Power distribution network information physical system risk analysis method, system and equipment based on grid division and storage medium

Abstract

The invention discloses a power distribution network information physical system risk analysis method, system and device based on grid division and a storage medium, and relates to the field of power system risk analysis. Dividing operation data of a physical system and an information system of the power distribution network into unit grids and grid groups; searching potential faults, and generating a fault event set; performing batch simulation on fault events based on a multi-thread programming framework by adopting a grid group parallel computing technology, quantifying load loss under different faults and generating load loss data; based on load loss data and introduced defense resource parameters, calculating a feeder risk coefficient and a system risk coefficient, and quantitatively analyzing a risk propagation path; according to the method, accurate risk positioning and rapid early warning can be realized, multi-dimensional coupling risks are comprehensively covered, the risk assessment real-time performance of the power distribution network is greatly improved, and safe and stable operation of a novel power system is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power system risk analysis, in particular to a power distribution network cyber-physical system risk analysis method, system, device and storage medium based on grid division. BACKGROUND

[0002] In recent years, with the wide application of digital power distribution network cyber-physical system in power system, the interaction and integration between information side and physical side become more and more complex. However, such integration is easy to cause information security vulnerabilities and communication delay risks on the information side. Moreover, it will form a complex interaction mechanism with the power flow exceeding the limit on the physical side, and then lead to frequent cross-domain cascading failures.

[0003] At present, most methods adopt centralized modeling ideas, lack of fine description of spatial grid level risk propagation path, and cannot accurately capture the dynamic coupling effect between the information-physical dual domains, nor can they reveal the nonlinear evolution mechanism of cascading failures. Specifically, it is often difficult to identify the vulnerability of local key nodes by taking larger areas such as substations and feeder lines as evaluation units, resulting in insufficient risk positioning accuracy. It also ignores the synergistic effect of information attacks and physical failures, as well as the potential conflicts between different communication protocols and control strategies, and does not comprehensively cover multi-dimensional coupled risks. Moreover, the complexity is relatively high, which cannot meet the rapid risk assessment requirements in the scenario of high proportion of new energy access in new power system and frequent interaction between source, network, load and storage, and the real-time performance is difficult to guarantee. SUMMARY

[0004] In view of the above problems, the present application is proposed.

[0005] Therefore, the technical problem solved by the present application is that the traditional power distribution network cyber-physical system risk analysis method has the limitations of insufficient risk positioning accuracy, incomplete coverage of multi-dimensional coupled risks, and difficult to guarantee real-time performance, and cannot accurately capture the dynamic coupling effect between information-physical dual domains and reveal the nonlinear evolution mechanism of cascading failures.

[0006] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a power distribution network cyber-physical system risk analysis method based on grid division, comprising: Analyzing the relevant data of the power distribution network, establishing the spatio-temporal correspondence of multi-source data, and dividing the operation data of the power distribution network physical system and information system into multiple unit grids and grid groups based on topological coupling analysis; Based on the grid division result, considering the threats faced by the power distribution network cyber-physical system, automatically searching for potential failures, and generating three types of fault event sets; With the generated fault event set, a sub-grid group parallel computing technology is adopted, fault events are simulated in batches based on a multi-thread programming framework, load loss under different faults is quantified, and load loss data is generated; Based on the load loss data and introduced defense resource parameters, a risk assessment model is constructed, feeder risk coefficients and system risk coefficients are calculated, and risk propagation paths are quantitatively analyzed.

[0007] As an optimal solution of the grid division-based power distribution network information physical system risk analysis method, wherein: The power distribution network related data is analyzed, a multi-source data space-time correspondence relationship is established, and the operation data of the power distribution network physical system and the information system are divided into a plurality of unit grids and grid groups based on topological coupling analysis, and the power distribution network physical system and the information system are divided into a plurality of unit grids and grid groups. All devices except switches are aggregated into the same unit grid, the unit grid is taken as the smallest risk assessment unit, a plurality of physical devices are contained, and the unit grids are connected only through switches; the grid group is composed of a group of unit grids connected through switches, and there is no electrical connection relationship between different grid groups.

[0008] The beneficial effects of the preferred technical solution are that the devices except switches are aggregated into the unit grid, the smallest risk assessment unit is defined, the power distribution network is facilitated for fine risk assessment, there is no electrical connection relationship between different grid groups, a basis is provided for subsequent parallel computing and analysis, and the analysis efficiency is improved.

[0009] As an optimal solution of the grid division-based power distribution network information physical system risk analysis method, wherein: The power distribution network related data is analyzed, a multi-source data space-time correspondence relationship is established, and the operation data of the power distribution network physical system and the information system are divided into a plurality of unit grids and grid groups based on topological coupling analysis, and the power distribution network physical system and the information system are divided into a plurality of unit grids and grid groups. In the fault positioning, isolation and recovery operation, the internal structure of the unit grid is ignored, and only whether the unit grid is faulty is determined; for the load transfer problem in fault recovery, all feeders having an electrical connection relationship are merged into one grid group, so that each grid group can be analyzed in parallel and does not interfere with each other; The physical topological data and the information topological data of the power distribution network are analyzed, the mapping relationship between the grid switches outside the unit grid and the FTU is established, and the real-time load data of each grid is labeled.

[0010] The advantages of this preferred technical solution are as follows: ignoring the internal structure of the cell grid during fault handling simplifies the fault diagnosis process and improves the speed of fault handling; merging electrically connected feeders into a grid group and enabling parallel analysis improves the efficiency of fault recovery and load transfer; establishing a mapping relationship between switches and FTUs and labeling real-time load data provides more comprehensive information for accurate risk assessment and effective fault handling.

[0011] As a preferred scheme for risk analysis of distribution network cyber-physical systems based on grid partitioning, wherein: Based on the grid partitioning results, and considering the threats faced by the cyber-physical system of the distribution network, the system automatically searches for potential faults and generates three sets of fault events, including: The scenario is initialized based on the operating status of the distribution network, setting information attack types and physical attack types; each unit grid is traversed to automatically generate three types of fault event sets: In response to possible abnormal situations in communication terminals, all information nodes are searched to generate a set of information failure events. The events in the set of information failure events are manifested as the communication terminal being unable to respond to controlled commands due to a specific attack.

[0012] As a preferred scheme for risk analysis of distribution network cyber-physical systems based on grid partitioning, wherein: The automatic search for potential faults and generation of three types of fault event sets based on grid partitioning results, considering the threats faced by the cyber-physical system of the distribution network, also includes: Traverse all cell grids to generate a set of physical fault events. Each physical fault in the set of physical fault events corresponds one-to-one with a cell grid. The fault manifests as a short circuit in the feeder segment, which leads to load loss in the system. The events in the generated information fault event set and physical fault event set are combined to obtain the collaborative fault event set.

[0013] As a preferred scheme for risk analysis of distribution network cyber-physical systems based on grid partitioning, wherein: The process of utilizing the generated set of fault events, employing grid-based group parallel computing technology, and performing batch simulations of the fault events based on a multi-threaded programming framework to quantify load loss under different faults and generate load loss data includes: Three types of fault events are simulated using the generated set of fault events: Physical Fault Simulation: Simulates the process of system triggering feeder automation control to locate, isolate, and restore the fault after a line short-circuit fault, analyzes the impact of equipment faults on power flow distribution in the distribution network, and calculates the load loss after fault recovery; Information Fault Simulation: Simulate the interference of information node anomalies on the main station's fault location, isolation, and recovery logic, and calculate the corresponding load loss; Collaborative fault simulation: Simulates the feeder automation process when a line short-circuit fault and an information node anomaly occur simultaneously, and calculates the load loss after fault recovery.

[0014] The advantages of this preferred technical solution are as follows: the use of grid-based parallel computing technology and multi-threaded programming framework for batch simulation improves simulation efficiency; the simulation and quantification of load loss for three types of fault events enable a comprehensive and accurate assessment of the impact of different faults on the distribution network.

[0015] As a preferred scheme for risk analysis of distribution network cyber-physical systems based on grid partitioning, wherein: The risk assessment model, constructed based on load loss data and introduced defense resource parameters, calculates feeder risk coefficients and system risk coefficients, and quantifies the risk propagation path, including: Based on the relationship between defense resources and defense effectiveness and the failure probability integral rule, the failure probability of a successful attack is calculated when each object is attacked under different defense resource configurations, where the amount of defense resources invested is positively correlated with the defense effectiveness. By using fault severity quantification rules and combining load loss data, the severity of physical equipment failures is calculated; based on the fault probability and severity of each physical node, the feeder risk coefficient of the current feeder is calculated; considering the weight of each feeder and combining the feeder risk coefficient, the system risk coefficient is calculated.

[0016] Secondly, the present invention provides a risk analysis system for a distribution network cyber-physical system based on grid partitioning, comprising: The power distribution network data grid partitioning module is used to parse relevant data of the power distribution network, establish the spatiotemporal correspondence of multi-source data, and divide the operation data of the power distribution network physical system and information system into multiple unit grids and grid groups based on topology coupling analysis. The potential fault event generation module is used to automatically search for potential faults and generate three types of fault event sets based on the grid partitioning results and considering the threats faced by the cyber-physical system of the distribution network. The fault load loss quantification module is used to perform batch simulation of fault events using the generated fault event set, employing grid-based group parallel computing technology and a multi-threaded programming framework, to quantify the load loss under different faults and generate load loss data. The risk assessment and propagation analysis module is used to build a risk assessment model based on load loss data and introduced defense resource parameters, calculate feeder risk coefficients and system risk coefficients, and quantitatively analyze risk propagation paths.

[0017] Thirdly, the present invention provides a computer device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the risk analysis method for the distribution network cyber-physical system based on grid partitioning.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of a grid-based cyber-physical system risk analysis method for power distribution networks.

[0019] The beneficial effects of this invention are as follows: The grid-based risk analysis method for distribution network cyber-physical systems has significant effects in practical applications; through hierarchical modeling of unit grids and grid groups, the accuracy of risk source location can be improved from the feeder level to the distribution area level, enabling early warning and rapid source tracing in the early stages of faults, and accurately identifying the vulnerability of local key nodes; simultaneously considering information attack, physical attack, and collaborative attack scenarios, it comprehensively depicts the risk propagation chain, overcoming the limitations of traditional single-dimensional analysis, and effectively addressing the complex interactive effects of information-side risks such as information security vulnerabilities and communication delays with physical-side risks such as power flow exceeding limits; the use of grid group parallel simulation technology significantly shortens the time required for risk assessment of large-scale distribution networks, meets the needs of online real-time analysis, and ensures rapid risk assessment in scenarios with high proportions of new energy access and frequent interaction between power sources, grids, loads, and storage in new power systems. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is an overall flowchart of a risk analysis method for a distribution network cyber-physical system based on grid partitioning provided by the present invention.

[0022] Figure 2 This is a schematic diagram of the risk assessment results of a distribution network cyber-physical system based on a grid partitioning-based risk analysis method provided by the present invention. Detailed Implementation

[0023] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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 protection scope of the present invention.

[0024] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a risk analysis method for a distribution network cyber-physical system based on grid partitioning, including: S1: Analyze relevant data of the distribution network, establish spatiotemporal correspondence of multi-source data, and divide the operation data of the physical system and information system of the distribution network into multiple unit grids and grid groups based on topology coupling analysis; S2: Based on the grid partitioning results, considering the threats faced by the cyber-physical system of the distribution network, automatically search for potential faults and generate three types of fault event sets; S3: Using the generated set of fault events, the grid-based parallel computing technology is employed to perform batch simulations of the fault events based on a multi-threaded programming framework, quantifying the load loss under different faults and generating load loss data. S4: Based on load loss data and introduced defense resource parameters, a risk assessment model is constructed to calculate the feeder risk coefficient and system risk coefficient, and to quantitatively analyze the risk propagation path.

[0025] It should be noted that through steps S1-S4, a hierarchical and regional refined analysis of the risks of the distribution network cyber-physical system is achieved. This not only accurately quantifies risks but also provides timely warnings when risks occur and formulates effective prevention and control decisions based on the characteristics of risk propagation, thus effectively ensuring the safe and stable operation of the distribution network in complex operating environments.

[0026] Example 2, refer to Figures 1-2 As an embodiment of the present invention, based on the previous embodiment, a risk analysis method for a distribution network cyber-physical system based on grid partitioning is provided, including: In this embodiment, step S1 above involves parsing distribution network-related data, establishing spatiotemporal correspondences of multi-source data, and dividing the operational data of the distribution network physical system and information system into multiple unit grids and grid groups based on topology coupling analysis, including: Based on the physical topology data, information topology data, and real-time measurement data of the distribution network, a spatiotemporal correspondence between multi-source data is established. Based on topology coupling analysis, the distribution network is divided into a two-layer structure of unit grid and grid group: the unit grid serves as the smallest risk assessment unit, containing multiple physical devices (such as loads, connecting lines, etc.), and the unit grids are connected by switches; the grid group consists of a group of unit grids connected by switches, and there is no electrical connection between different grid groups.

[0027] Furthermore, all equipment except switches is aggregated into the same cell grid, with cell grids connected only by switches. During fault location, isolation, and recovery operations, the internal structure of the cell grid can be temporarily ignored; only the fault status of the cell grid needs to be determined, thus simplifying the analysis process. For load transfer issues involved in fault recovery, all feeders with electrical connections are merged into a single grid group, allowing for parallel analysis between grid groups without interference. Simultaneously, by parsing the physical and information topology data of the distribution network, a mapping relationship is established between distribution network switches (physical equipment) and FTUs (Feeder Terminal Units) (information nodes) outside the cell grid, and the real-time load data of each grid is labeled.

[0028] In another possible implementation, establishing the spatiotemporal correspondence of multi-source data can be achieved using data mining techniques. This involves in-depth mining and analysis of data from different data sources in the power distribution network to identify the inherent temporal and spatial relationships between the data. For example, association rule mining algorithms can be used to analyze the correlation between voltage, current, and other data from different monitoring points at different times, thereby establishing an accurate spatiotemporal correspondence. Simultaneously, combining this with a Geographic Information System (GIS) allows for the association of the spatial location information of physical equipment with operational data, further refining the spatiotemporal correspondence.

[0029] In another possible implementation, establishing spatiotemporal correspondences for multi-source data can also involve introducing time series analysis methods to model and analyze time series data from various data sources in the distribution network. By learning from historical data, future data trends can be predicted, thereby establishing spatiotemporal correspondences for multi-source data. For example, using the ARIMA model to predict load data and matching the prediction results with other relevant data establishes a dynamic spatiotemporal correspondence. Furthermore, data fusion techniques can be used to combine different types of time series data, improving the accuracy of spatiotemporal correspondences.

[0030] In another possible implementation, when partitioning operational data based on topological coupling analysis, graph theory methods can be used to model the physical and information topology of the distribution network. Physical devices and information nodes are abstracted as nodes in the graph, and connections between devices and information transmission paths are abstracted as edges. By analyzing topological characteristics such as graph connectivity and clustering coefficients, the partitioning boundaries of unit grids and grid groups are determined. For example, a community detection algorithm can be used to divide the distribution network into different communities, with each community serving as a unit grid or grid group. Simultaneously, the partitioning results are adjusted and optimized based on the function and importance of the devices.

[0031] In another possible implementation, when partitioning operational data based on topology coupling analysis, the distribution network can also be partitioned based on electrical distance and information transmission delay. The electrical distance between physical devices and the transmission delay between information nodes are calculated, and devices and nodes with shorter electrical distances and lower information transmission delays are grouped into the same cell grid or grid group. For example, the shortest path algorithm is used to calculate the electrical distance, and the transmission delay is calculated based on the information transmission protocol and network topology. Furthermore, the operating status and reliability of the devices can be considered, and the partitioning results can be dynamically adjusted.

[0032] In another possible implementation, when aggregating devices other than switches into the same cell grid, aggregation can be based on the type and function of the devices. Devices with similar functions, such as transformers and capacitors, can be aggregated into the same cell grid. Simultaneously, considering the geographical location of the devices, adjacent devices can be grouped into the same cell grid. For example, within a substation area, all transformers and related auxiliary equipment can be aggregated into a single cell grid. Furthermore, the aggregation results can be further subdivided and adjusted based on the operating parameters and status of the devices.

[0033] In another possible implementation, when aggregating devices other than switches into the same cell grid, aggregation can also be based on the fault correlation of the devices. The fault propagation relationships between devices are analyzed, and devices that are likely to affect each other are aggregated into the same cell grid. For example, fault tree analysis can be used to identify the causal relationships of device failures, and devices with direct causal relationships are grouped into the same cell grid. Simultaneously, the aggregation results are optimized by combining the device's maintenance history and reliability data, improving the rationality and practicality of the cell grid.

[0034] In this embodiment, in step S2 above, based on the grid partitioning results and considering the threats faced by the distribution network cyber-physical system, potential faults are automatically searched, and three types of fault event sets are generated, including: The scenario initialization settings are performed based on the distribution network operation status, including setting information attack types and physical attack types, traversing each unit grid, and automatically generating three types of fault event sets: Information failure event set: A communication terminal is subjected to a DoS attack and is unable to respond to controlled commands; an information failure event set can be generated by searching all information nodes.

[0035] In another possible implementation, when searching all information nodes to generate a set of information fault events, active probing techniques can be employed. Probe data packets are periodically sent to all information nodes in the distribution network to monitor their response. If a node fails to respond or responds abnormally within a specified time, it is marked as a potentially faulty node, and a corresponding information fault event is generated. For example, the ping command or other network probing tools can be used to probe the information nodes. Simultaneously, the fault events are further analyzed and confirmed by combining the node's log records and status information.

[0036] In another possible implementation, while searching all information nodes to generate a set of information failure events, a health status assessment model for the information nodes can be established to monitor and evaluate their operational status in real time. By analyzing data such as node performance indicators and communication traffic, it can be determined whether the node is in a healthy state. If the node's health status falls below a set threshold, an information failure event is generated. For example, machine learning algorithms, such as Support Vector Machines (SVM) or neural networks, can be used to classify and predict the node's health status. Furthermore, an expert system can be incorporated to verify and correct the assessment results.

[0037] Physical fault event set: A short circuit in a feeder segment causes load loss in the system. Physical fault events can be generated by traversing all cell grids, with each physical fault corresponding to a cell grid.

[0038] Coordinated fault event set: For example, if a communication terminal is attacked by a DoS attack while a short-circuit fault occurs on one of its feeder segments, this coordinated attack event will cause the system fault range to expand, leading to a further increase in load loss. Coordinated fault events can be obtained by combining any information fault event and a physical fault event.

[0039] In this embodiment, step S3 above utilizes the generated set of fault events, employs grid-based group parallel computing technology, and performs batch simulations of the fault events based on a multi-threaded programming framework to quantify load losses under different faults and generate load loss data, including: We employ a grid-based cluster parallel computing technique to perform batch simulations on the generated risk fault set.

[0040] Specifically, firstly, different real-time libraries are assigned to the grid groups obtained from the grid generation module. Then, the device and topology connection data in the grid groups are pulled into the real-time libraries. Since different grid groups are not electrically connected, power flow calculations and fault simulations can be performed simultaneously in different real-time libraries. The simulation process for the three types of fault events is as follows: Physical Fault Simulation: This simulation follows a short-circuit fault in the line, triggering the feeder automation control to locate, isolate, and restore the fault. After recovery, the impact of the equipment failure on the power flow distribution of the distribution network is analyzed, and load losses are calculated. Physical equipment objects. System load loss due to fault express.

[0041] Information Fault Simulation: Simulates the interference of information node anomalies on the main station's fault location, isolation, and recovery logic, and calculates load loss. Information Equipment Objects System load loss due to fault express.

[0042] Collaborative Fault Simulation: Simulates a short-circuit fault on the line, along with an abnormal feeder automation process at the information node, and calculates load loss after fault recovery. (Object) and Load loss during coordinated failure express.

[0043] In this embodiment, step S4 above, based on load loss data and introduced defense resource parameters, constructs a risk assessment model, calculates the feeder risk coefficient and system risk coefficient, and quantifies the risk propagation path, including: Through the defense resource-defense effect function and the integral formula for failure probability Calculate the failure probability under different defense resource configurations; in, For object The defensive effect value; For object The defensive resources available For object The defense resource conversion coefficient (default is 1). Representative object The probability of a successful attack when under attack. The attack value is a random variable. In the interval The upper part follows a uniform distribution. Let be the probability density function of the attack value; the integral formula for the failure probability indicates that the probability of attack success is equal to the attack value. Exceeding the defensive effect The probability of.

[0044] It should be noted that, This is a non-negative real number representing the amount of defensive resources invested (such as firewall strength and security budget in network security). A larger value indicates better defensive effectiveness. In practical applications, it can be defined according to specific scenarios. The value can be set, for example, based on budget, technical capabilities, or historical data.

[0045] Quantifying physical equipment objects using fault hazard The severity of the malfunction is represented as follows: in, Representation Object The hazards of the malfunction Indicates the number of information nodes.

[0046] The formula for calculating the feeder risk factor is: in, Indicates the current feeder The feeder risk coefficient, Indicates the current feeder The number of physical nodes included.

[0047] The formula for calculating the system risk coefficient is as follows: in, Indicates the system risk coefficient. Indicates feeder The weight is 1 by default. This indicates the number of feeders currently included in the system.

[0048] The system risk coefficient and feeder risk coefficient are visualized and output, such as... Figure 2 As shown.

[0049] Example 3 illustrates a schematic scheme of a grid-based cyber-physical system risk analysis method for distribution networks. It should be noted that the technical solution of a grid-based cyber-physical system risk analysis system for distribution networks is based on the same concept as the aforementioned grid-based cyber-physical system risk analysis method. Details not described in detail in the grid-based cyber-physical system risk analysis system in this embodiment can be found in the description of the aforementioned grid-based cyber-physical system risk analysis method.

[0050] This embodiment also provides a grid-based cyber-physical system risk analysis system for distribution networks, including: The power distribution network data grid partitioning module is used to parse relevant data of the power distribution network, establish the spatiotemporal correspondence of multi-source data, and divide the operation data of the power distribution network physical system and information system into multiple unit grids and grid groups based on topology coupling analysis. The potential fault event generation module is used to automatically search for potential faults and generate three types of fault event sets based on the grid partitioning results and considering the threats faced by the cyber-physical system of the distribution network. The fault load loss quantification module is used to perform batch simulation of fault events using the generated fault event set, employing grid-based group parallel computing technology and a multi-threaded programming framework, to quantify the load loss under different faults and generate load loss data. The risk assessment and propagation analysis module is used to build a risk assessment model based on load loss data and introduced defense resource parameters, calculate feeder risk coefficients and system risk coefficients, and quantitatively analyze risk propagation paths.

[0051] This embodiment also provides an electronic device applicable to a risk analysis method for a distribution network cyber-physical system based on grid partitioning, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement a grid-based cyber-physical system risk analysis method for power distribution networks, as proposed in the above embodiments.

[0052] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a risk analysis method for a distribution network cyber-physical system based on grid partitioning as proposed in the above embodiments.

[0053] The storage medium proposed in this embodiment belongs to the same inventive concept as the risk analysis method for distribution network cyber-physical systems based on grid partitioning proposed in the above embodiment. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0054] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A risk analysis method for a distribution network cyber-physical system based on grid partitioning, characterized in that, include: Analyze relevant data of the distribution network, establish spatiotemporal correspondence of multi-source data, and divide the operation data of the physical system and information system of the distribution network into multiple unit grids and grid groups based on topology coupling analysis; Based on the grid partitioning results, considering the threats faced by the cyber-physical system of the distribution network, potential faults are automatically searched, and three types of fault event sets are generated. Using the generated set of fault events, a grid-based parallel computing technique is employed to perform batch simulations of the fault events based on a multi-threaded programming framework, quantifying load loss under different faults and generating load loss data. Based on load loss data and introduced defense resource parameters, a risk assessment model is constructed to calculate the feeder risk coefficient and system risk coefficient, and to quantitatively analyze the risk propagation path.

2. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 1, characterized in that, The process involves analyzing distribution network-related data, establishing spatiotemporal correspondences among multi-source data, and dividing the operational data of the distribution network's physical and information systems into multiple unit grids and grid groups based on topology coupling analysis. All devices except switches are aggregated into the same cell grid, making the cell grid the smallest unit of risk assessment. It contains multiple physical devices and the cell grids are connected only by switches. A grid group consists of a set of cell grids connected by switches, and there is no electrical connection between different grid groups.

3. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 2, characterized in that, The process of analyzing distribution network-related data, establishing spatiotemporal correspondences of multi-source data, and dividing the operational data of the distribution network physical system and information system into multiple unit grids and grid groups based on topology coupling analysis also includes: When performing fault location, isolation and recovery operations, the internal structure of the cell grid is ignored, and only the fault of the cell grid is determined; for the load transfer problem in fault recovery, all feeders with electrical connection are merged into a grid group, so that each grid group can be analyzed in parallel and without interference. The physical and information topology data of the distribution network are analyzed to establish the mapping relationship between distribution network switches and FTUs outside the unit grid, and the real-time load data of each grid is labeled.

4. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 3, characterized in that, Based on the grid partitioning results, and considering the threats faced by the cyber-physical system of the distribution network, the system automatically searches for potential faults and generates three sets of fault events, including: The scenario is initialized based on the operating status of the distribution network, setting information attack types and physical attack types; each unit grid is traversed to automatically generate three types of fault event sets: In response to possible abnormal situations in communication terminals, all information nodes are searched to generate a set of information failure events. The events in the set of information failure events are manifested as the communication terminal being unable to respond to controlled commands due to a specific attack.

5. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 4, characterized in that, The automatic search for potential faults and generation of three types of fault event sets based on grid partitioning results, considering the threats faced by the cyber-physical system of the distribution network, also includes: Traverse all cell grids to generate a set of physical fault events. Each physical fault in the set of physical fault events corresponds one-to-one with a cell grid. The fault manifests as a short circuit in the feeder segment, which leads to load loss in the system. The events in the generated information fault event set and physical fault event set are combined to obtain the collaborative fault event set.

6. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 5, characterized in that, The process of utilizing the generated set of fault events, employing grid-based group parallel computing technology, and performing batch simulations of the fault events based on a multi-threaded programming framework to quantify load loss under different faults and generate load loss data includes: Three types of fault events are simulated using the generated set of fault events: Physical Fault Simulation: Simulates the process of system triggering feeder automation control to locate, isolate, and restore the fault after a line short-circuit fault, analyzes the impact of equipment faults on power flow distribution in the distribution network, and calculates the load loss after fault recovery; Information Fault Simulation: Simulate the interference of information node anomalies on the main station's fault location, isolation, and recovery logic, and calculate the corresponding load loss; Collaborative fault simulation: Simulates the feeder automation process when a line short-circuit fault and an information node anomaly occur simultaneously, and calculates the load loss after fault recovery.

7. The method for risk analysis of a distribution network cyber-physical system based on grid partitioning as described in claim 6, characterized in that, The risk assessment model, constructed based on load loss data and introduced defense resource parameters, calculates feeder risk coefficients and system risk coefficients, and quantifies the risk propagation path, including: Based on the relationship between defense resources and defense effectiveness and the failure probability integral rule, the failure probability of a successful attack is calculated when each object is attacked under different defense resource configurations, where the amount of defense resources invested is positively correlated with the defense effectiveness. By using fault severity quantification rules and combining load loss data, the severity of physical equipment failures is calculated; based on the fault probability and severity of each physical node, the feeder risk coefficient of the current feeder is calculated; considering the weight of each feeder and combining the feeder risk coefficient, the system risk coefficient is calculated.

8. A risk analysis system for a distribution network cyber-physical system based on grid partitioning, using the method described in any one of claims 1 to 7, characterized in that, include: The power distribution network data grid partitioning module is used to parse relevant data of the power distribution network, establish the spatiotemporal correspondence of multi-source data, and divide the operation data of the power distribution network physical system and information system into multiple unit grids and grid groups based on topology coupling analysis. The potential fault event generation module is used to automatically search for potential faults and generate three types of fault event sets based on the grid partitioning results and considering the threats faced by the cyber-physical system of the distribution network. The fault load loss quantification module is used to perform batch simulation of fault events using the generated fault event set, employing grid-based group parallel computing technology and a multi-threaded programming framework, to quantify the load loss under different faults and generate load loss data. The risk assessment and propagation analysis module is used to build a risk assessment model based on load loss data and introduced defense resource parameters, calculate feeder risk coefficients and system risk coefficients, and quantitatively analyze risk propagation paths.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images