchronization layer. When the cloud-edge network is interrupted, the edge node switches to offline autonomous mode and caches the data to be synchronized. After the network is restored, it automatically performs data synchronization and consistency verification. The adaptive disaster recovery switching module is used to perform redundant backup of the core configuration data of edge nodes in the cloud through the adaptive disaster recovery layer, and dynamically switch the business tasks of the faulty node to the nearby edge node or the cloud disaster recovery node when a node failure is detected.

9. A computer device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the method as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images