Integrated disaster recovery method, system, device and equipment based on cloud side end technology

Abstract

The invention relates to the technical field of cloud computing, and discloses an integrated disaster recovery method, system, device and equipment based on a cloud side end technology. The method comprises the steps of obtaining real-time power grid data of a power distribution network; transmitting the real-time power grid data to an edge node, so that the edge node performs short-term load prediction by using a low-order prediction model to obtain a preliminary prediction result; when the preliminary prediction result is abnormal, performing preliminary fault response by using the edge node, and generating a preliminary response result; transmitting the real-time power grid data and the preliminary response result to a cloud center node, so that the cloud center node performs long-term load prediction by using a high-order prediction model and analyzes a fault reason; and generating a recovery execution instruction according to the fault reason, and sending the execution instruction to the fault terminal. Through cooperative work of the edge node and the cloud center node, prediction accuracy is effectively enhanced, rapid positioning and recovery of faults are realized, and flexibility and comprehensiveness of power grid management are improved.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to an integrated disaster recovery method, system, device and equipment based on cloud-edge technology. Background Art

[0002] With the digital transformation of the new power system, a large number of IoT intelligent sensing devices and new entities are connected to the power grid, which brings challenges to the safe and stable operation and management of the power grid. The existing dispatching systems mainly include independent backup dispatching systems with one master and one backup, integrated independent dispatching systems for prefectures and counties, and backup dispatching systems with multiple masters and one backup. These systems usually rely on cloud processing technology and perform centralized data processing and management through cloud centers.

[0003] Although the existing dispatching system meets the needs of power grid dispatching, it still has problems such as long and complex data transmission links and low data interaction efficiency. Large amounts of data are transmitted to the cloud center, which increases storage pressure and network bandwidth burden. The increasing computing power demand of the cloud center may lead to a shortage of computing resources. In addition, the centralized processing mode that relies on the cloud center will affect the stability and reliability of the system when the network fails or the cloud service is interrupted. Summary of the invention

[0004] The present invention provides an integrated disaster recovery method, system, device and equipment based on cloud-edge technology, which effectively enhances the accuracy of prediction and realizes rapid fault location and recovery through the collaborative work of edge nodes and cloud center nodes.

[0005] In order to solve the above technical problems, the present invention provides an integrated distribution network disaster recovery method based on cloud-edge-end technology, comprising:

[0006] Obtain real-time grid data of the distribution network;

[0007] Transmitting the real-time power grid data to edge nodes so that the edge nodes use low-order prediction models to perform short-term load prediction and obtain preliminary prediction results;

[0008] When the preliminary prediction result is abnormal, using the edge node to perform a preliminary fault response and generate a preliminary response result;

[0009] Transmitting the real-time power grid data and the preliminary response results to a cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault;

[0010] A recovery execution instruction is generated according to the fault cause, and the execution instruction is sent to the fault terminal.

[0011] Furthermore, the real-time power grid data of the distribution network is obtained, specifically:

[0012] Use the local dispatching subsystem to collect real-time grid data from each main device in the distribution network;

[0013] Among them, the main equipment includes smart meters, charging piles, distributed energy and energy storage devices; the real-time power grid data includes equipment operation data, environmental data, user behavior data and system operation data.

[0014] Furthermore, the edge node uses a low-order prediction model to perform short-term load prediction and obtains preliminary prediction results, specifically:

[0015] Performing data preprocessing on the real-time power grid data; wherein the data preprocessing includes data cleaning and data format conversion;

[0016] The real-time power grid data after data preprocessing is input into the low-order prediction model for short-term load forecasting, and the preliminary prediction results are obtained, specifically:

[0017] Use time series analysis methods to perform load forecasting on real-time power grid data after data preprocessing to generate preliminary analysis results;

[0018] The multivariate regression analysis method is used to optimize the preliminary analysis results and obtain preliminary prediction results.

[0019] Further, when the preliminary prediction result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result, specifically:

[0020] When the preliminary prediction result is abnormal, matching a preliminary response operation in a preset response strategy library according to the preliminary prediction result;

[0021] executing the preliminary response operation using the edge node, and recording fault data corresponding to the preliminary response operation;

[0022] A preliminary response result is generated according to the preliminary response operation and the fault data.

[0023] Furthermore, the real-time power grid data and the preliminary response result are transmitted to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault, specifically:

[0024] Acquire fault data in the preliminary response result;

[0025] Extracting key features from the real-time power grid data and the fault data; wherein the key features include current change rate and voltage change rate;

[0026] The key features are input into the high-order prediction model to obtain the cause of the failure, which is specifically:

[0027] Using a support vector machine algorithm to identify the key features, and derive the fault type and the fault impact range;

[0028] Analyze the cause of the fault based on the fault type and the fault impact scope.

[0029] Further, generating a recovery execution instruction according to the fault cause and sending the execution instruction to the fault terminal is specifically:

[0030] Analyze the cause of the failure using the cloud center node and generate a recovery strategy;

[0031] Using the optimization algorithm of the cloud center node to process the recovery strategy and generate execution instructions;

[0032] The cloud center node is controlled to send the execution instruction to the edge node, so that the edge node forwards the execution instruction to the fault terminal.

[0033] Furthermore, the integrated distribution network disaster recovery method based on cloud-edge-device technology also includes:

[0034] After the faulty terminal receives and executes the execution instruction, collecting recovery operation data of the faulty terminal;

[0035] Analyze the fault recovery effect of the fault terminal based on the recovery operation data, and generate fault feedback information;

[0036] The low-order prediction model and the high-order prediction model are updated based on the fault feedback information.

[0037] The present invention also provides an integrated distribution network disaster recovery system based on cloud-edge-end technology, which is used to execute the above-mentioned integrated distribution network disaster recovery method based on cloud-edge-end technology, including: a local scheduling subsystem, an edge node and a cloud center node;

[0038] The local dispatching subsystem is connected to each main device in the distribution network to obtain real-time power grid data of each main device in the distribution network;

[0039] The edge node is used to perform short-term load forecasting based on the real-time power grid data using a low-order forecasting model to obtain a preliminary forecasting result; and when the preliminary forecasting result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result;

[0040] The cloud center node is used to transmit the real-time power grid data and the preliminary response result to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; and generate a recovery execution instruction based on the cause of the fault.

[0041] The present invention also provides an integrated distribution network disaster recovery device based on cloud-edge-end technology, including: a data acquisition module, a short-term prediction module, a preliminary response module, a long-term prediction module and an instruction issuing module;

[0042] The data acquisition module is used to acquire real-time power grid data of the power distribution network;

[0043] The short-term prediction module is used to transmit the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load prediction and obtain a preliminary prediction result;

[0044] The preliminary response module is used to use the edge node to perform a preliminary fault response and generate a preliminary response result when the preliminary prediction result is abnormal;

[0045] The long-term prediction module is used to transmit the real-time power grid data and the preliminary response result to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault;

[0046] The instruction issuing module is used to generate a recovery execution instruction according to the fault cause, and send the execution instruction to the fault terminal.

[0047] The present invention also provides an integrated distribution network disaster recovery device based on cloud-edge-terminal technology, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the integrated distribution network disaster recovery method based on cloud-edge-terminal technology as described above.

[0048] The present invention provides an integrated distribution network disaster recovery method, system, device and equipment based on cloud-edge technology to obtain real-time power grid data of the distribution network; transmit the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load forecasting and obtain preliminary prediction results; when the preliminary prediction result is abnormal, use the edge node to perform preliminary fault response and generate preliminary response results; transmit the real-time power grid data and preliminary response results to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; generate recovery execution instructions according to the cause of the fault, and send the execution instructions to the fault terminal. The present invention effectively enhances the accuracy of the prediction through the collaborative work of the edge node and the cloud center node, realizes the rapid location and recovery of the fault, thereby effectively reducing the scope of fault impact, and realizes long-term and short-term load forecasting and decision optimization, improving the flexibility and comprehensiveness of power grid management. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1A flow chart of an embodiment of an integrated distribution network disaster recovery method based on cloud-edge-end technology provided by the present invention;

[0050] Figure 2 A structural diagram of an embodiment of an integrated distribution network disaster recovery system based on cloud-edge-end technology provided by the present invention;

[0051] Figure 3 A structural schematic diagram of another embodiment of the integrated distribution network disaster recovery system based on cloud-edge-end technology provided by the present invention;

[0052] Figure 4 A structural schematic diagram of an embodiment of an integrated distribution network disaster recovery device based on cloud-edge-end technology provided by the present invention;

[0053] Figure 5 A structural schematic diagram of an embodiment of an integrated distribution network disaster recovery device based on cloud-edge-end technology provided by the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0056] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0057] Example 1

[0058] See also Figure 1 , is a flow chart of an embodiment of an integrated distribution network disaster recovery method based on cloud-edge-end technology provided by the present invention, the method includes steps 101 to 105, each step is specifically as follows:

[0059] Step 101: Acquire real-time power grid data of the distribution network.

[0060] Furthermore, in the first embodiment of the present invention, real-time power grid data of the distribution network is obtained, specifically:

[0061] Use the local dispatching subsystem to collect real-time grid data from each main device in the distribution network;

[0062] Among them, the main equipment includes smart meters, charging piles, distributed energy and energy storage devices; the real-time power grid data includes equipment operation data, environmental data, user behavior data and system operation data.

[0063] In the first embodiment of the present invention, the real-time grid data of the distribution network can be obtained by using the local dispatching subsystem. Specifically, the local dispatching subsystem is connected to the main equipment such as smart meters, charging piles, distributed energy and energy storage devices, and equipment operation data including voltage, current, power and energy consumption can be collected, as well as environmental data such as temperature and humidity, as well as user behavior data such as electricity usage habits and load curves, and system operation data such as switch status and alarm information.

[0064] Step 102: The real-time power grid data is transmitted to the edge node, so that the edge node uses a low-order prediction model to perform short-term load prediction and obtain a preliminary prediction result.

[0065] Furthermore, in the first embodiment of the present invention, the edge node uses a low-order prediction model to perform short-term load prediction and obtains a preliminary prediction result, which is specifically:

[0066] Performing data preprocessing on the real-time power grid data; wherein the data preprocessing includes data cleaning and data format conversion;

[0067] The real-time power grid data after data preprocessing is input into the low-order prediction model for short-term load forecasting, and the preliminary prediction results are obtained, specifically:

[0068] Use time series analysis methods to perform load forecasting on real-time power grid data after data preprocessing to generate preliminary analysis results;

[0069] The multivariate regression analysis method is used to optimize the preliminary analysis results and obtain preliminary prediction results.

[0070] In the first embodiment of the present invention, after receiving the real-time power grid data, the edge node first pre-processes the received data, including data cleaning and data format conversion, to improve the accuracy of data processing. The trained low-order prediction model is used to perform load forecasting and judge the operation status of each device. Among them, the low-order prediction model is trained based on historical power grid data, and the time series analysis method and multivariate regression analysis method are used during training. For example, the autoregressive moving average (ARIMA) model is used to perform hourly load forecasting, and the multivariate regression analysis method is used in combination with environmental data to correct and optimize the load forecasting results. The trained low-order prediction model can achieve rapid response and processing of local data, so that it can immediately respond to environmental changes and provide real-time load warnings. Based on the low-order prediction model, the edge node can generate warnings in real time, and provide scheduling management and rapid response suggestions, which greatly improves the system's response capability to emergencies and overall operating efficiency.

[0071] Step 103: When the preliminary prediction result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result.

[0072] Further, in the first embodiment of the present invention, when the preliminary prediction result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result, specifically:

[0073] When the preliminary prediction result is abnormal, matching a preliminary response operation in a preset response strategy library according to the preliminary prediction result;

[0074] executing the preliminary response operation using the edge node, and recording fault data corresponding to the preliminary response operation;

[0075] A preliminary response result is generated according to the preliminary response operation and the fault data.

[0076] In the first embodiment of the present invention, after the low-order prediction model is used to process the real-time power grid data to obtain the preliminary prediction results, it is determined whether each device has a fault such as a local load anomaly according to the preliminary prediction results. When a potential fault is detected, the edge node immediately performs a preset preliminary response operation, such as isolating the fault area, starting the backup power supply, etc., to minimize the impact of the fault on the local network.

[0077] Step 104: Transmit the real-time power grid data and the preliminary response result to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault.

[0078] Furthermore, in the first embodiment of the present invention, the real-time power grid data and the preliminary response result are transmitted to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault, specifically:

[0079] Acquire fault data in the preliminary response result;

[0080] Extracting key features from the real-time power grid data and the fault data; wherein the key features include current change rate and voltage change rate;

[0081] The key features are input into the high-order prediction model to obtain the cause of the failure, which is specifically:

[0082] Using a support vector machine algorithm to identify the key features, and derive the fault type and the fault impact range;

[0083] Analyze the cause of the fault based on the fault type and the fault impact scope.

[0084] In the first embodiment of the present invention, when the edge node performs a preliminary response operation, it records the relevant fault data and operation records and sends them to the cloud center node. After the cloud center node collects the fault data from multiple edge nodes, it integrates and preprocesses the received data, and uses a high-order prediction model to perform global long-term load forecasting. Specifically, the data processing performed by the high-order prediction model includes: extracting key features such as the current change rate and the voltage change rate, and then using the support vector machine algorithm (SVM) to classify and recognize fault types, evaluate the scope of fault impact, and analyze the root causes, such as equipment aging or operation errors.

[0085] In the first embodiment of the present invention, the training process of the high-order prediction model involves complex data processing processes and algorithms, including extracting load characteristics from a huge data set and constructing multi-dimensional feature vectors. Detailed load data modeling and training are performed using a deep neural network (DNN). The cloud center node integrates and compares the long-term load forecast results obtained using the high-order prediction model and the short-term forecast results of the edge nodes, thereby obtaining a more comprehensive perspective at the macro and micro levels, ensuring the continuity and consistency of the forecast. By combining long-term and short-term forecast results, the present invention can effectively adjust and optimize real-time response strategies, effectively improve decision-making support capabilities, and make power grid operations more accurate and efficient.

[0086] Step 105: Generate a recovery execution instruction according to the fault cause, and send the execution instruction to the fault terminal.

[0087] Further, in the first embodiment of the present invention, a recovery execution instruction is generated according to the fault cause, and the execution instruction is sent to the fault terminal, specifically:

[0088] Analyze the cause of the failure using the cloud center node and generate a recovery strategy;

[0089] Using the optimization algorithm of the cloud center node to process the recovery strategy and generate execution instructions;

[0090] The cloud center node is controlled to send the execution instruction to the edge node, so that the edge node forwards the execution instruction to the fault terminal.

[0091] In the first embodiment of the present invention, after obtaining the cause of the fault, the cloud center node can formulate a targeted recovery strategy according to the cause of the fault, such as equipment upgrade or grid parameter adjustment, so as to improve the overall stability and reliability of the system. Specifically, after the recovery strategy and operation plan are obtained through the optimization algorithm processing of the cloud center node, a specific execution instruction can be generated, and then the execution instruction is sent to the corresponding edge node. After the edge node receives the execution instruction from the cloud center, it sends the received execution instruction to the connected terminal device, such as a smart meter, distributed energy equipment, etc., to perform specific operations such as isolating the fault area and starting the backup power supply. The present invention greatly improves the emergency response capability and overall operational flexibility of the power grid through the flow of instructions from the cloud center to the edge node and then to the specific terminal, ensuring the rapid and accurate execution of the execution instruction while maintaining the high efficiency of the power grid operation and the stability of the power supply.

[0092] Furthermore, in the first embodiment of the present invention, it also includes:

[0093] After the faulty terminal receives and executes the execution instruction, collecting recovery operation data of the faulty terminal;

[0094] Analyze the fault recovery effect of the fault terminal based on the recovery operation data, and generate fault feedback information;

[0095] The low-order prediction model and the high-order prediction model are updated based on the fault feedback information.

[0096] In the first embodiment of the present invention, when the fault terminal receives an execution instruction, it executes the execution instruction and obtains the power grid data of the fault terminal after the fault is recovered, so as to analyze the effect of the fault recovery, and then update the system configuration and fault response strategy to prevent the occurrence of similar faults in the future, and use this data to adjust the primary prediction model and the above-mentioned high-order prediction model to improve the prediction accuracy of the model.

[0097] In summary, the first embodiment of the present invention provides an integrated distribution network disaster recovery method based on cloud-edge technology, which obtains real-time power grid data of the distribution network; transmits the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load forecasting and obtains preliminary prediction results; when the preliminary prediction result is abnormal, the edge node is used to perform preliminary fault response and generate preliminary response results; the real-time power grid data and the preliminary response results are transmitted to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; generates a recovery execution instruction according to the cause of the fault, and sends the execution instruction to the fault terminal. The present invention effectively enhances the accuracy of the prediction through the collaborative work of the edge node and the cloud center node, realizes the rapid location and recovery of the fault, thereby effectively reducing the scope of fault impact, and realizes long-term and short-term load forecasting and decision optimization, and improves the flexibility and comprehensiveness of power grid management.

[0098] Example 2

[0099] See also Figure 2 , is a structural diagram of an embodiment of an integrated distribution network disaster recovery system based on cloud-edge-end technology provided by the present invention, which is used to execute the above-mentioned integrated distribution network disaster recovery method based on cloud-edge-end technology, including a local scheduling subsystem 201, an edge node 202 and a cloud center node 203;

[0100] The local dispatching subsystem 201 is connected to each main device in the distribution network to obtain real-time power grid data of each main device in the distribution network;

[0101] The edge node 202 is used to perform short-term load forecasting based on the real-time power grid data using a low-order forecasting model to obtain a preliminary forecasting result; and when the preliminary forecasting result is abnormal, perform a preliminary fault response using the edge node and generate a preliminary response result;

[0102] The cloud center node 203 is used to transmit the real-time power grid data and the preliminary response results to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; and generate recovery execution instructions based on the cause of the fault.

[0103] In the second embodiment of the present invention, the prefectural-level dispatching subsystem 201 is connected to the main equipment such as smart meters, charging piles, distributed energy and energy storage devices, and collects a large amount of equipment data generated by these subject equipment, including smart meter data, charging pile data, energy storage system data and distributed energy data. In addition, the prefectural-level dispatching subsystem 201 can also collect a large amount of system data generated by the system in coordinated control, including monitoring data, operation data, environmental data and alarm information. At the same time, the prefectural-level dispatching subsystem 201 also records user behavior data, such as electricity usage habits, load curves, identity information and contract information. The prefectural-level dispatching subsystem 201 also records the event data generated by the system, including fault time, fault type, fault location and repair time. Through these data, the prefectural-level dispatching subsystem 201 can achieve comprehensive monitoring, optimized dispatching and rapid response to ensure efficient, safe and stable operation of the power system.

[0104] In the second embodiment of the present invention, the edge node 202 is responsible for data processing and storage, real-time calculation and rapid decision-making. Specifically, data is obtained from the local dispatch subsystem 201, preprocessed, and the data is classified, and the classified data is stored in the local storage unit after protocol conversion. In addition, the edge node 202 has real-time computing functions, such as load forecasting, power generation forecasting, efficiency analysis and low-order prediction model training. The rapid decision-making function includes dispatching management, fault warning, meteorological warning and status monitoring. Through these functions, the edge node 202 can improve data processing efficiency and system response speed, and support the stable and efficient operation of the power grid. Among them, the low-order prediction model refers to a model with low computational complexity and high real-time requirements, which is suitable for scenarios that require rapid response such as preliminary fault detection and short-term load forecasting.

[0105] Optionally, the edge node 202 mainly performs real-time or near real-time decision-making tasks, uses its rapid response capabilities to process instant data, and provides real-time load warnings and emergency response suggestions. For example, when the system detects an abnormal increase in load, the edge node 202 can immediately analyze the data and predict the load trend in the short term. Based on this information, the system can quickly adjust the grid operation parameters or activate the demand response strategy to avoid potential power supply problems or overload conditions. This fast and local decision support is the key to maintaining grid stability and responding to emergencies.

[0106] In the second embodiment of the present invention, the cloud center node 203 is responsible for the collection and unified scheduling of computing resources and data, and has data collection and storage functions, high-order model training functions, computing resource monitoring functions and communication collaborative scheduling functions. Large-scale data from the edge node 202 is collected and managed through a distributed storage system, including real-time monitoring data, user behavior data and system operation data. Using the big data platform and artificial intelligence platform, the cloud center node 203 performs high-order model training, and performs in-depth analysis and modeling of the data to improve the accuracy of prediction and decision-making. In addition, the cloud center node 203 monitors and manages computing resources through cloud native technology, realizes virtualization and dynamic scheduling of server resources, and coordinates and schedules communication resources to ensure efficient transmission of data between the edge node 202 and the cloud center. Through these functions, the cloud center node 203 realizes centralized management and optimization of power system data and resources, and improves the intelligence level and operation efficiency of the system. Among them, the high-order model refers to a model with high computational complexity and large data processing and model training requirements. It is suitable for processing large data sets, performing deep learning training, complex statistical analysis and long-term trend prediction, etc.

[0107] Optionally, the cloud center node 203 is responsible for formulating fault recovery strategies and performing advanced calculations and decisions, and issuing instructions to be executed by the edge node 202. The cloud center node 203 analyzes the fault data collected from the edge node 202 to formulate update strategies for resource scheduling and parameter configuration, such as increasing backup power output and mobilizing power support from neighboring areas, and reconfiguring grid operation parameters, including adjusting transformer taps, modifying line switch status, and voltage control settings.

[0108] Optionally, the cloud center node 203 is mainly used to formulate a more comprehensive and long-term grid operation strategy. For example, by analyzing long-term load data and environmental trends, the cloud center node 203 can predict changes in power demand in the next few weeks or months, guide power market purchasing strategies, optimize power generation resource allocation, and even support infrastructure investment decisions. The cloud center node 203 can also integrate and analyze data from multiple edge nodes 202 to provide a macro perspective, thereby making grid operations more efficient and economical.

[0109] In the second embodiment of the present invention, the interaction between the local dispatching subsystem 201 and the edge node 202 includes the local dispatching subsystem 201 uploading data to the edge node 202, and the edge node 202 sending control instructions to the local dispatching subsystem 201; similarly, the interaction between the edge node 202 and the cloud center node 203 includes the edge node 202 uploading data to the cloud center node 203, and the cloud center node 203 sending control instructions to the edge node 202. Through the synergy of the rapid response of the edge node 202 and the deep analysis of the cloud center node 203, the system can achieve comprehensive optimization from immediate response to long-term planning, providing powerful decision-making support for power grid operations. Regularly updating the model enables the system to continuously adapt to new data patterns and operational changes, ensuring the timeliness and accuracy of decision-making, thereby critically improving the intelligence and automation level of the power grid.

[0110] In the second embodiment of the present invention, after the fault is restored, the cloud center node 203 continues to monitor the power grid status, analyzes the effect of the fault recovery, and updates the system configuration and fault response strategy to prevent the occurrence of similar faults in the future. The fault detection model and response strategy of the edge node 202 are adjusted according to the latest operation data.

[0111] As an example of the second embodiment of the present invention, see Figure 3, is a structural diagram of another embodiment of the integrated distribution network disaster recovery system based on cloud-edge-end technology provided by the present invention. The prefectural-level distribution network dispatching system is a collection of equipment terminals containing a large number of new entities, and has various sensors and controllers deployed in the local dispatching system, such as voltage sensors, current sensors, and switch controllers. The prefectural-level distribution network dispatching system is used to obtain operating data such as microgrids, electric vehicles, and new energy, and connect to edge servers through communication base stations to receive and transmit data. The edge node includes an edge server, an edge gateway, a communication base station, and a control device. Data synchronization between edge servers, the edge server integrates a computing unit, a storage unit, and a network interface, and runs and manages applications through Docker, which is responsible for the acquisition, aggregation, and preliminary processing of real-time data. The edge gateway converts data of different protocols into a standard protocol and performs preliminary processing to reduce the amount of data transmission. The communication base station ensures efficient transmission of data between nodes. The control device includes a remote control unit RTU, which is used to remotely monitor and control equipment, manage intelligent switches and circuit breakers, and protect and manage power equipment. The cloud center node includes a cloud server processing center, a data center, and network equipment. The cloud server processing center includes a big data platform Hadoop and an artificial intelligence platform TensorFlow for large-scale data processing and analysis. The data center uses the distributed storage system Ceph to provide large-scale data storage and management functions. Through cloud native technology, a data computing, storage and transmission technology support system with server resource virtualization, data standardization and business application service is established to achieve efficient data management and processing.

[0112] In summary, the first embodiment of the present invention provides an integrated distribution network disaster recovery system based on cloud-edge-end technology, including a local dispatching subsystem, edge nodes, and cloud center nodes. The system introduces cloud-edge-end collaborative technology into the dispatching and operation management of distribution networks, solving the resource scheduling, business calculation, and data interaction needs of multiple business entities. The real-time computing capability of the edge nodes enables the dispatching terminal to respond quickly to instructions, while the powerful data processing capability of the cloud center ensures the efficient execution of complex tasks. Distributed data processing and storage are implemented between edge nodes and cloud center nodes, which significantly improves data processing and storage efficiency, reduces bandwidth pressure and delays in data transmission, optimizes task allocation and resource utilization, and improves the real-time and flexibility of the system.

[0113] Example 3

[0114] See also Figure 4 , is a structural diagram of an embodiment of an integrated distribution network disaster recovery device based on cloud-edge-end technology provided by the present invention, the device includes a data acquisition module 301, a short-term prediction module 302, a preliminary response module 303, a long-term prediction module 304 and an instruction issuing module 305;

[0115] The data acquisition module 301 is used to acquire real-time power grid data of the distribution network;

[0116] The short-term prediction module 302 is used to transmit the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load prediction and obtain a preliminary prediction result;

[0117] The preliminary response module 303 is used to use the edge node to perform a preliminary fault response and generate a preliminary response result when the preliminary prediction result is abnormal;

[0118] The long-term prediction module 304 is used to transmit the real-time power grid data and the preliminary response results to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault;

[0119] The instruction issuing module 305 is used to generate a recovery execution instruction according to the fault cause, and send the execution instruction to the fault terminal.

[0120] Furthermore, in the third embodiment of the present invention, real-time power grid data of the distribution network is obtained, specifically:

[0121] Use the local dispatching subsystem to collect real-time grid data from each main device in the distribution network;

[0122] Among them, the main equipment includes smart meters, charging piles, distributed energy and energy storage devices; the real-time power grid data includes equipment operation data, environmental data, user behavior data and system operation data.

[0123] Furthermore, in the third embodiment of the present invention, the edge node uses a low-order prediction model to perform short-term load prediction and obtains a preliminary prediction result, which is specifically:

[0124] Performing data preprocessing on the real-time power grid data; wherein the data preprocessing includes data cleaning and data format conversion;

[0125] The real-time power grid data after data preprocessing is input into the low-order prediction model for short-term load forecasting, and the preliminary prediction results are obtained, specifically:

[0126] Use time series analysis methods to perform load forecasting on real-time power grid data after data preprocessing to generate preliminary analysis results;

[0127] The multivariate regression analysis method is used to optimize the preliminary analysis results and obtain preliminary prediction results.

[0128] Further, in the third embodiment of the present invention, when the preliminary prediction result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result, specifically:

[0129] When the preliminary prediction result is abnormal, matching a preliminary response operation in a preset response strategy library according to the preliminary prediction result;

[0130] executing the preliminary response operation using the edge node, and recording fault data corresponding to the preliminary response operation;

[0131] A preliminary response result is generated according to the preliminary response operation and the fault data.

[0132] Furthermore, in a third embodiment of the present invention, the real-time power grid data and the preliminary response result are transmitted to a cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault, specifically:

[0133] Acquire fault data in the preliminary response result;

[0134] Extracting key features from the real-time power grid data and the fault data; wherein the key features include current change rate and voltage change rate;

[0135] The key features are input into the high-order prediction model to obtain the cause of the failure, which is specifically:

[0136] Using a support vector machine algorithm to identify the key features, and derive the fault type and the fault impact range;

[0137] Analyze the cause of the fault based on the fault type and the fault impact scope.

[0138] Further, in the third embodiment of the present invention, a recovery execution instruction is generated according to the fault cause, and the execution instruction is sent to the fault terminal, specifically:

[0139] Analyze the cause of the failure using the cloud center node and generate a recovery strategy;

[0140] Using the optimization algorithm of the cloud center node to process the recovery strategy and generate execution instructions;

[0141] The cloud center node is controlled to send the execution instruction to the edge node, so that the edge node forwards the execution instruction to the fault terminal.

[0142] Furthermore, in a third embodiment of the present invention, the integrated distribution network disaster recovery method based on cloud-edge-device technology further includes:

[0143] After the faulty terminal receives and executes the execution instruction, collecting recovery operation data of the faulty terminal;

[0144] Analyze the fault recovery effect of the fault terminal based on the recovery operation data, and generate fault feedback information;

[0145] The low-order prediction model and the high-order prediction model are updated based on the fault feedback information.

[0146] In summary, the third embodiment of the present invention provides an integrated distribution network disaster recovery device based on cloud-edge technology, which obtains real-time power grid data of the distribution network based on the organic combination of modules; transmits the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load forecasting and obtains preliminary prediction results; when the preliminary prediction result is abnormal, the edge node is used to perform preliminary fault response and generate preliminary response results; the real-time power grid data and the preliminary response results are transmitted to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; generates a recovery execution instruction according to the cause of the fault, and sends the execution instruction to the fault terminal. The present invention effectively enhances the accuracy of the prediction through the collaborative work of the edge node and the cloud center node, realizes the rapid location and recovery of the fault, thereby effectively reducing the scope of fault impact, and realizes long-term and short-term load forecasting and decision optimization, and improves the flexibility and comprehensiveness of power grid management.

[0147] The present invention also provides an integrated distribution network disaster recovery device based on cloud-edge-end technology, see Figure 5 , is a structural diagram of an embodiment of an integrated distribution network disaster recovery device based on cloud-edge-terminal technology provided by the present invention. The device can be a server, including a processor, a memory and a network interface connected through a system bus, wherein the memory can include a non-volatile storage medium and an internal memory.

[0148] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0149] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, which, when executed, can enable the processor to execute any one of the integrated distribution network disaster recovery methods based on cloud-edge-end technology.

[0150] The internal memory provides an environment for the operation of computer programs in non-volatile storage media. When the computer program is executed by the processor, the processor can execute any integrated distribution network disaster recovery method based on cloud-edge-terminal technology.

[0151] The network interface is used for network communication, such as sending assigned tasks.

[0152] Those skilled in the art will understand that Figure 5The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0153] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0154] The specific embodiments described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. An integrated distribution network disaster recovery method based on cloud-edge-end technology, characterized in that: include: Obtain real-time grid data of the distribution network; Transmitting the real-time power grid data to edge nodes so that the edge nodes use low-order prediction models to perform short-term load prediction and obtain preliminary prediction results; When the preliminary prediction result is abnormal, using the edge node to perform a preliminary fault response and generate a preliminary response result; Transmitting the real-time power grid data and the preliminary response results to a cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault; A recovery execution instruction is generated according to the fault cause, and the execution instruction is sent to the fault terminal.

2. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 1 is characterized in that: The real-time power grid data of the distribution network is obtained as follows: Use the local dispatching subsystem to collect real-time grid data from each main device in the distribution network; Among them, the main equipment includes smart meters, charging piles, distributed energy and energy storage devices; the real-time power grid data includes equipment operation data, environmental data, user behavior data and system operation data.

3. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 2 is characterized in that: The edge node uses a low-order prediction model to perform short-term load prediction and obtains preliminary prediction results, specifically: Performing data preprocessing on the real-time power grid data; wherein the data preprocessing includes data cleaning and data format conversion; The real-time power grid data after data preprocessing is input into the low-order prediction model for short-term load forecasting, and the preliminary prediction results are obtained, specifically: Use time series analysis methods to perform load forecasting on real-time power grid data after data preprocessing to generate preliminary analysis results; The multivariate regression analysis method is used to optimize the preliminary analysis results and obtain preliminary prediction results.

4. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 3 is characterized in that: When the preliminary prediction result is abnormal, the edge node is used to perform a preliminary fault response and generate a preliminary response result, specifically: When the preliminary prediction result is abnormal, matching a preliminary response operation in a preset response strategy library according to the preliminary prediction result; executing the preliminary response operation using the edge node, and recording fault data corresponding to the preliminary response operation; A preliminary response result is generated according to the preliminary response operation and the fault data.

5. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 4 is characterized in that: The real-time power grid data and the preliminary response result are transmitted to the cloud center node so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault, specifically: Acquire fault data in the preliminary response result; Extracting key features from the real-time power grid data and the fault data; wherein the key features include current change rate and voltage change rate; The key features are input into the high-order prediction model to obtain the cause of the failure, which is specifically: Using a support vector machine algorithm to identify the key features, and derive the fault type and the fault impact range; Analyze the cause of the fault based on the fault type and the fault impact scope.

6. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 5 is characterized in that: The generating of the recovery execution instruction according to the fault cause and sending the execution instruction to the fault terminal is specifically: Analyze the cause of the failure using the cloud center node and generate a recovery strategy; Using the optimization algorithm of the cloud center node to process the recovery strategy and generate execution instructions; The cloud center node is controlled to send the execution instruction to the edge node, so that the edge node forwards the execution instruction to the fault terminal.

7. The integrated distribution network disaster recovery method based on cloud-edge-end technology according to claim 6 is characterized in that: Also includes: After the faulty terminal receives and executes the execution instruction, collecting recovery operation data of the faulty terminal; Analyze the fault recovery effect of the fault terminal based on the recovery operation data, and generate fault feedback information; The low-order prediction model and the high-order prediction model are updated based on the fault feedback information.

8. An integrated distribution network disaster recovery system based on cloud-edge-end technology, characterized in that: Used to execute the integrated distribution network disaster recovery method based on cloud-edge-end technology as described in any one of claims 1 to 7, including a local scheduling subsystem, an edge node and a cloud center node; The local dispatching subsystem is connected to each main device in the distribution network to obtain real-time power grid data of each main device in the distribution network; The edge node is used to perform short-term load forecasting based on the real-time power grid data using a low-order forecasting model to obtain a preliminary forecasting result; and when the preliminary prediction result is abnormal, using the edge node to perform a preliminary fault response and generate a preliminary response result; The cloud center node is used to transmit the real-time power grid data and the preliminary response result to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load forecasting and analyze the cause of the fault; and generate a recovery execution instruction based on the cause of the fault.

9. An integrated distribution network disaster recovery device based on cloud-edge-end technology, characterized in that: include: Data acquisition module, short-term prediction module, initial response module, long-term prediction module and instruction issuing module; The data acquisition module is used to acquire real-time power grid data of the power distribution network; The short-term prediction module is used to transmit the real-time power grid data to the edge node, so that the edge node uses a low-order prediction model to perform short-term load prediction and obtain a preliminary prediction result; The preliminary response module is used to use the edge node to perform a preliminary fault response and generate a preliminary response result when the preliminary prediction result is abnormal; The long-term prediction module is used to transmit the real-time power grid data and the preliminary response result to the cloud center node, so that the cloud center node uses a high-order prediction model to perform long-term load prediction and analyze the cause of the fault; The instruction issuing module is used to generate a recovery execution instruction according to the fault cause, and send the execution instruction to the fault terminal.

10. An integrated distribution network disaster recovery device based on cloud-edge-end technology, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the integrated distribution network disaster recovery method based on cloud-edge technology as described in any one of claims 1 to 7.

Images