A smart grid fault diagnosis method, device, medium and equipment

Abstract

A smart grid fault diagnosis method, device, medium and equipment are disclosed. The method comprises: extracting protection and circuit breaker action signals from fault recording files collected after a power grid fault to form an observation event set; circumscribing a fault range according to a primary power grid structure and the observation event set to form a suspected fault equipment set; performing single-end fault location on a power transmission line in the suspected fault equipment using the fault recording files to form a fault location event set; performing reasoning on each suspected fault equipment in the suspected fault equipment set using a fault causal network to form an expected event set; and iteratively calculating the support degree of each event in the observation event set and the fault location event set using the expected event set to obtain a fault equipment set of the power grid fault.

Description

Technical Field

[0001] This invention relates to the field of power system relay protection technology, and more specifically, to a method, apparatus, medium, and equipment for fault diagnosis in smart grids. Background Technology

[0002] Modern society is increasingly reliant on a reliable power supply. When a power grid fault occurs, dispatchers need to quickly determine the location and cause of the fault in order to handle it correctly. However, complex power grid faults generate a large amount of event information in a short period, posing significant challenges to dispatchers in quickly and accurately diagnosing the fault. Therefore, complex power grid fault diagnosis technology has received widespread attention.

[0003] The development of power grid fault diagnosis technology has basically followed the development path of artificial intelligence (AI). Early methods, represented by expert systems, achieved diagnostic reasoning based on logical matching by building rule bases. However, bottlenecks existed in knowledge acquisition, updating, and maintenance, which hindered the widespread application of this technology.

[0004] As AI technology enters the data-driven stage, researchers widely employ methods such as deep learning, graphical models, and ensemble learning to uncover potential patterns between alarm information and equipment status, achieving significant progress in fault identification accuracy and efficiency. To improve model interpretability and the efficiency of prior knowledge utilization, some studies have introduced probabilistic graphical models such as Bayesian networks, combined with methods like rough sets and Hausdorff distance, effectively enhancing the ability to handle uncertain information. However, these methods rely on a large number of high-quality samples, have complex modeling structures, and lack interpretability and generality.

[0005] To overcome the data dependency problem, some research has shifted to a mechanism-driven approach, constructing explicit models based on primary system topology and protection configuration, and describing fault behavior through analytical modeling, optimization algorithms, or formal methods. However, this modeling process is complex and lacks universality.

[0006] After exploring expert systems, data-driven approaches, and mechanism-driven approaches, researchers have begun to investigate knowledge-driven methods, represented by knowledge graphs. This method possesses semantically structured expression and reasoning capabilities and is gradually being applied to key business processes in power systems: In dispatching and maintenance, research has constructed knowledge graph frameworks that integrate multi-source heterogeneous information to support grid operation status identification and anomaly handling assistance; in equipment management, existing research has constructed graphs based on logs, sensor data, and expert experience to achieve defect prediction, maintenance recommendations, and status assessment, promoting refined operation and maintenance; in fault diagnosis, some works attempt to integrate topology, protection configuration, and historical case knowledge to construct graphs, achieving unified modeling of fault handling logic, while other studies have introduced graph neural networks and graph embedding mechanisms to improve semantic modeling capabilities and fault identification accuracy. However, most current research remains at the stage of graph construction and static relationship modeling, lacking a unified graph structure covering primary systems and protection configurations, as well as reasoning mechanisms adaptable to multiple fault scenarios, making it difficult to meet the needs of engineering practice for clarity of diagnostic logic and system scalability.

[0007] Regarding the information sources relied upon for power grid fault diagnosis, existing methods all depend on alarm information such as protection actions and circuit breaker trips, without integrating fault location results. In fact, fault recorders collect fault information independently of protection systems. In complex scenarios such as main protection or circuit breaker failure to operate, faults being cleared by remote backup protection actions, and loss of action event information, fault location can provide independent and direct fault range determination results. Human experts also typically regard fault location as an important auxiliary decision-making basis for fault diagnosis. Therefore, integrating fault location results in fault diagnosis has significant value. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method, apparatus, medium, and equipment for diagnosing faults in smart grids.

[0009] According to one aspect of the present invention, a method for diagnosing faults in a smart grid is provided, comprising:

[0010] Extract protection and circuit breaker action signals from fault recording files collected after a power grid fault to form an observation event set;

[0011] Based on the primary power grid structure and the set of observed events, the fault range is delineated, and a set of suspected faulty equipment is formed.

[0012] For transmission lines in suspected faulty equipment, single-end fault location is performed using fault recording files to form a fault location event set;

[0013] Based on each suspected faulty device in the suspected faulty device set, reasoning is performed using a fault causal network to form a set of expected events;

[0014] By iteratively calculating the support of each event in the expected event set and the fault location event set, the set of faulty devices in the power grid fault is obtained.

[0015] Optionally, based on the primary power grid structure and the set of observed events, the fault range is delineated to form a set of suspected faulty equipment, including:

[0016] Construct a power grid diagram structure based on the primary power grid structure;

[0017] Extract the set of tripped circuit breakers and construct all pairs of tripped circuit breakers in the set;

[0018] Based on the power grid diagram structure, calculate the shortest path for each pair of circuit breakers in the full combination to determine multiple equipment sets;

[0019] The intersection of multiple device sets is used to obtain the set of suspected faulty devices.

[0020] Optionally, based on each suspected faulty device in the suspected faulty device set, reasoning is performed using a fault causal network to form a set of expected events, including:

[0021] Construct a fault causal network for multiple line faults;

[0022] Search for the fault causal network of each suspected faulty device in the suspected faulty device set;

[0023] Based on the fault causal network of each suspected faulty device, the expected event set of each suspected faulty device in the suspected faulty device set is obtained.

[0024] Optionally, the structure of the fault causal network is as follows:

[0025] The root node of each fault causal network represents a single equipment fault event, the edges represent causal relationships, and the intermediate nodes and leaf nodes represent protection action events, circuit breaker tripping events, or fault location events.

[0026] Each edge represents an ordered pair of events.<c,e,T> Where 'c' represents the cause event; 'e' represents the effect event; and 'T' represents the time constraint. This means that event e must occur after event c, and that e cannot occur earlier than t and cannot occur later than t relative to c.

[0027] Optionally, the expression for calculating the event support in the observed event set is:

[0028]

[0029] In the formula, τ represents the temporal consistency between the observed event and the expected event, specifically assigned the following value: τ = 1: Satisfaction and Time constraints between them; τ = 0.5: Beyond and The upper limit of the time constraint between them; τ = 0: lower than The lower bound of the time constraint, or not appearing in In the full time constraint table; D represents the protection action type, and the specific values ​​are as follows: D=1: main protection action or corresponding circuit breaker tripping; D=2: failure protection action or corresponding circuit breaker tripping; D=3: backup protection action or corresponding circuit breaker tripping;

[0030] The expression for calculating the event support in the fault location event set is:

[0031] m ij =tL

[0032] In the formula, t is used to characterize whether the occurrence time of the fault location event meets the constraints, and the specific values ​​are as follows: t = 1: The fault recording on which the fault location event depends is generated within the specified time limit; t = 0.5: It means that the fault recording starts too slowly, and the reliability of the fault location event generated based on it is not high; t = 0: The fault location event is earlier than the time of a previous equipment failure, which means that the corresponding fault recording data does not correspond to this fault; L is used to set the extent to which the fault location result can support the fault hypothesis, and the values ​​are as follows: L = 1: If the single-end fault location result is 0% to 90% of the total line length; L = 0: The single-end fault location result is greater than 110% of the total line length, or the fault location fails; L ∈ (0, 1): The single-end fault location result is in the range of 90% to 110% of the total line length, and L is taken as a value in the range of 0 to 1. When the location result is 100% of the total line length, L = 0.5.

[0033] Optionally, the support degree of each event in the observed event set and the fault location event set is calculated iteratively using the expected event set to obtain the set of faulty devices in the power grid fault, including:

[0034] Calculate the support of each event in the expected event set and the fault location event set using the expected event set;

[0035] The support levels for all anticipated events for each suspected faulty device are fused to obtain fused evidence.

[0036] The suspected faulty device corresponding to the maximum value of the fused evidence is designated as the faulty device.

[0037] Remove faulty devices and recalculate the support of each event iteratively. In each iteration, select the suspected faulty device corresponding to the maximum value of the fused evidence as the faulty device, until there are no suspected faulty devices with fused evidence greater than the preset value, then stop the iteration and obtain the set of faulty devices.

[0038] Optionally, the computational expression for fused evidence is:

[0039]

[0040] In the formula, m1 and m2 are two sources of evidence, namely two observed events or fault location events; A represents the proposition to be judged, which here refers to the suspected faulty equipment; m1(Ai) represents the confidence of evidence source m1 in relation to the suspected fault Ai; m2(Bj) represents the confidence of evidence source m2 in relation to the suspected fault Bj; Ai∩Bj=A means that the intersection of Ai and Bj is A; m(A) represents the confidence of A after fusing evidence sources m1 and m2.

[0041] According to another aspect of the present invention, a smart grid fault diagnosis device is provided, comprising:

[0042] The first generation module is used to extract protection and circuit breaker action signals from the fault recording files collected after a power grid fault, and to form an observation event set;

[0043] The second forming module is used to delineate the fault range based on the primary power grid structure and the set of observed events, and to form a set of suspected faulty equipment.

[0044] The ranging module is used to perform single-end fault ranging for transmission lines in suspected faulty equipment using fault waveform files, forming a fault ranging event set.

[0045] The reasoning module is used to reason based on each suspected faulty device in the suspected faulty device set, using a fault causal network to form a set of expected events;

[0046] The calculation module is used to iteratively calculate the support of each event in the expected event set and the fault location event set to obtain the set of faulty devices in the power grid fault.

[0047] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the method described in any of the preceding aspects of the present invention.

[0048] Therefore, this invention consists of backward fault reasoning, single-end fault location, forward fault reasoning, "event-fault device" support assignment, and iterative evidence fusion. Compared with existing fault diagnosis methods, this invention has the following characteristics and advantages: (1) It proposes to represent fault diagnosis knowledge using a fault causal network, enabling explicit representation of knowledge and making the diagnosis results interpretable; (2) It proposes a bidirectional reasoning mechanism consisting of backward and forward reasoning, improving the coverage and precision of fault diagnosis; (3) It proposes a unified fault diagnosis framework that integrates fault location results and protection and circuit breaker action events, establishes a support calculation method for multiple fault events and an iterative evidence fusion method, enabling this invention to handle complex fault types such as multiple faults, incorrect protection actions, and missing fault information. Attached Figure Description

[0049] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures:

[0050] Figure 1 This is a flowchart illustrating a smart grid fault diagnosis method provided in an exemplary embodiment of the present invention;

[0051] Figure 2 This is another schematic flowchart of a smart grid fault diagnosis method provided in an exemplary embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of a single-line power grid topology provided in an exemplary embodiment of the present invention;

[0053] Figure 4 This is a schematic diagram of a Neo4j graph model of a single-line graph provided in an exemplary embodiment of the present invention;

[0054] Figure 5 This is a schematic diagram of the fault causal network of line L2 provided in an exemplary embodiment of the present invention;

[0055] Figure 6 This is a schematic diagram of a method for calculating the value of a single-ended fault location event L provided in an exemplary embodiment of the present invention;

[0056] Figure 7 This is a schematic diagram of an iterative evidence fusion diagnostic process for multiple fault scenarios provided by an exemplary embodiment of the present invention;

[0057] Figure 8 This is a schematic diagram of the on-site situation and expert diagnosis results of multiple faults in transformer D provided by an exemplary embodiment of the present invention;

[0058] Figure 9 This is a schematic diagram of the power grid topology under a single fault, provided in an exemplary embodiment of the present invention;

[0059] Figure 10 This is a schematic diagram of the power grid topology under multiple faults provided in an exemplary embodiment of the present invention;

[0060] Figure 11 This is a schematic diagram of the structure of a smart grid fault diagnosis device provided in an exemplary embodiment of the present invention;

[0061] Figure 12 This is the structure of an electronic device provided in an exemplary embodiment of the present invention. Detailed Implementation

[0062] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0063] It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention.

[0064] Those skilled in the art will understand that the terms "first," "second," etc., in the embodiments of the present invention are only used to distinguish different steps, devices, or modules, and do not represent any specific technical meaning, nor do they indicate a necessary logical order between them.

[0065] It should also be understood that in the embodiments of the present invention, "multiple" can refer to two or more, and "at least one" can refer to one, two or more.

[0066] It should also be understood that any component, data or structure mentioned in the embodiments of the present invention can generally be understood as one or more unless explicitly defined or given contrary instructions in the context.

[0067] Furthermore, the term "and / or" in this invention is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this invention generally indicates that the preceding and following related objects have an "or" relationship.

[0068] It should also be understood that the description of the various embodiments in this invention emphasizes the differences between the various embodiments, and the similarities or similarities can be referred to each other. For the sake of brevity, they will not be described in detail.

[0069] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0070] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use.

[0071] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the specification.

[0072] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0073] The embodiments of this invention can be applied to electronic devices such as terminal devices, computer systems, and servers, and can operate together with a wide range of other general-purpose or special-purpose computing system environments or configurations. Well-known examples of terminal devices, computing systems, environments, and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, and servers include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments including any of the above systems, etc.

[0074] Electronic devices such as terminal devices, computer systems, and servers can be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. Typically, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in distributed cloud computing environments, where tasks are executed by remote processing devices linked through communication networks. In distributed cloud computing environments, program modules can reside on local or remote computing system storage media, including storage devices.

[0075] Exemplary methods

[0076] Figure 1 This is a schematic flowchart of a smart grid fault diagnosis method provided in an exemplary embodiment of the present invention. This embodiment can be applied to electronic devices, such as... Figure 1 As shown, the smart grid fault diagnosis method 100 includes the following steps:

[0077] Step 101: Extract protection and circuit breaker action signals from the fault recording files collected after the power grid fault to form an observation event set;

[0078] Step 102: Delineate the fault range based on the primary power grid structure and the set of observed events, and form a set of suspected faulty equipment;

[0079] Step 103: For the transmission lines in the suspected faulty equipment, use the fault waveform file to perform single-end fault location and form a fault location event set.

[0080] Step 104: Based on each suspected faulty device in the suspected faulty device set, use the fault causal network to reason and form a set of expected events;

[0081] Step 105: Iteratively calculate the support of each event in the expected event set and the fault location event set to obtain the set of faulty devices in the power grid fault.

[0082] Specifically, in response to the technical problems existing in the background technology, this invention proposes a smart grid fault diagnosis method that integrates single-end fault location and protection action logic, which has the following characteristics and advantages: (1) It proposes to represent fault diagnosis knowledge with a fault causal network, so that the knowledge can be explicitly represented and the diagnosis results can be interpreted; (2) It proposes a bidirectional reasoning mechanism composed of backward reasoning and forward reasoning to improve the coverage and precision of fault diagnosis; (3) It proposes a unified fault diagnosis framework that integrates fault location results and protection and circuit breaker action events, establishes a support degree calculation method for multiple fault events and an iterative evidence fusion method, so that this invention can handle complex fault types such as multiple faults, incorrect protection actions and missing fault information.

[0083] The overall framework for smart grid fault diagnosis constructed in this invention is as follows: Figure 2 As shown, it consists of several steps, including backward fault reasoning, single-end fault ranging, forward fault reasoning, "event-fault device" support assignment, and iterative evidence fusion.

[0084] (1) Backward reasoning after a fault: Using fault recording files collected after a power grid fault as input, the protection and circuit breaker action signals are extracted to form an observation event set. This set is then combined with primary power grid structure information to delineate the fault range and form a set of suspected faulty devices. In this invention, "equipment" is used to represent primary equipment such as lines, busbars, transformers, and circuit breakers, while "device" is used to represent intelligent electronic devices (IEDs) such as relay protection and fault recorders.

[0085] (2) Single-end fault location: For transmission lines with concentrated suspected faulty equipment, high-precision single-end fault location is performed using fault waveform files to form fault location events. The events are used to describe the distance from the suspected fault point to the beginning of the line. The location event can support or deny a fault on a certain line and can be used for subsequent reasoning.

[0086] (3) Fault forward reasoning: For each primary device in the suspected faulty device set, the present invention uses the fault causal network to reason and form the expected event set. That is, assuming that the primary device is a faulty device, then according to the protection function logic and the configuration of the fault recorder, what events should be generated (including protection action events, circuit breaker tripping events and fault location events, etc.).

[0087] (4) Support assignment for “event-faulty equipment”: This invention treats each observed event as an independent source of evidence, and each source of evidence provides different levels of support to each suspected faulty equipment. To this end, this invention proposes a support assignment algorithm.

[0088] (5) Iterative evidence fusion. A support matrix is ​​formed based on the support assignment results. Each row in this matrix represents a suspected faulty device, and each column represents an observed event (including protection action events, circuit breaker action events, and fault location events). Since power grid faults may involve complex situations such as multiple faults, protection malfunctions, and missing fault information, this invention adopts an iterative evidence fusion method, which forms the final fault diagnosis result by applying the DS evidence fusion method multiple times.

[0089] In the aforementioned fault diagnosis process, backward reasoning aims to improve the coverage of fault diagnosis, that is, to ensure that no real faulty equipment is missed; while other processes, including forward reasoning, aim to improve the precision of fault diagnosis, that is, to exclude non-faulty equipment as much as possible.

[0090] Compared to traditional rule-based fault diagnosis methods, this invention represents knowledge using a fault causal network. This knowledge directly originates from protection configurations and coordination relationships, offering advantages such as visualization and intuitiveness. Furthermore, the fault causal network can conveniently represent the temporal constraints of events. Compared to data-driven fault diagnosis methods, the knowledge in this invention's framework has explicit representation capabilities, allowing for independent improvement of reasoning methods. This enhances knowledge sharing, strengthens system scalability and interpretability, and effectively overcomes the "black box" problem of knowledge and reasoning in data-driven methods. In addition, this invention's framework integrates multiple event types such as fault location, protection actions, and circuit breaker tripping, improving the diagnostic capabilities for complex faults.

[0091] Step 1: Backward reasoning about the fault

[0092] 1. Obtain primary power grid structure information

[0093] Primary power grid structure information includes the electrical connections between devices such as buses, lines, transformers, and circuit breakers, and can be represented as a graph structure. To improve the efficiency of topology analysis, this invention uses Neo4j to store and manage primary power grid structure information and leverages its built-in Cypher language to efficiently perform graph query operations. However, it should be noted that this invention is not limited to using Neo4j; any graph database can also be used.

[0094] Figure 3 It demonstrates a simple power grid topology. Figure 4 This corresponds to the Neo4j graph model.

[0095] 2. Fault backward reasoning method

[0096] When a power grid fault occurs, fault isolation is ultimately achieved by circuit breaker tripping. The area delineated by the tripped circuit breaker defines the fault zone, within which primary equipment is considered potentially faulty. The backward reasoning task can be formally described as follows:

[0097]

[0098] In the formula: The set of observed events includes actual observed protection actions and circuit breaker tripping events. The set index J represents the number of events. This is a set of suspected faulty devices, with the set index I representing the number of suspected devices. The set index will be omitted from the following text unless it causes confusion.

[0099] As can be seen from the above, backward fault reasoning infers the scope of a fault by analyzing the results of actions. Its goal is to narrow down the search space for subsequent fault diagnosis without overlooking any truly faulty devices. The specific steps are as follows:

[0100] 1) Extracting the set of tripped circuit breakers: Summarizing the information of tripped circuit breakers into a set. And construct all pairs of this set. Where O represents the total number of tripped circuit breakers. To improve the coverage of fault diagnosis, this invention integrates circuit breaker trip alarm and protection action alarm information into this set, assuming that the two types of information will not be lost simultaneously. Thus, even if one type of alarm information is missing (e.g., only protection action information is available, but circuit breaker trip information is lacking), potential faulty devices will not be missed.

[0101] 2) Search for suspected faulty equipment: for each circuit breaker pair This invention queries the shortest path between two points in the primary power grid structure and extracts the primary equipment contained in the path into a set of equipment.

[0102] This invention uses Dijkstra's algorithm, supported by the Cypher language, to implement the aforementioned shortest path query. Figure 3 Taking the simple power grid shown as an example, if we query the shortest path between circuit breakers CB3 and CB6, the query results will include the following primary equipment: bus B3, lines L2 and L3, and circuit breakers CB4 and CB5.

[0103] 3) Form a set of suspected faulty devices: Traverse all tripped circuit breaker pairs to form a device set family. Take the union of these sets to form the final set of suspected faulty devices.

[0104] by Figure 3 Taking a simple power grid fault as an example, assume that the alarm information observed after the fault is as shown in Table 1. Using the above-described fault backward reasoning method, a set of suspected faulty devices can be obtained.

[0105] Table 1. Alarm information of protection and circuit breakers during simple power grid faults.

[0106]

[0107] Step 2: Single-ended fault location

[0108] If the fault is cleared by the remote backup protection, the power outage area will increase, and the set of suspected faulty devices may include more devices. This invention makes full use of fault recording data to target the set of suspected faulty devices. In this paper, single-ended fault location is performed on the transmission lines. The fault location results are also regarded as an observation event and used in subsequent fault reasoning, thereby further improving the accuracy of fault diagnosis, that is, eliminating non-faulty equipment as much as possible.

[0109] Compared to double-ended fault location, single-ended fault location can be performed even when fault waveform data on the opposite side is lost, thus improving the reliability of fault analysis. Furthermore, this invention requires that the single-ended fault location algorithm can still accurately calculate the fault location even in long lines or situations with transition resistance. To meet these requirements, this invention employs a single-ended fault location method that considers the distributed parameters of the transmission line to perform single-ended fault location.

[0110] by Figure 3 Taking the simple power grid fault shown as an example, the single-end fault location results are also listed in Table 1. The location results of L2B2 and L2B3 both support a fault in line L2, especially the location result of L2B3, which weakens the possibility of a fault caused by L2B. 3m Uncertainty caused by protection failure to operate; L3B4 ranging results not supporting line L3 faults, all of these will enhance the certainty of fault diagnosis results.

[0111] Step 3: Forward Fault Reasoning

[0112] 1. Fault-cause network

[0113] This invention proposes and constructs a Fault Causal Network, which is a tree structure with a single equipment failure event as the root node. Figure 3 Taking the simple power grid shown as an example, Figure 5 The fault causal network of line L2 is shown. The meaning of this fault causal network is further explained as follows:

[0114] 1) The root node of each fault causal network represents a primary equipment failure event, the edges represent causal relationships, and the intermediate and leaf nodes represent protection action events, circuit breaker tripping events, or fault location events. The entire fault causal network represents the set of events that may be triggered according to the protection function logic, assuming a primary equipment failure corresponding to the root node. This invention refers to this as the expected event set under this fault assumption.

[0115] 2) Each edge represents an ordered pair of events.<c,e,T> Where 'c' represents the cause event; 'e' represents the effect event; and 'T' represents the time constraint. This means that event e must occur after event c, and that e cannot occur earlier than t and cannot occur later than t relative to c.

[0116] This invention uses directed edges and sets edge weights to represent the above causal relationship and time constraint, and stipulates that (1) the weight of the directed edge from c to e is set to a positive number. (2) The weight of the directed edge from e to c is set to a negative number -t. In particular, for fault location events, this invention uses the start time of fault recording as the event time.

[0117] 3) The protection devices on both sides of the line are identified by “equipment + busbar”. The subscripts m, f, and b are used to distinguish between main protection, failure protection and backup protection. FL is used to represent fault location.

[0118] For example, in Figure 3 In the diagram, the bidirectional directed edge between root node L2 and L2B3m indicates that when a fault occurs on line L2, the main protection of line L2, which is closer to bus B3, needs to operate within a time window of [10,40] ms.

[0119] In order to construct the above-mentioned fault causal network, the present invention requires the following two types of input information: (1) Field protection configuration and coordination information: including the main protection, failure protection, and backup protection configuration of primary equipment, as well as the correlation between protection and tripping circuit breaker; (2) Action time constraints: the action time of failure protection and backup protection can be obtained from the protection setting sheet; for the action time of main protection and circuit breaker, the present invention adopts typical protection and circuit breaker action delay data.

[0120] Similar to the system topology information, this invention also uses Neo4j to store and manage the fault causal network of each device.

[0121] 2. Forward reasoning about faults

[0122] A set of suspected faulty devices is obtained through backward reasoning based on the fault in step 1 of this invention. Then, for each device in the set Perform forward reasoning. Its formal description is:

[0123]

[0124] In the formula: If If the device is faulty, then the set of events that should occur according to the causal network of the fault (referred to as the expected event set) is as follows.

[0125] It is important to emphasize that in fault causal networks<c,e> This represents the causal rules between events. However, the known conditions for fault diagnosis problems are usually the fault result e, so fault causal networks cannot be directly used for fault deduction. To this end, this invention first performs backward reasoning to obtain a set of suspected faulty devices, thereby transforming the problem into equation (2), and then fault causal networks can be applied for fault diagnosis.

[0126] This invention completes the forward fault reasoning according to the following steps:

[0127] 1) Retrieve suspected faulty devices from the Neo4j database. The corresponding fault causal network, when expanded, yields the expected set of events. For example, the fault causal network corresponding to line L2 is as follows: Figure 3 As shown.

[0128] 2) Convert the fault causal network into an All-Paris Time Constraints table, which gives the time constraint relationships between each pair of events in the fault causal network.

[0129] This invention utilizes Cypher queries to generate all-pair time constraints. Due to the presence of negative weights in the fault causal network (e.g.... Figure 3 In this invention, the Floyd-Warshall algorithm, which supports negative weights, is selected for implementation in this query. Figure 3 For example, the corresponding all-correct time constraint table is shown in Table 2.

[0130] Table 2. Time Constraints for All Correct Answers on Line L2 (ms)

[0131]

[0132] Table 2 shows that the shortest time from a fault in line L2 to the tripping of circuit breaker CB6 is 1140 ms, while the shortest time from the tripping of circuit breaker CB6 to a fault in line L2 is -620 ms. These results indicate that if line L2 is faulty, the remote tripping event time constraint for circuit breaker CB6 is [620, 1140] ms. Using Table 1, it is possible to quickly determine whether the time constraint is satisfied between any two events, providing a crucial basis for subsequent fault reasoning.

[0133] Step 4: Assigning support values ​​for "Event-Faulty Devices"

[0134] because It may not be an actual faulty device, and is also affected by complex factors such as protection malfunctions and missing information, resulting in an expected event set. Possibly related to the actual observed event set Not entirely consistent. Therefore, this invention uses m ij Represents observed events For the fault hypothesis Support. In multi-source evidence fusion theory, support represents the original, localized measure of trust that a single source of evidence gives to a proposition.

[0135] The support calculation method of this invention for protection actions and circuit breaker tripping events is as follows:

[0136]

[0137] In the formula: (1) τ: the temporal consistency between the observed event and the expected event, specifically assigned the following value: τ=1: Satisfaction and Time constraints between them; τ = 0.5: Beyond and The upper limit of the time constraint between them (may be due to slow operation of protection or circuit breaker); τ = 0: lower than The lower bound of the time constraint, or not appearing in The time constraint table may be affected by protection or circuit breaker malfunction or failure to issue an action signal.

[0138] (2) D: Protection action type. The specific values ​​are as follows: D=1: Main protection action or corresponding circuit breaker tripping; D=2: Failure protection action or corresponding circuit breaker tripping; D=3: Backup protection action or corresponding circuit breaker tripping.

[0139] For fault location events, this invention uses the following support calculation formula:

[0140] m ij =tL (4)

[0141] In the formula: (1) t: similar to formula (3), it represents whether the occurrence time of the fault location event meets the constraint. The specific values ​​are as follows: t = 1: the fault recording on which the fault location event depends is generated within the specified time limit; t = 0.5: it means that the fault recording starts too slowly, and the reliability of the fault location event generated accordingly is not high; t = 0: the fault location event is earlier than the time of a previous equipment failure, which means that the corresponding fault recording data does not correspond to this failure.

[0142] (2) L: Used to set the extent to which fault location results support the fault hypothesis. See [link to relevant documentation] for the assignment method. Figure 6 The explanation is as follows: L=1: If the single-end fault location result is 0% to 90% of the total line length; L=0: If the single-end fault location result is greater than 110% of the total line length, or the fault location fails; L∈(0,1): If the single-end fault location result is within the range of 90% to 110% of the total line length, then... Figure 4 The linearization method takes L as a value in the range of 0 to 1. Among them, when the ranging result is 100% of the total length of the line, L = 0.5.

[0143] It should be noted that the above algorithm takes into account the inherent error in fault location. As the accuracy of the single-end fault location algorithm improves, the calculation method for the L value can be modified. For example, when the single-end fault location result is sufficiently accurate, L = 1 can be used when the location result is between 0% and 100%, while L = 0 is used in all other cases.

[0144] by Figure 3 Taking the simple power grid fault shown and the alarm information in Table 1 as examples, the support is calculated using the above method, and all support values ​​m are... ij The organization is represented by a support matrix of size I×J, as shown in Table 3. In this matrix, each row corresponds to a suspected faulty device, each column corresponds to an observed event, and the value in the cell represents the support of that observed event for the faulty device.

[0145] Table 3 Support Matrix for Simple Power Grid Faults

[0146]

[0147] Step 5: Iterative Evidence Fusion

[0148] In Table 3, if each column is considered an independent source of evidence, the problem lies in how to integrate J sources of evidence to determine the fault of I suspected faulty devices. To address this, the present invention introduces the DS evidence theory. [6] .

[0149] Dempster's evidence theory is a non-probabilistic uncertainty reasoning method applicable to information fusion from multiple evidence sources. Its core mechanism involves fusing multiple evidence sources within an identification framework Θ using Dempster's rule to obtain the final belief assignment. The belief function is m:2. Θ →[0,1] should satisfy:

[0150]

[0151] In the formula: Let A be a proposition, and m(A) be the reliability of A.

[0152] Without loss of generality, suppose there are two mutually exclusive sources of evidence, m1 and m2. Then the fusion result regarding proposition A is:

[0153]

[0154] In the formula: The 1-K in the denominator is to ensure that the fusion result still satisfies the normalization condition of equation (6).

[0155] To address the multiple fault diagnosis problem of this invention, evidence fusion faces two challenges: first, multiple evidence sources may assign a support degree of 1 to different faulty devices, easily leading to typical "complete conflict"; second, a single fault moment may correspond to multiple actual faulty devices, requiring multi-target identification. Therefore, this invention proposes an iterative evidence fusion method for multiple fault scenarios, thereby improving the precision of fault diagnosis. Specifically, it eliminates non-faulty devices through evidence fusion and iteratively processes multiple fault scenarios. The overall process of iterative evidence fusion includes three stages: evidence fusion preprocessing, fusion calculation and preliminary judgment, and iterative judgment of faulty devices. The specific diagnostic process is as follows: Figure 7 As shown.

[0156] This invention will collect suspected faulty devices. As an identification framework for evidence fusion, additional rows are introduced in the summary table to meet the support normalization requirement in equation (6). The bank assigned a support level that could not be attributed to any faulty device to indicate insufficient evidence.

[0157] In the case of multiple failures, if multiple sources of evidence assign a support of 1 to different propositions, it may lead to fusion failure (conflict factor K = 1). To mitigate this problem, this invention merges all columns that assign a value of 1 to only a single proposition, and after uniform normalization, they are treated as a group of evidence to participate in fusion, thereby reducing the intensity of conflict and improving stability.

[0158] Based on Table 3, Tables 4 and 5 are obtained through this diagnostic process, indicating that the actual faulty device in this fault is line L2.

[0159] Table 4. Preliminary evidence fusion results for simple power grid faults.

[0160]

[0161]

[0162] Table 5. Iterative Evidence Fusion Results for Simple Power Grid Faults

[0163]

[0164] Ultimately, the fault diagnosis results for this simple power grid fault event are as follows: Line L2 experienced a fault at 38ms; main protection devices L2B2m and L2B3m operated at 50ms and 51ms respectively, causing circuit breaker CB3 to trip at 65ms; backup protection device L3B4b operated at 700ms, causing circuit breaker CB6 to trip at 725ms. Circuit breaker CB9 malfunctioned at 300ms. Circuit breaker CB4 failed to operate.

[0165] In a specific embodiment of the present invention, Figure 8 The on-site situation and expert diagnosis results of multiple faults in transformer D are presented.

[0166] To facilitate the explanation of the principle of the method of this invention, the preceding text adopts... Figure 3 The simplified power grid shown is used for illustration. In this section, two historical fault cases from the East China Power Grid are selected for verification: one is a single fault event that occurred on the SM line; the other is a multiple fault event that occurred at the D substation.

[0167] 1. Fault diagnosis of a single fault

[0168] Fault alarms were received from both transformer M and transformer S. At the time of the fault, the topology of the two stations was as follows: Figure 9 As shown in Table 6, the relevant alarm information is as follows.

[0169] Table 6. Alarm information of protection and circuit breakers during a single fault (hour:minute:second)

[0170]

[0171]

[0172] Given the implementation of a dual configuration of the main protection devices on site, entries in the table with subscripts m,1 or m,2 represent alarm information generated by the first or second set of protection devices, respectively. Furthermore, since most related protection and circuit breaker alarms occur instantaneously after a fault, only the time in seconds of alarm occurrence is recorded.

[0173] 1) Backward reasoning about faults

[0174] Based on the primary power grid structure of transformers M and S and Table 6, a set of suspected faulty equipment is obtained through backward reasoning.

[0175] 2) Forward reasoning about faults

[0176] Based on the fault causal network of each suspected faulty device and Table 6, the support matrix is ​​obtained by forward reasoning, as shown in Table 7.

[0177] Table 7 Support Matrix for Single Fault

[0178]

[0179] 3) Iterative evidence fusion

[0180] According to Table 7, the actual faulty device in the case of a single fault is the SM line, as determined by evidence fusion.

[0181] 4) Fault diagnosis results

[0182] Based on the above reasoning, the fault diagnosis result and the complete fault occurrence process can be obtained as follows: At 23:563 on [Date], a single fault occurred on the SM line. The first and second sets of main protection for the M transformer on the SM line operated at 23:574 and 23:575 respectively, causing M transformer XXX2 and M transformer XXX3 to trip at 23:609 and 23:609 respectively; the first and second sets of main protection for the S transformer on the SM line operated at 23:579 and 23:587 respectively, causing S transformer XXX2 and S transformer XXX1 to trip at 23:611 and 23:622 respectively.

[0183] The above fault diagnosis results are consistent with the actual situation.

[0184] 2. Fault diagnosis of multiple faults

[0185] On [Date] at [Time]: 41.433 seconds and 48.758 seconds, the dispatch center received consecutive fault alarms from substation D. At the time of the fault, the topology of substation D was as follows: Figure 10 As shown in Table 8, the relevant alarm information is as follows. This fault is very typical, and the specific on-site situation and expert diagnosis results are shown in the appendix.

[0186] Table 8. Alarm information for protection and circuit breakers during multiple faults (hours:minutes:seconds)

[0187]

[0188] 1) Backward reasoning after failure

[0189] Based on the fault causal network of variable D and Table 8, a set of suspected faulty devices is obtained through backward reasoning.

[0190] It should be noted that, due to the lack of fault recording files from the opposite substation, it is impossible to identify all suspected faulty devices solely based on the shortest path query. To improve the completeness of fault identification, this invention introduces additional protection action alarm information as an auxiliary judgment basis in the backward reasoning when diagnosing multiple faults. Specifically, if the fault causal network of a device contains observed main protection action events, then that device is also determined to be a suspected faulty device.

[0191] Based on the above supplementary rules, the set of suspected faulty devices in this incident is as follows:

[0192] 2) Forward reasoning about faults

[0193] Based on the fault causal network of each suspected faulty device and Table 8, forward reasoning was performed on the two fault zero moments of X hour X minute 41.433 seconds and X hour X minute 48.758 seconds to obtain Tables 9 and 10.

[0194] 3) Iterative evidence fusion

[0195] Based on Tables 9 and 10, the evidence fusion results shown in Table 11 obtained through this diagnostic process indicate that the actual faulty equipment during the first fault of the multiple faults were the YX line and the WX line; the actual faulty equipment during the second fault was the 500kV busbar 2 of transformer D.

[0196] Table 9 Support matrix for the first failure in multiple failure scenarios

[0197]

[0198] Table 10 Support matrix of the second fault in multiple fault scenarios

[0199]

[0200] Table 11 Iterative evidence fusion results under multiple faults

[0201]

[0202] 4) Fault diagnosis results

[0203] Based on the above reasoning, the fault diagnosis results and the complete fault occurrence process can be obtained as follows:

[0204] (1) At 41:433 on [Date], 2021, both the WX and YX lines experienced faults simultaneously. The first and second sets of main protection devices for transformer D on the WX line and the first and second sets of main protection devices for transformer D on the YX line operated at 41:453, 41:455, 41:450, and 41:454, respectively, causing transformer D XXX1 to trip at 41:483. Transformers D XXX1 and D XXX2 were in hot standby mode before the fault.

[0205] (2) At 48:758 on [Date], a fault occurred on the 500kV Busbar 2 of Substation D. The first set of main protection on Busbar 2 of Substation D operated at 48:769, causing Substation XXX3, Substation XXXX3, Substation XXX3, and Substation XXX3 to trip at 48:788, 48:790, 48:791, and 48:792, respectively. Substation XXX3 was in hot standby before the fault; Substation XXX3 had already tripped during the first fault.

[0206] The above fault diagnosis results are consistent with the actual situation.

[0207] Therefore, a fault diagnosis method integrating single-end fault location and protection action logic is proposed for complex power grid faults. The method consists of backward fault reasoning, single-end fault location, forward fault reasoning, "event-fault equipment" support assignment, and iterative evidence fusion. Compared with existing fault diagnosis methods, this invention has the following features and advantages: (1) It proposes to represent fault diagnosis knowledge with a fault causal network, so that the knowledge can be explicitly represented and the diagnosis results can be interpreted; (2) It proposes a bidirectional reasoning mechanism consisting of backward and forward reasoning to improve the coverage and precision of fault diagnosis; (3) It proposes a unified fault diagnosis framework that integrates fault location results and protection and circuit breaker action events, establishes a support calculation method for multiple fault events and an iterative evidence fusion method, so that this invention can handle complex fault types such as multiple faults, incorrect protection actions, and missing fault information.

[0208] Exemplary device

[0209] Figure 11 This is a schematic diagram of the structure of a smart grid fault diagnosis device provided in an exemplary embodiment of the present invention. Figure 11 As shown, the device 1100 includes:

[0210] The first forming module 1110 is used to extract protection and circuit breaker action signals from the fault recording files collected after the power grid fault, and form an observation event set;

[0211] The second forming module 1120 is used to delineate the fault range based on the primary power grid structure and the set of observed events, and to form a set of suspected faulty equipment.

[0212] The ranging module 1130 is used to perform single-end fault ranging for transmission lines in suspected faulty equipment using fault waveform files, and form a fault ranging event set.

[0213] The reasoning module 1140 is used to reason based on each suspected faulty device in the suspected faulty device set using a fault causal network to form a set of expected events.

[0214] The calculation module 1150 is used to iteratively calculate the support of each event in the expected event set and the fault location event set to obtain the set of faulty devices in the power grid fault.

[0215] Optionally, the second forming module 1120 includes:

[0216] Construct a power grid diagram structure based on the primary power grid structure;

[0217] Extract the set of tripped circuit breakers and construct all pairs of tripped circuit breakers in the set;

[0218] Based on the power grid diagram structure, calculate the shortest path for each pair of circuit breakers in the full combination to determine multiple equipment sets;

[0219] The intersection of multiple device sets is used to obtain the set of suspected faulty devices.

[0220] Optionally, the inference module 1140 includes:

[0221] Construct a fault causal network for multiple line faults;

[0222] Search for the fault causal network of each suspected faulty device in the suspected faulty device set;

[0223] Based on the fault causal network of each suspected faulty device, the expected event set of each suspected faulty device in the suspected faulty device set is obtained.

[0224] Optionally, the structure of the fault causal network is as follows:

[0225] The root node of each fault causal network represents a single equipment fault event, the edges represent causal relationships, and the intermediate nodes and leaf nodes represent protection action events, circuit breaker tripping events, or fault location events.

[0226] Each edge represents an ordered pair of events.<c,e,T> Where 'c' represents the cause event; 'e' represents the effect event; and 'T' represents the time constraint. This means that event e must occur after event c, and that e cannot occur earlier than t and cannot occur later than t relative to c.

[0227] Optionally, the expression for calculating the event support in the observed event set is:

[0228]

[0229] In the formula, τ represents the temporal consistency between the observed event and the expected event, specifically assigned the following value: τ = 1: Satisfaction and Time constraints between them; τ = 0.5: Beyond and The upper limit of the time constraint between them; τ = 0: lower than The lower bound of the time constraint, or not appearing in In the full time constraint table; D represents the protection action type, and the specific values ​​are as follows: D=1: main protection action or corresponding circuit breaker tripping; D=2: failure protection action or corresponding circuit breaker tripping; D=3: backup protection action or corresponding circuit breaker tripping;

[0230] The expression for calculating the event support in the fault location event set is:

[0231] m ij =tL

[0232] In the formula, t is used to characterize whether the occurrence time of the fault location event meets the constraints, and the specific values ​​are as follows: t = 1: The fault recording on which the fault location event depends is generated within the specified time limit; t = 0.5: It means that the fault recording starts too slowly, and the reliability of the fault location event generated based on it is not high; t = 0: The fault location event is earlier than the time of a previous equipment failure, which means that the corresponding fault recording data does not correspond to this fault; L is used to set the extent to which the fault location result can support the fault hypothesis, and the values ​​are as follows: L = 1: If the single-end fault location result is 0% to 90% of the total line length; L = 0: The single-end fault location result is greater than 110% of the total line length, or the fault location fails; L ∈ (0, 1): The single-end fault location result is in the range of 90% to 110% of the total line length, and L is taken as a value in the range of 0 to 1. When the location result is 100% of the total line length, L = 0.5.

[0233] Optionally, the computing module includes:

[0234] Calculate the support of each event in the expected event set and the fault location event set using the expected event set;

[0235] The support levels for all anticipated events for each suspected faulty device are fused to obtain fused evidence.

[0236] The suspected faulty device corresponding to the maximum value of the fused evidence is designated as the faulty device.

[0237] Remove faulty devices and recalculate the support of each event iteratively. In each iteration, select the suspected faulty device corresponding to the maximum value of the fused evidence as the faulty device, until there are no suspected faulty devices with fused evidence greater than the preset value, then stop the iteration and obtain the set of faulty devices.

[0238] Optionally, the computational expression for fused evidence is:

[0239]

[0240] In the formula, m1 and m2 are two sources of evidence, namely two observed events or fault location events; A represents the proposition to be judged, which here refers to the suspected faulty equipment; m1(Ai) represents the confidence of evidence source m1 in relation to the suspected fault Ai; m2(Bj) represents the confidence of evidence source m2 in relation to the suspected fault Bj; Ai∩Bj=A means that the intersection of Ai and Bj is A; m(A) represents the confidence of A after fusing evidence sources m1 and m2.

[0241] Exemplary electronic devices

[0242] Figure 12 This is the structure of an electronic device provided in an exemplary embodiment of the present invention. For example... Figure 12 As shown, the electronic device 120 includes one or more processors 121 and memory 122.

[0243] The processor 121 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0244] The memory 122 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 121 may execute the program instructions to implement the methods of the software programs of the various embodiments of the present invention described above, and / or other desired functions. In one example, the electronic device may also include an input device 123 and an output device 124, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0245] In addition, the input device 123 may also include, for example, a keyboard, a mouse, etc.

[0246] The output device 124 can output various information to the outside. The output device 124 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0247] Of course, for the sake of simplicity, Figure 12 Only some of the components of the electronic device relevant to the present invention are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device may include any other suitable components depending on the specific application.

[0248] Exemplary computer program products and computer-readable storage media

[0249] In addition to the methods and apparatus described above, embodiments of the present invention may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps of the methods according to various embodiments of the present invention described in the "Exemplary Methods" section above.

[0250] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of the present invention. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0251] Furthermore, embodiments of the present invention may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps of the methods according to various embodiments of the present invention described in the "Exemplary Methods" section above.

[0252] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0253] The basic principles of the present invention have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in the present invention are merely examples and not limitations, and should not be considered as essential features of each embodiment of the present invention. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the present invention to the necessity of employing the aforementioned specific details.

[0254] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For system embodiments, since they largely correspond to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0255] The block diagrams of devices, systems, devices, and systems involved in this invention are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, systems, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0256] The methods and systems of the present invention may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the methods is for illustrative purposes only, and the steps of the methods of the present invention are not limited to the order specifically described above unless otherwise specifically stated. Furthermore, in some embodiments, the present invention may also be implemented as a program recorded on a recording medium, the program comprising machine-readable instructions for implementing the methods according to the present invention. Thus, the present invention also covers recording media storing programs for performing the methods according to the present invention.

[0257] It should also be noted that in the systems, apparatus, and methods of the present invention, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered equivalents of the present invention. The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the invention. Therefore, the invention is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0258] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the invention to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A smart grid fault diagnosis method integrating single-ended fault location and protection action logic, characterized in that, include: Extract protection and circuit breaker action signals from fault recording files collected after a power grid fault to form an observation event set; Based on the primary power grid structure and the observed event set, the fault range is delineated, forming a set of suspected faulty equipment, specifically: In the formula: The set of observed events includes actual observed protection actions and circuit breaker tripping events. The set index J represents the number of events. It is the set of suspected faulty devices, which is the set of expected events. The set index I is the number of suspected faulty devices. For the transmission lines in the suspected faulty equipment, the fault recording file is used to perform single-end fault location and form a fault location event set. Based on each suspected faulty device in the set of suspected faulty devices, reasoning is performed using a fault causal network to form a set of expected events; The support of each event in the observed event set and the fault location event set is calculated iteratively using the expected event set to obtain the set of faulty devices in the power grid fault. Based on each suspected faulty device in the set of suspected faulty devices, reasoning is performed using a fault causal network to form a set of expected events, including: Construct a fault causal network for multiple line faults; Retrieve the fault causal network of each suspected faulty device in the suspected faulty device set; Based on the fault causal network of each suspected faulty device, the expected event set of each suspected faulty device in the suspected faulty device set is obtained; The event support of the observed event set m ij The calculation expression is: In the formula, m ij For observing events For suspected faulty equipment Support τ To ensure temporal consistency between observed and expected events, the specific values ​​are assigned as follows: τ =1: Satisfaction and Time constraints between them; τ =0.5: Beyond and The upper limit of the time constraint between; τ =0: lower than The lower bound of the time constraint between, or Not appeared The full-pair time constraint table, wherein the full-pair time constraint table is transformed based on the fault causal network of suspected faulty equipment; D To specify the protection action type, the values ​​are as follows: D =1: The main protection action or the corresponding circuit breaker trips; D =2: Failure protection action or corresponding circuit breaker trip; D =3: Backup protection action or corresponding circuit breaker trip; The event support of the fault location event set m ij ’ The calculation expression is: m ij ’ ’ = tL In the formula, t This is used to characterize whether the occurrence time of the fault location event meets the constraint, and the specific values ​​are assigned as follows: t =1: The fault recording on which the fault location event depends was generated within the specified time limit; t =0.5: This indicates that the fault recording starts too slowly, and the reliability of the fault location event generated based on this is not high; t =0: The fault ranging event is earlier than the time of a previous equipment failure, which means that the corresponding fault waveform data does not correspond to this fault. L To what extent can the fault location results be used to support suspected faulty equipment? The values ​​are assigned as follows: L =1: If the single-end fault location result is 0%~90% of the total line length; L =0: The single-end fault location result is greater than 110% of the total line length, or the fault location failed; L ∈(0,1): Single-ended fault location results are within 90%~110% of the total line length. L The value is taken as a range of 0 to 1, where the distance measurement result is 100% of the total length of the line. L =0.

5.

2. The method according to claim 1, characterized in that, Based on the primary power grid structure and the observed event set, the fault range is delineated, forming a set of suspected faulty equipment, including: Construct a power grid diagram structure based on the primary power grid structure; Extract the set of tripped circuit breakers and construct all pairs of tripped circuit breakers in the set. Based on the power grid diagram structure, calculate the shortest path for each pair of circuit breakers in the full combination to determine multiple equipment sets; The intersection of the multiple sets of devices is used to obtain the set of suspected faulty devices.

3. The method according to claim 1, characterized in that, The structure of the fault causal network is as follows: The root node of each fault causal network represents a single equipment fault event, the edges represent causal relationships, and the intermediate nodes and leaf nodes represent protection action events, circuit breaker tripping events, or fault location events. Each edge represents an ordered pair of events.<c, e, T> Where 'c' represents the cause event; 'e' represents the effect event; and 'T' represents the time constraint. This means that event e should occur after event c, and that e cannot occur earlier than c. And it cannot be later than .

4. The method according to claim 1, characterized in that, By iteratively calculating the support of each event in the observed event set and the fault location event set using the expected event set, a set of faulty devices for power grid faults is obtained, including: Calculate the support of each event in the observed event set and the fault ranging event set using the expected event set; The support levels for all anticipated events for each suspected faulty device are fused to obtain fused evidence. The suspected faulty device corresponding to the maximum value of the fused evidence is designated as the faulty device; The support of each event is recalculated iteratively after removing the faulty devices. In each iteration, the suspected faulty device corresponding to the maximum value of the fused evidence is selected as the faulty device. The iteration is stopped when there are no suspected faulty devices with the fused evidence greater than the preset value, and a set of faulty devices is obtained.

5. The method according to claim 4, characterized in that, The calculation expression for the fused evidence is as follows: In the formula, m1 and m2 are two sources of evidence, namely two observed events or fault location events; A represents the proposition to be judged, here referring to the set of suspected faulty equipment; m1(A i () represents evidence source m1 for suspected faulty equipment set A i Reliability; m2(B j (m2 represents evidence source m2 for suspected faulty equipment set B) j Reliability; A i ∩B j =A represents A i and B j The intersection of these two sources is A; m(A) represents the confidence level of A after fusing evidence sources m1 and m2.

6. A smart grid fault diagnosis device integrating single-ended fault location and protection action logic, used to implement the method of claim 1, characterized in that, include: The first generation module is used to extract protection and circuit breaker action signals from the fault recording files collected after a power grid fault, and to form an observation event set; The second forming module is used to delineate the fault range based on the primary power grid structure and the observed event set, and to form a set of suspected faulty equipment. The ranging module is used to perform single-end fault ranging for the transmission line in the suspected faulty equipment using the fault waveform file, and form a fault ranging event set. The reasoning module is used to reason based on each suspected faulty device in the suspected faulty device set using a fault causal network to form a set of expected events; The calculation module is used to iteratively calculate the support of each event in the observed event set and the fault ranging event set using the expected event set, so as to obtain the set of faulty devices in the power grid fault.

7. The apparatus according to claim 6, characterized in that, The second forming module includes: Construct a power grid diagram structure based on the primary power grid structure; Extract the set of tripped circuit breakers and construct all pairs of tripped circuit breakers in the set. Based on the power grid diagram structure, calculate the shortest path for each pair of circuit breakers in the full combination to determine multiple equipment sets; The intersection of the multiple sets of devices is used to obtain the set of suspected faulty devices.

8. The apparatus according to claim 6, characterized in that, The structure of the fault causal network is as follows: The root node of each fault causal network represents a single equipment fault event, the edges represent causal relationships, and the intermediate nodes and leaf nodes represent protection action events, circuit breaker tripping events, or fault location events. Each edge represents an ordered pair of events.<c, e, T> Where 'c' represents the cause event; 'e' represents the effect event; and 'T' represents the time constraint. This means that event e should occur after event c, and that e cannot occur earlier than c. And it cannot be later than .

9. The apparatus according to claim 6, characterized in that, The calculation module includes: Calculate the support of each event in the observed event set and the fault ranging event set using the expected event set; The support levels for all anticipated events for each suspected faulty device are fused to obtain fused evidence. The suspected faulty device corresponding to the maximum value of the fused evidence is designated as the faulty device; The support of each event is recalculated iteratively after removing the faulty devices. In each iteration, the suspected faulty device corresponding to the maximum value of the fused evidence is selected as the faulty device. The iteration is stopped when there are no suspected faulty devices with the fused evidence greater than the preset value, and a set of faulty devices is obtained.

10. The apparatus according to claim 9, characterized in that, The calculation expression for the fused evidence is as follows: In the formula, m1 and m2 are two sources of evidence, namely two observed events or fault location events; A represents the proposition to be judged, here referring to the suspected faulty equipment; m1(A i () represents evidence source m1 regarding suspected fault A i Reliability; m2(B j () represents evidence source m2 for suspected fault B j Reliability; A i ∩B j =A represents A i and B j The intersection of these two sources is A; m(A) represents the confidence level of A after fusing evidence sources m1 and m2.

11. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for performing the method described in any one of claims 1-5.

12. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the executable instructions to implement the method described in any one of claims 1-5.

Images