Intelligent power grid fault deduction method and device based on knowledge graph, and electronic equipment

Abstract

The invention discloses an intelligent power grid fault deduction method and device based on a knowledge graph and electronic equipment, and relates to the technical field of power grid fault diagnosis, and the method comprises the steps: receiving protection action alarm information and circuit breaker tripping alarm information when a target intelligent power grid breaks down; based on the protection action alarm information and the circuit breaker tripping alarm information, performing hypothesis reasoning by using a preset power grid topology knowledge graph to obtain a suspected fault equipment set; performing deductive reasoning on each suspected fault device in the suspected fault device set by using a preset protection configuration knowledge graph based on the protection action alarm information and the circuit breaker tripping alarm information to obtain observation event confidence corresponding to each suspected fault device; and evidence fusion reasoning is carried out on each suspected fault device based on each expected observation event confidence, and actual fault information of the smart power grid is obtained. The method can improve the accuracy of the fault diagnosis result of the smart power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid fault diagnosis, and in particular to a method, device, storage medium and electronic device for smart grid fault deduction based on a knowledge graph. Background Art

[0002] Modern society is increasingly reliant on a reliable power supply. When a power grid fault occurs, dispatchers must quickly determine its location and cause to address it appropriately. However, complex power grid faults generate a large amount of event information in a short period of time, making it difficult for dispatchers to quickly and accurately analyze the fault. Consequently, complex power grid fault diagnosis technology has attracted widespread attention.

[0003] The development of power grid fault diagnosis technology has generally followed the path of artificial intelligence (AI) technology. Early fault diagnosis primarily employed rule-based expert systems. However, expert systems have always faced significant bottlenecks in knowledge representation, acquisition, and updating to handle the uncertainty caused by protection misoperation and refusal to operate, hindering the widespread application of this technology. As AI enters the data-driven phase, related technologies are also being applied to fault diagnosis. However, current data-driven AI suffers from poor interpretability and requires a large number of training samples. The probability of complex power grid faults is low, and when samples are insufficient, the generalization ability of data-driven methods is subject to significant uncertainty. Existing methods for diagnosing smart grid faults do not accurately restore incidents. Summary of the Invention

[0004] In view of this, the present invention provides a method, device, storage medium and electronic device for smart grid fault deduction based on knowledge graph, the main purpose of which is to solve the current problem of inaccurate power grid fault diagnosis.

[0005] To solve the above problems, this application provides a smart grid fault deduction method based on knowledge graph, including:

[0006] Receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0007] Based on the protection action alarm information and the circuit breaker trip alarm information, a preset power grid topology knowledge graph is used to perform hypothesis reasoning to obtain a set of suspected fault devices;

[0008] Based on the protection action alarm information and the circuit breaker trip alarm information, a preset protection configuration knowledge graph is used to perform deductive reasoning on each suspected faulty device in the suspected faulty device set to obtain an observed event confidence level corresponding to each suspected faulty device;

[0009] Based on the confidence level of each expected observation event, evidence fusion reasoning is performed on each suspected faulty device to obtain actual fault information of the smart grid.

[0010] Optionally, before performing hypothesis reasoning using a preset power grid topology knowledge graph based on the protection action alarm information and the circuit breaker trip alarm information, the method further includes: constructing a preset power grid topology knowledge graph;

[0011] The construction of the preset power grid topology knowledge graph specifically includes:

[0012] Obtaining a topological network structure diagram of the target smart grid;

[0013] Building a model based on the topological network structure diagram to obtain a CIM / E model corresponding to the target smart grid;

[0014] A knowledge graph is constructed based on the CIM / E model to obtain the preset grid topology knowledge graph corresponding to the target smart grid, which represents the topological connection relationship between devices.

[0015] Optionally, the performing of hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected faulty devices specifically includes:

[0016] Extracting data based on the protection action alarm information and the circuit breaker trip alarm information to obtain a tripped circuit breaker set;

[0017] Constructing a mapping relationship between each tripped circuit breaker in the tripped circuit breaker set to obtain each tripped circuit breaker group;

[0018] For each of the tripped circuit breaker groups, a first query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain a first suspected fault device of the shortest path corresponding to each of the tripped circuit breaker groups;

[0019] For the line primary devices in the first suspected faulty devices, a second query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain the second suspected faulty device of the shortest path corresponding to each of the line primary devices;

[0020] Performing a union calculation on the first suspected faulty device and the second suspected faulty device to obtain the suspected faulty device set;

[0021] The primary equipment includes lines, busbars and transformers.

[0022] Optionally, before performing deductive reasoning on each suspected faulty device in the suspected faulty device set using a preset protection configuration knowledge graph based on the protection action alarm information and the circuit breaker trip alarm information, the method further includes: constructing a preset protection configuration knowledge graph;

[0023] The construction of the preset protection configuration knowledge graph specifically includes:

[0024] Parsing the SCD file of the target smart grid to obtain protection configuration information of the target smart grid, the protection configuration information including primary protection and backup protection configuration information of a primary device and an association relationship between protection and a tripped circuit breaker;

[0025] Using a second query statement to perform a shortest path query on the preset power grid topology knowledge graph to obtain remote backup protection information;

[0026] Obtain the action time constraint of the backup protection action from the preset protection setting list;

[0027] A knowledge graph is constructed based on the protection configuration information, the remote backup protection information and the action time constraint to obtain the preset protection configuration knowledge graph for characterizing the protection configuration and action time constraint of the primary device.

[0028] Optionally, the performing of deductive reasoning on each suspected faulty device in the set of suspected faulty devices using a preset protection configuration knowledge graph based on the protection action alarm information and the circuit breaker trip alarm information to obtain an observed event confidence level corresponding to each suspected faulty device specifically includes:

[0029] Based on each of the suspected faulty devices, a third query statement is used to query the preset protection configuration knowledge graph to obtain an expected event corresponding to each of the suspected faulty devices and a time constraint for the occurrence of the expected event;

[0030] Performing data screening based on the protection action alarm information and the circuit breaker trip alarm information to obtain an observation event;

[0031] Confidence calculation is performed based on the expected event and the observed event to obtain the confidence of the observed event corresponding to each of the suspected faulty devices.

[0032] Optionally, the performing confidence calculation based on the expected event set and the observed event set to obtain the confidence of the observed event corresponding to each of the suspected faulty devices specifically includes:

[0033] Determine a first indicator value based on the actual time of occurrence of the observed event and the time constraint of the observed event in the preset protection configuration knowledge graph;

[0034] determining a second indicator value based on the protection action type of the observed event;

[0035] A preset confidence function is used to perform calculation based on the first indicator value and the second indicator value to obtain the confidence of the observed event, so as to obtain the confidence of the observed event corresponding to each of the suspected faulty devices.

[0036] Optionally, performing evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observed events to obtain actual fault information of the smart grid fault specifically includes:

[0037] Supplement the confidence of each expected observation event with evidence to obtain a target confidence matrix;

[0038] Based on the target confidence matrix, the DS evidence fusion method is used to perform evidence fusion to obtain an evidence fusion result;

[0039] Based on the evidence fusion result, the protection action alarm information and the circuit breaker tripping alarm information when the target smart grid fails, the fault is restored using the preset protection configuration knowledge graph to obtain the actual fault information of the target smart grid.

[0040] To solve the above problems, this application proposes a smart grid fault deduction device based on knowledge graph, including:

[0041] A receiving module, used to receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0042] A hypothesis reasoning module, configured to perform hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices;

[0043] a deductive reasoning module, configured to perform deductive reasoning on each suspected faulty device in the set of suspected faulty devices based on the protection action alarm information and the circuit breaker trip alarm information using a preset protection configuration knowledge graph, and obtain an observed event confidence level corresponding to each suspected faulty device;

[0044] The evidence fusion reasoning module is used to perform evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observation events to obtain actual fault information of the smart grid.

[0045] In order to solve the above problems, the present application provides a storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned knowledge graph-based smart grid fault deduction method are implemented.

[0046] In order to solve the above problems, the present application provides an electronic device, which includes at least a memory and a processor. The memory stores a computer program, and the processor implements the steps of the above-mentioned knowledge graph-based smart grid fault deduction method when executing the computer program on the memory.

[0047] Beneficial effects of this application: This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0048] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0050] Figure 1 A schematic diagram of a process for a smart grid fault deduction method based on a knowledge graph provided in an embodiment of the present application is shown;

[0051] Figure 2 A schematic diagram of a flow chart of a smart grid fault deduction method based on a knowledge graph provided in another embodiment of the present application is shown;

[0052] Figure 3 A single-line diagram of a simple power grid topology network structure according to an embodiment of the present application is shown;

[0053] Figure 4 The CIM / E model corresponding to the topology network structure single line diagram according to an embodiment of the present application is shown;

[0054] Figure 5 The power grid topology knowledge graph corresponding to the topology network structure single-line diagram of the embodiment of the present application is shown;

[0055] Figure 6 The protection configuration knowledge graph of line L2 in an embodiment of the present application is shown;

[0056] Figure 7 A structural block diagram of a smart grid fault deduction device based on a knowledge graph is shown in another embodiment of the present application;

[0057] Figure 8 A schematic diagram of a power grid structure showing a single fault mode of a calculation example power grid according to an embodiment of the present application is shown;

[0058] Figure 9 A schematic diagram of a power grid structure showing multiple failure modes of an example power grid according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0059] Various aspects and features of the present application are described herein with reference to the accompanying drawings.

[0060] It should be understood that various modifications may be made to the embodiments of the present application. Therefore, the above description should not be considered as limiting, but merely as an example of an embodiment. Other modifications within the scope and spirit of the present application will occur to those skilled in the art.

[0061] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0062] These and other characteristics of the present application will become apparent from the following description of a preferred form of embodiment given as a non-limiting example with reference to the accompanying drawings.

[0063] It should also be understood that although the present application has been described with reference to certain specific examples, those skilled in the art will readily be able to implement many other equivalent forms of the present application.

[0064] The above and other aspects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings.

[0065] Specific embodiments of the present application will be described hereinafter with reference to the accompanying drawings; however, it should be understood that the embodiments described are merely examples of the present application and may be implemented in a variety of ways. Familiar and / or repetitive functions and structures are not described in detail to avoid obscuring the present application with unnecessary or redundant details. Therefore, the specific structural and functional details described herein are not intended to be limiting, but rather serve merely as a basis and representative basis for the claims to teach those skilled in the art to variously utilize the present application with substantially any suitable detailed structure.

[0066] This specification may use the phrases "in one embodiment," "in another embodiment," "in yet another embodiment," or "in other embodiments," which may all refer to one or more of the same or different embodiments according to the present application.

[0067] The present application embodiment provides a method for smart grid fault deduction based on knowledge graph, such as Figure 1 Shown, including:

[0068] Step S101: receiving protection action alarm information and circuit breaker trip alarm information when a fault occurs in a target smart grid;

[0069] During the implementation of this step, the fault zero time is determined through the fault recorder and protection activation information, and protection action alarm information and circuit breaker trip alarm information are received. Specifically, when a fault occurs in the target smart grid, the circuit breaker must ultimately trip to isolate the fault. The range of the tripped circuit breaker defines the fault area. The fault recorder is a critical device in the power system that records changes in electrical quantities during a fault. When a grid fault occurs, the fault recorder automatically activates and records the waveforms and effective values of electrical quantities such as current and voltage before and after the fault. This recorded data can be used to analyze the nature and type of the fault, as well as the specific time of occurrence, known as the fault zero time. The fault recorder determines the fault zero time by detecting sudden changes in current or voltage. When a fault is detected, the protection device activates and issues a protection action signal. The protection activation information includes the protection device's action time and action type. When a fault occurs, the circuit breaker trips to isolate the fault area. The circuit breaker alarm information includes the trip time and tripping reason.

[0070] Step S102: performing hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices;

[0071] During the specific implementation of this step, data extraction is performed based on the protection action alarm information and the circuit breaker tripping alarm information to obtain a tripped circuit breaker set; a mapping relationship between each tripped circuit breaker in the tripped circuit breaker set is constructed to obtain each tripped circuit breaker group; for each of the tripped circuit breaker groups, a first query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain each primary device of the shortest path corresponding to each of the tripped circuit breaker groups, the primary devices including lines, busbars and transformers, to obtain the set of suspected fault devices.

[0072] Step S103: performing deductive reasoning on each suspected faulty device in the suspected faulty device set based on the protection action alarm information and the circuit breaker trip alarm information using a preset protection configuration knowledge graph to obtain an observed event confidence level corresponding to each suspected faulty device;

[0073] During the specific implementation of this step, the preset protection configuration knowledge graph is searched based on each suspected faulty device to obtain the expected events corresponding to each suspected faulty device and the time constraints for the occurrence of the expected events; data is screened based on the protection action alarm information and the circuit breaker tripping alarm information to obtain observed events; confidence calculation is performed based on the expected events and the observed events to obtain the confidence of the observed events corresponding to each suspected faulty device.

[0074] Step S104: performing evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observed events to obtain actual fault information of the smart grid.

[0075] During the specific implementation of this step, evidence is supplemented for the confidence of each expected observation event to obtain a target confidence matrix; based on the target confidence matrix, evidence fusion is performed using the DS evidence fusion method to obtain an evidence fusion result; based on the evidence fusion result, the protection action alarm information when a fault occurs in the target smart grid, and the circuit breaker tripping alarm information, the preset protection configuration knowledge graph is used to restore the fault to obtain the actual fault information of the target smart grid.

[0076] This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0077] Another embodiment of the present application provides another smart grid fault deduction method based on knowledge graph, such as Figure 2 Shown, including:

[0078] Step S201: receiving protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0079] During the specific implementation of this step, the fault zero time (i.e., the time when the target smart grid fault occurred) is determined through the fault recording device and protection start information, and protection action alarm information and circuit breaker trip alarm information are received. The protection action alarm information and circuit breaker trip alarm information can be collected in real time by the fault recording device.

[0080] Step S202: constructing a preset power grid topology knowledge graph;

[0081] During the specific implementation of this step, a topological network structure diagram of the target smart grid is obtained; Figure 3 The figure shows a simple single-line diagram of the topological network structure of a power grid. Based on the topological network structure diagram, a model is constructed to obtain a CIM / E model corresponding to the target smart grid; for example, when the target smart grid is Figure 3 When the topology network is as shown, Figure 4 As shown, it is a CIM / E model corresponding to the target smart grid. Based on the CIM / E model, a knowledge graph is constructed to obtain the preset grid topology knowledge graph corresponding to the target smart grid, which represents the topological connection relationship between devices. Figure 5 As shown, it is the grid topology knowledge graph corresponding to the target smart grid; the dispatching system has widely adopted the CIM / E model to describe the primary system topology. Since there is a clear mapping relationship between the CIM / E model and the grid topology knowledge graph model, the grid topology knowledge graph can be automatically constructed according to the CIM / E model. The accompanying drawings of this application only explain this application with a simple topological structure, and this application can also be applied to more complex target smart grids. The preset grid topology knowledge graph of this application is a semantic network that reveals entities and their relationships; the grid topology knowledge graph constructed by this application is used to describe the topological relationship of the entire network. This application formalizes the preset grid topology knowledge graph into the following undirected graph G ， As shown in the following formula (1):

[0082]

[0083] Where V is the vertex set and E is the edge set. The vertex set includes the primary equipment set EQ and the electrical connection node set CN. In the power grid topology knowledge graph, each vertex represents a primary equipment or electrical connection node, and each edge is an unordered pair (x, y) representing the topological connection relationship between devices (interconnected through electrical connection nodes).

[0084] Step S203: performing hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices;

[0085] During the specific implementation of this step, data is extracted based on the protection action alarm information and the circuit breaker trip alarm information to obtain a set of tripped circuit breakers. When a power grid fault occurs, the circuit breaker must trip to isolate the fault. The range of tripped circuit breakers defines the fault area, and the primary equipment within this range is considered to be a suspected faulty device. Assume that the reasoning task can be formally described as shown in the following formula (2):

[0086]

[0087] in, is a set of suspected faulty devices, and I is the number of suspected faulty devices. Construct a mapping relationship between each tripped circuit breaker in the tripped circuit breaker set to obtain each tripped circuit breaker group; hypothesis reasoning is to infer the fault range through the action results and form a fault hypothesis set. Its goal is to narrow the search space for subsequent fault reasoning without missing the real faulty device. Summarize the tripped circuit breaker information into a set Where O is the total number of tripped circuit breakers. This set combines the circuit breaker trip alarm and protection action alarm information. The present invention assumes that the above two types of information will not be lost at the same time. Construct a full pair combination of tripped circuit breakers For each of the tripped circuit breaker groups, a first query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain the first suspected fault device of the shortest path corresponding to each of the tripped circuit breaker groups; for the line primary devices in the first suspected fault devices, a second query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain the second suspected fault device of the shortest path corresponding to each of the line primary devices; a union calculation is performed on the first suspected fault device and the second suspected fault device to obtain the suspected fault device set. The primary equipment includes lines, busbars and transformers. The first query statement can be: MATCH path = shortestpath((:CB{name:“CB3”})-[:CONNECTED_TO*]-(:CB{name:“CB6”}))RETURN path; the second query statement can be: MATCH path = shortestpath((:Line{name:'L2'})-[:CONNECTED_TO*]-(l:Line))RETURN l; Figure 5 For example, assuming that the alarm information of protection and circuit breaker observed after the fault is shown in Table 1, the following set of suspected fault devices can be obtained through hypothetical reasoning: It can be expressed as the following formula (3):

[0088]

[0089] Table 1 Alarm information of protection and circuit breaker

[0090] Failure Zero Moment 38ms Protection action alarm information <h2 style=";text-align:left;direction:ltr"><![CDATA[L2B <h2 style=";text-align:left;direction:ltr"> 2m <h2 style=";text-align:left;direction:ltr"> (50ms),L2B<h2 style=";text-align:left;direction:ltr"> 3m <h2 style=";text-align:left;direction:ltr"> (51ms),L3B<h2 style=";text-align:left;direction:ltr"> 4b <h2 style=";text-align:left;direction:ltr"> (700ms)]]><h2 style=";text-align:left;direction:ltr"> Circuit breaker trip alarm information <![CDATA[CB3(65ms),CB9(300ms),CB6(725ms)]]>

[0091] Step S204: Constructing a preset protection configuration knowledge graph;

[0092] During the specific implementation of this step, the SCD file of the target smart grid is parsed to obtain the protection configuration information of the target smart grid, which includes the primary protection of the primary equipment, the near backup protection configuration information, and the association between the protection and the tripping circuit breaker; the protection configuration information can be automatically extracted by parsing the SCD file; the second query statement is used to perform the shortest path query on the preset power grid topology knowledge graph to obtain the far backup protection information; the action time constraint of the backup protection action is obtained from the preset protection setting list; a knowledge graph is constructed based on the protection configuration information, the far backup protection information, and the action time constraint to obtain the preset protection configuration knowledge graph for characterizing the protection configuration and action time constraint of the primary equipment. Figure 6 The protection configuration knowledge graph of line L2 is shown. The protection configuration knowledge graph constructed by the present invention is used to describe the protection configuration and action time constraints of the primary equipment. The present invention formalizes it as the following directed graph G, which can be expressed as the following formula (4):

[0093]

[0094] Among them, E EQ 、E CB and E PR They represent a set of equipment failure events, a set of circuit breaker tripping and closing events, and a set of protection action events respectively; each vertex v represents an event in the above set; each edge is a set of ordered event pairs<p,q> , indicating that events p and q have a causal relationship; T maps the event set E to the real number set That is, each event pair<p,q> Assign a real number as the edge weight to represent the time constraint between p and q. Suppose event p will lead to event q, and the time of q relative to p must fall within the interval [t min ,t max ], t min ,t max ≥ 0. To represent the above causal relationship and time constraints in the protection configuration knowledge graph, the present invention sets the weights according to the following rules: the weight of the directed edge from p to q is set to t max ; The weight of the directed edge from q to p is set to -t min Therefore, the protection configuration knowledge graph is a directed graph containing negative weight edges. Figure 3 As an example of a simple power grid, Figure 6 The protection configuration knowledge graph related to line L2 is shown, with L2 as the root vertex. Conventions in the figure: L2B 3m Indicates the protection of line L2 near busbar B3; subscripts m, f, and b represent main protection, failure protection, and backup protection respectively. For example, Figure 6 Middle vertices L2 and L2B 3m The two directed edges between them indicate that if L2 fails, the action time constraint of the line main protection near B3 is [10,40]ms.

[0095] Step S205: Based on each of the suspected faulty devices, a third query statement is used to query the preset protection configuration knowledge graph to obtain an expected event corresponding to each of the suspected faulty devices and a time constraint for the occurrence of the expected event;

[0096] In the specific implementation process of this step, after obtaining the suspected fault equipment set After that, for each suspected faulty device Perform deductive reasoning. Its formal mathematical description can be expressed as follows:

[0097]

[0098] Where p→q is the protection rule in formula (4), which is represented by the protection configuration knowledge graph; If For a faulty device, the set of protection and circuit breaker actions expected to occur according to the rule p→q is referred to as the expected event set in this invention. In this way, the accident deduction problem is transformed into a typical deductive reasoning problem, which can be used to utilize the knowledge graph. Related preset protection configuration knowledge graphs (e.g. Figure 5 ), combined with the full time constraint table constructed using the third query statement, we can get To and its time constraints. So far, the deductive reasoning of formula (5) is completed. The third query statement is used to query the all-pairs time constraints between events in the protection configuration knowledge graph. This query can be completed by the All-Pairs Shortest Path algorithm. The present invention adopts the Floyd-Warshall algorithm that supports negative weight edges. Figure 6Taking the protection configuration knowledge graph as an example, the results of the third query, Query3, form the full-pair time constraint table in Table 2. For example, the table shows that the shortest distance from line L2 to circuit breaker CB6 is 1140ms, while the shortest distance from CB6 to L2 is -620ms. This means that if line L2 fails, the time constraint for the remote tripping event of circuit breaker CB6 is [620, 1140]ms. Using Table 2, we can quickly determine whether the time constraint between any two events is met. This provides an important basis for implementing fault reasoning.

[0099] Table 2 Full pair time constraint table (ms)

[0100] vertex <![CDATA[L2]]> <![CDATA[L2B 2m ]]> <![CDATA[L2B 3m ]]> <![CDATA[L3B 4b ]]> <![CDATA[CB3]]> <![CDATA[CB6]]> <![CDATA[L2]]> 0 40 40 1100 80 1140 <![CDATA[L2B 2m ]]> -10 0 30 1090 40 1130 <![CDATA[L2B 3m ]]> -10 30 0 1090 70 1130 <![CDATA[L3B 4b ]]> -600 -560 -560 0 -520 40 <![CDATA[CB3]]> -30 -20 10 1070 0 1110 <![CDATA[CB6]]> -620 -580 -580 -20 -540 0

[0101] Step S206: performing data screening based on the protection action alarm information and the circuit breaker trip alarm information to obtain an observation event;

[0102] During the specific implementation of this step, It is not necessarily a real faulty device, but may be affected by complex factors such as protection misoperation and information loss. Possibly related to the actual observed event set Data is filtered based on the protection action alarm information and the circuit breaker trip alarm information to obtain an observed event.

[0103] Step S207: performing confidence calculation based on the expected event and the observed event to obtain the confidence of the observed event corresponding to each of the suspected faulty devices;

[0104] During the specific implementation of this step, It is not necessarily a real faulty device, but may be affected by complex factors such as protection misoperation and information loss. Possibly related to the actual observed event set The present invention calculates the confidence by evaluating the consistency of the two. The calculation process can be formalized as the following formula (6):

[0105]

[0106] Among them, m ij Indicates that the fault hypothesis Next, observe the event The confidence level of the observed event is determined based on the actual time of the observed event and the time constraint of the observed event in the preset protection configuration knowledge graph; the observed event is judged according to the full time constraint table. Whether the time constraint is met, When the time constraint is met, the first index value r is determined to be 1; If it appears in the full pair time constraint table but exceeds the upper limit of the time constraint (possibly due to slow protection or circuit breaker operation), the first index value r is determined to be 0.5; If it does not appear in the full time constraint table or is lower than the time constraint lower limit, the first index value r is determined to be 0. Based on the protection action type of the observed event, a second index value is determined; the second index value D is used to assign weights according to the protection action type: specifically, if the main protection action or the associated circuit breaker trips, the second index value D is determined to be 1; if the failure protection action or the associated circuit breaker trips, the second index value D is determined to be 2; if the backup protection action or the associated circuit breaker trips, the second index value D is determined to be 3; based on the first index value and the second index value, a preset confidence function is used to perform calculation processing to obtain the observed event confidence, so as to obtain the observed event confidence m corresponding to each of the suspected faulty devices. ij The confidence level of the observed event m ij The mathematical formula for calculating can be expressed as follows:

[0107]

[0108] by Figure 3 Taking Table 1 as an example, the calculated m ij Putting them in an I×J two-dimensional table, we can get the confidence summary table representing the confidence matrix as shown in Table 3.

[0109] Table 3 Confidence summary table

[0110] Suspected faulty equipment <![CDATA[L2B 2m ]]> <![CDATA[L2B 3m ]]> <![CDATA[L3B 4b ]]> <![CDATA[CB3]]> <![CDATA[CB6]]> <![CDATA[CB9]]> <![CDATA[B2]]> 0 0 0 0 0 0 <![CDATA[B3]]> 0 0 0.125 0 0.125 0 <![CDATA[L2]]> 1 1 0.125 1 0.125 0 <![CDATA[L3]]> 0 0 0 0 0 0 <![CDATA[L4]]> 0 0 0 0 0 0

[0111] Step S208: Supplement the confidence of each expected observation event with evidence to obtain a target confidence matrix;

[0112] During the specific implementation of this step, the confidence summary table representing the confidence matrix is supplemented for insufficient evidence, and the confidence that cannot be attributed to any faulty device is supplemented with evidence to obtain a target confidence matrix; the target confidence of the target confidence matrix is summarized as shown in Table 4 below:

[0113] Table 4 Summary of target confidence

[0114]

[0115] Step S209: performing evidence fusion based on the target confidence matrix using the DS evidence fusion method to obtain an evidence fusion result;

[0116] In the specific implementation of this step, in Table 4, if each column is regarded as an independent source of evidence, the next question is: how to fuse J sources of evidence to perform fault identification on I suspected faulty devices. To this end, this application introduces the DS evidence fusion theory for final reasoning.

[0117] For the problem of this application, the identification framework Θ of the evidence fusion problem is a set of suspected faulty devices In this recognition framework, the confidence function m:2 Θ →[0,1] should satisfy the following mathematical formula (8):

[0118]

[0119] in: is a proposition, and m(A) is the confidence level of A.

[0120] Without loss of generality, suppose there are two mutually exclusive evidence sources m1 and m2, then the fusion result of proposition A is as shown in the following formula (9):

[0121]

[0122] in: The 1-K in the denominator is to ensure that the fusion result still meets the normalization condition of formula (8). Table 4 is fused using the above formula, and the results are shown in Table 5. The additional row at the bottom of the table assigns the confidence level that cannot be attributed to any faulty device, which is used to indicate insufficient evidence.

[0123] Table 5 Evidence fusion results

[0124]

[0125] Step S210: Based on the evidence fusion result, the protection action alarm information when the target smart grid fails, and the circuit breaker tripping alarm information, the preset protection configuration knowledge graph is used to restore the fault to obtain the actual fault information of the target smart grid.

[0126] In the specific implementation process of this step, after obtaining the evidence fusion results in Table 5, the original alarm information in Table 1 and Figure 6 The protection configuration knowledge graph provides a complete explanation of the accident process. The specific implementation steps are as follows:

[0127] Step 1: Determine the faulty device. Based on the "Evidence Fusion Result" column in Table 5, determine the suspected device corresponding to the maximum value as the actual faulty device. If the maximum value is lower than the evidence fusion result, the evidence fusion is considered invalid and the faulty device cannot be determined.

[0128] Step 2: Determine the protection and circuit breaker action behavior. After confirming the actual faulty equipment, further analyze the protection and circuit breaker action behavior:

[0129] ① Misoperation event: If an observed event is assigned a confidence level of 0, the event is considered a misoperation;

[0130] ② Refusal to operate event: Based on the protection configuration knowledge graph of the actual faulty equipment, if a protection or circuit breaker should operate but is not observed, it is judged as refusal to operate.

[0131] Step 3: Restore the accident process. Finally, combine the protection and circuit breaker action sequence to fully restore the accident process. The details are as follows:

[0132] Line L2 fails at 38ms, L2B 2m 、L2B 3m They operate at 50ms and 51ms respectively, CB3 trips at 65ms, and CB4 refuses to operate; L3B 4b It operates at 700ms, CB6 trips at 725ms, and CB9 malfunctions at 300ms.

[0133] This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0134] The following is an example of the deduction process of this application in combination with specific applications:

[0135] This section establishes an accident simulation example based on the actual topology and protection configuration of four 500kV substations in the XX power grid, namely, “A-B-C-D”. Figure 8 、 9 The figure shows the substation identification, and the protection is configured in a dual configuration. For example, L1S1 represents the substation S1 side of line L1, and B1m,1 represents the first set of main protection for busbar B1.

[0136] exist Figure 6 In the example, assume that line L2 fails. The system determines the zero time of the fault through the fault recorder and protection start-up information, and receives the protection and circuit breaker alarm information shown in Table 6.

[0137] Table 6 Alarm information of protection and circuit breaker in case of single fault

[0138] Failure Zero Moment 50ms Protection action alarm information <![CDATA[L2S 1m,1 (62ms),L2S 2m,1 (63ms),L2S 1m，2 (64ms),L2S 2m，2 (65ms),5033 f (356ms)]]> Circuit breaker trip alarm information <![CDATA[CB 5032 (85ms),CB 5061 (86ms),CB 5062 (87ms),CB 5081 (377ms),CB 5013 (380ms),CB 5023 (382ms)]]>

[0139] Step 1: Hypothesis reasoning: Based on the grid topology knowledge graph of the example grid and Table 6, the set of suspected faulty devices is obtained by hypothesis reasoning.

[0140] Step 2: Deductive reasoning: Based on Table 6 and the protection configuration knowledge graph of suspected faulty equipment, the confidence level of each event is obtained by deductive reasoning, as shown in Table 7.

[0141] Table 7 Confidence summary table

[0142] Suspected faulty equipment <![CDATA[L2S 1m,1 ]]> <![CDATA[L2S 2m,1 ]]> <![CDATA[L2S 1m，2 ]]> <![CDATA[L2S 2m，2 ]]> <![CDATA[5033 f ]]> <![CDATA[CB 5032 ]]> <![CDATA[CB 5061 ]]> <![CDATA[CB 5062 ]]> <![CDATA[CB 5081 ]]> <![CDATA[CB 5013 ]]> <![CDATA[CB 5023 ]]> <![CDATA[B2]]> 0 0 0 0 0.5 0 0 0 0 0.5 0.5 <![CDATA[B5]]> 0 0 0 0 0 0 0 0 0.5 0 0 <![CDATA[L2]]> 1 1 1 1 0.5 1 1 1 0 0.5 0.5 <![CDATA[L3]]> 0 0 0 0 0 0 0 0 0.5 0 0.5 <![CDATA[L4]]> 0 0 0 0 0 1 0 0 0.5 0 0

[0143] Step 3: Fusion reasoning: Perform evidence fusion on Table 7. The results are shown in Table 8.

[0144] Table 8 Evidence fusion results

[0145]

[0146] Step 4. Accident process explanation: According to the evidence fusion results in Table 8, it can be seen that line L2 fails at 50ms and the main protection L2S 1m,1 and L2S 2m,1 Actions at 62ms and 63ms, L2S 1m,2 and L2S 2m,2 Action at 64ms and 65ms; circuit breaker CB 5032 , CB 5061 , CB 5062 Trip at 85ms, 86ms and 87ms, CB 5033 Refusal to operate; Failure protection 5033 f The circuit breaker CB is operated at 356ms. 5013 and CB 5023 Trip at 380ms and 382ms. Circuit breaker CB 5081 It malfunctioned at 377ms.

[0147] The method of this application can also be used to deduce accidents with multiple faults. The specific examples are as follows:

[0148] exist Figure 9 In this example, lines L2 and L4 successively faulted. After the faults resolved, the system received the protection and circuit breaker alarm information shown in Table 9. The fault recording and protection activation information detected two fault zero moments.

[0149] Table 9 Alarm information of protection and circuit breaker in case of multiple faults

[0150]

[0151] Step 1: Hypothesis reasoning: Based on the grid topology knowledge graph of the example grid and Table 9, the set of suspected faulty devices is obtained through hypothesis reasoning.

[0152] Step 2: Deductive reasoning: Based on the set of suspected faulty devices, two fault zero moments are selected for deductive reasoning:

[0153] 1) Taking 40ms as the zero fault time, combined with the protection configuration knowledge graph and Table 9, the reasoning results in Table 10 are obtained.

[0154] 2) Taking 776ms as the fault zero time, combined with the protection configuration knowledge graph and Table 9, the reasoning results in Table 11 are obtained.

[0155] Table 10 Summary of the credibility of first failure

[0156] Suspected faulty equipment <![CDATA[B 2m,1 ]]> <![CDATA[B 2m，2 ]]> <![CDATA[L4S 1m，1 ]]> <![CDATA[L4S 3m，1 ]]> <![CDATA[L4S 1m，2 ]]> <![CDATA[L4S 3m，2 ]]> <![CDATA[5033 f ]]> <![CDATA[CB 5013 ]]> <![CDATA[CB 5023 ]]> <![CDATA[CB 5032 ]]> <![CDATA[CB 5081 ]]> <![CDATA[CB 5061 ]]> <![CDATA[CB 5062 ]]> <![CDATA[CB 5031 ]]> <![CDATA[CB 5071 ]]> <![CDATA[CB 5072 ]]> <![CDATA[B2]]> 1 1 0 0 0 0 0.5 1 1 0.5 0 0.5 0.5 0 0.125 0.125 <![CDATA[B5]]> 0 0 0 0 0 0 0 0 0 0.5 0.5 0 0 0.25 0.5 0.25 <![CDATA[L2]]> 0 0 0 0 0 0 0.5 0 0 0 0 0.5 0.5 0.125 0.25 0.25 <![CDATA[L3]]> 0 0 0 0 0 0 0 0 0 0 0.5 0 0 0.125 0.125 0.125 <![CDATA[L4]]> 0 0 0.5 0.5 0.5 0.5 0 0 0 0.5 0 0.25 0.25 0.5 0.5 0.5

[0157] Table 11 Summary of the credibility of the second failure

[0158] Suspected faulty equipment <![CDATA[B 2m,1 ]]> <![CDATA[B 2m,2 ]]> <![CDATA[L4S 1m,1 ]]> <![CDATA[L4S 3m,1 ]]> <![CDATA[L4S 1m，2 ]]> <![CDATA[L4S 3m，2 ]]> <![CDATA[5033 f ]]> <![CDATA[CB 5013 ]]> <![CDATA[CB 5023 ]]> <![CDATA[CB 5032 ]]> <![CDATA[CB 5081 ]]> <![CDATA[CB 5061 ]]> <![CDATA[CB 5062 ]]> <![CDATA[CB 5031 ]]> <![CDATA[CB 5071 ]]> <![CDATA[CB 5072 ]]> <![CDATA[B2]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <![CDATA[B5]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.5 0 <![CDATA[L2]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <![CDATA[L3]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <![CDATA[L4]]> 0 0 1 1 1 1 0 0 0 0 0 0 0 1 1 1

[0159] Step 3: Fusion reasoning: Based on Table 10 and Table 11, Table 12 and Table 13 are obtained by evidence fusion respectively.

[0160] Table 12 Evidence fusion results of first failure

[0161]

[0162] Table 13 Evidence fusion results of the second fault

[0163]

[0164] Step 4. Accident process explanation: According to the evidence fusion results in Table 12, busbar B2 is determined to be the first actual fault device; according to Table 13, line L4 is determined to be the second actual fault device. Circuit breaker CB 5081 The confidence levels of busbar B2 and line L4 are both 0, which is considered a false trip event. According to the protection configuration knowledge graph of busbar B2, the circuit breaker CB 5033 No tripping, determined as a refusal to operate event. The complete fault process is as follows:

[0165] First fault: Busbar B2 fails at 40ms, main protection B 2m,1 and B 2m,2 Action at 52ms and 54ms; circuit breaker CB 5013 , CB 5023 Trip at 75ms, 76ms, CB 5033Refusal to operate; Failure protection 5033 f The circuit breaker CB will be activated at 350ms. 5032 , CB 5061 and CB 5062 Trips at 377ms, 577ms and 579ms.

[0166] Second fault: Line L4 fails at 776ms, main protection L4S 1m,1 and L4S 3m,1 Action at 790ms and 791ms, L4S 1m,2 and L4S 3m,2 Action at 794ms and 795ms; circuit breaker CB 5031 , CB 5071 and CB 5072 Trips at 815ms, 817ms and 818ms.

[0167] Circuit Breaker CB 5081 It malfunctions at 300ms.

[0168] Another embodiment of the present application provides a smart grid fault deduction device based on knowledge graph, such as Figure 7 Shown include:

[0169] Receiving module 1, used to receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0170] Hypothesis reasoning module 2, configured to perform hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices;

[0171] Deductive reasoning module 3, configured to perform deductive reasoning on each suspected faulty device in the suspected faulty device set based on the protection action alarm information and the circuit breaker trip alarm information using a preset protection configuration knowledge graph, and obtain an observed event confidence level corresponding to each suspected faulty device;

[0172] The evidence fusion reasoning module 4 is used to perform evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observed events to obtain actual fault information of the smart grid.

[0173] During the specific implementation process, the device also includes a preset power grid topology knowledge graph construction module, which is specifically used to: obtain a topological network structure diagram of the target smart grid; construct a model based on the topological network structure diagram to obtain a CIM / E model corresponding to the target smart grid; construct a knowledge graph based on the CIM / E model to obtain the preset power grid topology knowledge graph corresponding to the target smart grid, which represents the topological connection relationship between devices.

[0174] During the specific implementation process, the hypothesis reasoning module 2 is specifically used to: extract data based on the protection action alarm information and the circuit breaker tripping alarm information to obtain a tripped circuit breaker set; construct a mapping relationship between each tripped circuit breaker in the tripped circuit breaker set to obtain each tripped circuit breaker group; for each of the tripped circuit breaker groups, use a first query statement to perform a shortest path query on the preset power grid topology knowledge graph to obtain the first suspected fault device of the shortest path corresponding to each of the tripped circuit breaker groups; for the line primary equipment in the first suspected fault equipment, use a second query statement to perform a shortest path query on the preset power grid topology knowledge graph to obtain the second suspected fault device of the shortest path corresponding to each of the line primary equipment; perform a union calculation on the first suspected fault equipment and the second suspected fault equipment to obtain the suspected fault device set; wherein, the primary equipment includes lines, busbars and transformers.

[0175] During the specific implementation process, the device also includes: a preset protection configuration knowledge graph construction module, and the preset protection configuration knowledge graph construction module is specifically used to: parse the SCD file of the target smart grid to obtain the protection configuration information of the target smart grid, the protection configuration information including the primary protection of the primary equipment, the near backup protection configuration information and the association between the protection and the tripping circuit breaker; use the second query statement to perform the shortest path query on the preset power grid topology knowledge graph to obtain the far backup protection information; obtain the action time constraint of the backup protection action from the preset protection setting list; construct a knowledge graph based on the protection configuration information, the far backup protection information and the action time constraint to obtain the preset protection configuration knowledge graph for characterizing the protection configuration and action time constraint of the primary equipment.

[0176] During the specific implementation process, the deductive reasoning module 3 is specifically used to: query the preset protection configuration knowledge graph using a third query statement based on each of the suspected faulty devices to obtain the expected events corresponding to each of the suspected faulty devices and the time constraints for the occurrence of the expected events; perform data screening based on the protection action alarm information and the circuit breaker tripping alarm information to obtain observed events; perform confidence calculation based on the expected events and the observed events to obtain the confidence of the observed events corresponding to each of the suspected faulty devices.

[0177] During the specific implementation process, the deductive reasoning module 3 is also used to: determine a first indicator value based on the actual time of occurrence of the observed event and the time constraint of the observed event in the preset protection configuration knowledge graph; determine a second indicator value based on the protection action type of the observed event; and calculate and process the first indicator value and the second indicator value using a preset credibility function to obtain the confidence of the observed event, so as to obtain the confidence of the observed event corresponding to each suspected faulty device.

[0178] During the specific implementation process, the evidence fusion reasoning module 4 is specifically used to: supplement the confidence of each expected observation event with evidence to obtain a target confidence matrix; use the DS evidence fusion method to fuse evidence based on the target confidence matrix to obtain an evidence fusion result; based on the evidence fusion result, the protection action alarm information when a fault occurs in the target smart grid, and the circuit breaker tripping alarm information, use the preset protection configuration knowledge graph to restore the fault to obtain the actual fault information of the target smart grid.

[0179] This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0180] Another embodiment of the present application provides a storage medium storing a computer program. When the computer program is executed by a processor, the following method steps are implemented:

[0181] Step 1: Receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0182] Step 2: Based on the protection action alarm information and the circuit breaker trip alarm information, a preset power grid topology knowledge graph is used to perform hypothesis reasoning to obtain a set of suspected fault devices;

[0183] Step 3: Based on the protection action alarm information and the circuit breaker trip alarm information, a preset protection configuration knowledge graph is used to perform deductive reasoning on each suspected faulty device in the suspected faulty device set to obtain the confidence level of the observed event corresponding to each suspected faulty device;

[0184] Step 4: Based on the confidence level of each expected observation event, evidence fusion reasoning is performed on each of the suspected faulty devices to obtain actual fault information of the smart grid.

[0185] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0186] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0187] The specific implementation process of the above method steps can be found in any of the above-mentioned embodiments of the smart grid fault deduction method based on knowledge graph, and this embodiment will not be repeated here.

[0188] This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0189] Another embodiment of the present application provides an electronic device, which may be a server, and the electronic device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external client via a network connection. When the electronic device program is executed by the processor, it implements the functions or steps on the server side of a smart grid fault deduction method based on a knowledge graph.

[0190] In one embodiment, an electronic device is provided, which may be a client. The electronic device includes a processor, a memory, a network interface, a display screen, and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external server via a network connection. When the electronic device program is executed by the processor, it implements the functions or steps on the client side of a smart grid fault deduction method based on a knowledge graph.

[0191] Another embodiment of the present application provides an electronic device, comprising at least a memory and a processor, wherein the memory stores a computer program, and the processor implements the following method steps when executing the computer program in the memory:

[0192] Step 1: Receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid;

[0193] Step 2: Based on the protection action alarm information and the circuit breaker trip alarm information, a preset power grid topology knowledge graph is used to perform hypothesis reasoning to obtain a set of suspected fault devices;

[0194] Step 3: Based on the protection action alarm information and the circuit breaker trip alarm information, a preset protection configuration knowledge graph is used to perform deductive reasoning on each suspected faulty device in the suspected faulty device set to obtain the confidence level of the observed event corresponding to each suspected faulty device;

[0195] Step 4: Based on the confidence level of each expected observation event, evidence fusion reasoning is performed on each of the suspected faulty devices to obtain actual fault information of the smart grid.

[0196] The specific implementation process of the above method steps can be found in any of the above-mentioned embodiments of the smart grid fault deduction method based on knowledge graph, and this embodiment will not be repeated here.

[0197] This application establishes a power grid topology knowledge graph and a protection configuration knowledge graph, which constitute a power grid accident deduction knowledge graph. On this basis, an accident deduction framework including hypothetical reasoning, deductive reasoning and fusion reasoning is proposed, which realizes the delineation of fault scope, hypothesis verification of suspected faults and fusion of multi-source evidence. The advantage of the knowledge graph-driven accident deduction framework proposed in this application is that, through explicit knowledge representation and independent reasoning mechanism, it not only improves the sharing, scalability and interpretability of knowledge, but also effectively handles time constraints and multi-source conflict evidence. The method of this application is more accurate in restoring smart grid faults.

[0198] The above embodiments are merely exemplary embodiments of the present application and are not intended to limit the scope of the present application. The scope of protection of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and scope of protection of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the scope of protection of the present application.

Claims

1. A smart grid fault deduction method based on knowledge graph, characterized in that: include: Receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid; Based on the protection action alarm information and the circuit breaker trip alarm information, a preset power grid topology knowledge graph is used to perform hypothesis reasoning to obtain a set of suspected fault devices; Based on the protection action alarm information and the circuit breaker trip alarm information, a preset protection configuration knowledge graph is used to perform deductive reasoning on each suspected faulty device in the suspected faulty device set to obtain an observed event confidence level corresponding to each suspected faulty device; Based on the confidence level of each expected observation event, evidence fusion reasoning is performed on each suspected faulty device to obtain actual fault information of the smart grid.

2. The method according to claim 1, wherein Before performing hypothesis reasoning using a preset power grid topology knowledge graph based on the protection action alarm information and the circuit breaker trip alarm information, the method further includes: constructing a preset power grid topology knowledge graph; The construction of the preset power grid topology knowledge graph specifically includes: Obtaining a topological network structure diagram of the target smart grid; Building a model based on the topological network structure diagram to obtain a CIM / E model corresponding to the target smart grid; A knowledge graph is constructed based on the CIM / E model to obtain the preset grid topology knowledge graph corresponding to the target smart grid, which represents the topological connection relationship between devices.

3. The method according to claim 1, wherein The method of performing hypothetical reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices specifically includes: Extracting data based on the protection action alarm information and the circuit breaker trip alarm information to obtain a tripped circuit breaker set; Constructing a mapping relationship between each tripped circuit breaker in the tripped circuit breaker set to obtain each tripped circuit breaker group; For each of the tripped circuit breaker groups, a first query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain a first suspected fault device of the shortest path corresponding to each of the tripped circuit breaker groups; For the line primary devices in the first suspected faulty devices, a second query statement is used to perform a shortest path query on the preset power grid topology knowledge graph to obtain the second suspected faulty device of the shortest path corresponding to each of the line primary devices; Performing a union calculation on the first suspected faulty device and the second suspected faulty device to obtain the suspected faulty device set; The primary equipment includes lines, busbars and transformers.

4. The method according to claim 1, wherein Before performing deductive reasoning on each suspected faulty device in the suspected faulty device set using a preset protection configuration knowledge graph based on the protection action alarm information and the circuit breaker trip alarm information, the method further includes: constructing a preset protection configuration knowledge graph; The construction of the preset protection configuration knowledge graph specifically includes: Parsing the SCD file of the target smart grid to obtain protection configuration information of the target smart grid, the protection configuration information including primary protection and backup protection configuration information of a primary device and an association relationship between protection and a tripped circuit breaker; Using a second query statement to perform a shortest path query on the preset power grid topology knowledge graph to obtain remote backup protection information; Obtain the action time constraint of the backup protection action from the preset protection setting list; A knowledge graph is constructed based on the protection configuration information, the remote backup protection information and the action time constraint to obtain the preset protection configuration knowledge graph for characterizing the protection configuration and action time constraint of the primary device.

5. The method according to claim 1, wherein The method of performing deductive reasoning on each suspected faulty device in the suspected faulty device set based on the protection action alarm information and the circuit breaker trip alarm information using a preset protection configuration knowledge graph to obtain the confidence level of the observed event corresponding to each suspected faulty device specifically includes: Based on each of the suspected faulty devices, a third query statement is used to query the preset protection configuration knowledge graph to obtain an expected event corresponding to each of the suspected faulty devices and a time constraint for the occurrence of the expected event; Performing data screening based on the protection action alarm information and the circuit breaker trip alarm information to obtain an observation event; Confidence calculation is performed based on the expected event and the observed event to obtain the confidence of the observed event corresponding to each of the suspected faulty devices.

6. The method according to claim 5, wherein The confidence calculation process based on the expected event set and the observed event set to obtain the confidence of the observed event corresponding to each of the suspected faulty devices specifically includes: Determine a first indicator value based on the actual time of occurrence of the observed event and the time constraint of the observed event in the preset protection configuration knowledge graph; determining a second indicator value based on the protection action type of the observed event; A preset confidence function is used to perform calculation based on the first indicator value and the second indicator value to obtain the confidence of the observed event, so as to obtain the confidence of the observed event corresponding to each of the suspected faulty devices.

7. The method according to claim 1, wherein The performing of evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observed events to obtain actual fault information of the smart grid fault specifically includes: Supplement the confidence of each expected observation event with evidence to obtain a target confidence matrix; Based on the target confidence matrix, the DS evidence fusion method is used to perform evidence fusion to obtain an evidence fusion result; Based on the evidence fusion result, the protection action alarm information and the circuit breaker tripping alarm information when the target smart grid fails, the fault is restored using the preset protection configuration knowledge graph to obtain the actual fault information of the target smart grid.

8. A smart grid fault deduction device based on knowledge graph, characterized in that: include: A receiving module, used to receive protection action alarm information and circuit breaker trip alarm information when a fault occurs in the target smart grid; A hypothesis reasoning module, configured to perform hypothesis reasoning based on the protection action alarm information and the circuit breaker trip alarm information using a preset power grid topology knowledge graph to obtain a set of suspected fault devices; a deductive reasoning module, configured to perform deductive reasoning on each suspected faulty device in the set of suspected faulty devices based on the protection action alarm information and the circuit breaker trip alarm information using a preset protection configuration knowledge graph, and obtain an observed event confidence level corresponding to each suspected faulty device; The evidence fusion reasoning module is used to perform evidence fusion reasoning on each of the suspected faulty devices based on the confidence level of each of the expected observation events to obtain actual fault information of the smart grid.

9. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the smart grid fault deduction method based on the knowledge graph according to any one of claims 1 to 7 are implemented.

10. An electronic device, characterized in that: The method comprises at least a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program in the memory, the method implements the steps of the knowledge graph-based smart grid fault deduction method according to any one of claims 1 to 7.

Images