Switchgear intelligent monitoring system and related methods based on knowledge graph

Abstract

This application provides a knowledge graph-based intelligent monitoring system and related methods for switchgear. A data acquisition module obtains first working data of a preset location of the intelligent switchgear within a first time period. An anomaly diagnosis module inputs the first working data into a first anomaly diagnosis model to obtain a first diagnosis result. An acquisition module acquires first attribute information of the preset location and second attribute information of the first working data. A reasoning module locates nodes in the corresponding knowledge graph of the intelligent switchgear monitoring system based on the first and second attribute information to obtain target nodes and acquires the first knowledge graph corresponding to the target nodes. Reasoning is performed on the first diagnosis result based on the first knowledge graph to obtain a first reasoning result. A feedback adjustment module adjusts the first anomaly diagnosis model based on the first reasoning result to obtain a second anomaly diagnosis model. Using this embodiment can improve the accuracy of anomaly diagnosis for switchgear.

Description

Technical Field

[0001] This application relates to the fields of computer technology or artificial intelligence technology, specifically to a knowledge graph-based intelligent monitoring system for switchgear and related methods. Background Technology

[0002] Switchgear, as a crucial electrical device in power systems, directly impacts the power supply quality and safety performance due to its operational reliability. Typically, switchgear insulation condition monitoring and anomaly diagnosis are essential technical means for timely maintenance, replacement, and prevention of insulation abnormalities. Switchgear insulation condition monitoring primarily targets partial discharge and the operating environment, utilizing various sensor technologies, computer technologies, network technologies, and communication technologies to assess the insulation status of power equipment. The development trend of switchgear condition monitoring technology is towards digitalization and intelligence. Through data acquisition and analysis technologies, real-time monitoring and maintenance of the switchgear operation process can be achieved, improving equipment reliability and safety. Applying artificial intelligence technology to intelligent online monitoring systems for switchgear enables functions such as equipment anomaly prediction, diagnosis, and repair.

[0003] Currently, the accuracy of anomaly diagnosis in switchgear is generally low. Therefore, the issue of how to improve the accuracy of anomaly diagnosis in switchgear urgently needs to be addressed. Summary of the Invention

[0004] This application provides a knowledge graph-based intelligent monitoring system and related methods for switchgear, which can improve the accuracy of anomaly diagnosis in switchgear.

[0005] In a first aspect, embodiments of this application provide a knowledge graph-based intelligent monitoring system for switchgear. This system includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module.

[0006] The acquisition module is used to acquire the first working data of the preset position of the intelligent switch cabinet in the first time period.

[0007] The anomaly diagnosis module is used to input the first working data into the first anomaly diagnosis model to obtain the first diagnosis result;

[0008] The acquisition module is used to acquire the first attribute information of the preset position and the second attribute information of the first working data;

[0009] The reasoning module is used to locate nodes in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, obtain target nodes, and acquire the first knowledge graph corresponding to the target nodes; and to reason about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result.

[0010] The feedback adjustment module is used to adjust the first anomaly diagnosis model according to the first inference result to obtain a second anomaly diagnosis model.

[0011] The acquisition module is also used to acquire second working data of the preset position of the intelligent switch cabinet in a second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the second time period.

[0012] The anomaly diagnosis module is also used to input the second working data into the second anomaly diagnosis model to obtain a second diagnosis result.

[0013] Secondly, embodiments of this application provide a knowledge graph-based intelligent monitoring method for switchgear. This method is applied to a knowledge graph-based intelligent monitoring system for switchgear, which includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The method includes:

[0014] The acquisition module obtains the first working data of the preset location of the intelligent switch cabinet in the first time period.

[0015] The first working data is input into the first anomaly diagnosis model through the anomaly diagnosis module to obtain the first diagnosis result;

[0016] The acquisition module obtains the first attribute information of the preset location and the second attribute information of the first working data.

[0017] The reasoning module locates the target node in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, and obtains the first knowledge graph corresponding to the target node; it then infers the first diagnostic result based on the first knowledge graph to obtain the first reasoning result.

[0018] The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model.

[0019] The acquisition module obtains the second working data of the preset position of the intelligent switch cabinet in the second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the second time period.

[0020] The second working data is input into the second anomaly diagnosis model through the anomaly diagnosis module to obtain the second diagnosis result.

[0021] Thirdly, embodiments of this application provide an electronic device, including a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for performing the steps in the second aspect of embodiments of this application.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the second aspect of embodiments of this application.

[0023] Fifthly, embodiments of this application provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the second aspect of embodiments of this application. The computer program product may be a software installation package.

[0024] Implementing the embodiments of this application has the following beneficial effects:

[0025] As can be seen, the knowledge graph-based intelligent monitoring system and related methods for switchgear described in this application embodiment include: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The data acquisition module acquires first working data of a preset location of the intelligent switchgear within a first time period. The anomaly diagnosis module inputs the first working data into a first anomaly diagnosis model to obtain a first diagnosis result. The acquisition module acquires first attribute information of the preset location and second attribute information of the first working data. The inference module locates nodes in the corresponding knowledge graph of the intelligent switchgear monitoring system based on the first and second attribute information to obtain target nodes and acquires the first knowledge graph corresponding to the target nodes. The first diagnosis result is inferred based on the first knowledge graph to obtain a first inference result. The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model. The data acquisition module acquires second working data of the preset location of the intelligent switchgear within a second time period. The start time of the second time period is later than the start time of the first time period. The second time period ends later than the first time period ends; the anomaly diagnosis module inputs the second working data into the second anomaly diagnosis model to obtain the second diagnosis result. Firstly, since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics, characterizing the changes in the working status of the switchgear at the preset location, the relevant nodes can be quickly located based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics, and the relevant knowledge graph can be accurately obtained to ensure the accuracy of knowledge graph acquisition, which helps to ensure the efficiency of subsequent reasoning. Moreover, since the relevant knowledge graph is accurately obtained, the accuracy and efficiency of the reasoning of the diagnosis result can be guaranteed. Secondly, not only can the anomaly diagnosis model be optimized using the reasoning result to ensure the accuracy of diagnosis, but also the working data after the first time period can be re-collected. Considering the continuity and gradual change of the anomaly, the randomness of anomaly detection can be prevented, and the anomaly can be accurately captured when it actually occurs, further ensuring and improving the accuracy of anomaly diagnosis of the switchgear, and also ensuring the intelligence of anomaly diagnosis of the switchgear. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the structure of a knowledge graph-based intelligent monitoring system for switchgear provided in an embodiment of this application;

[0028] Figure 2 This is a schematic diagram of another intelligent monitoring system for switchgear based on knowledge graphs provided in this application embodiment;

[0029] Figure 3 This is a schematic diagram of another knowledge graph-based intelligent monitoring system for switchgear provided in an embodiment of this application;

[0030] Figure 4 This is a schematic diagram illustrating an application scenario of a knowledge graph-based intelligent monitoring system for switchgear, as provided in an embodiment of this application.

[0031] Figure 5 This is a schematic diagram illustrating another application scenario of a knowledge graph-based intelligent monitoring system for switchgear provided in this application embodiment;

[0032] Figure 6 This is a flowchart illustrating a knowledge graph-based intelligent monitoring method for switchgear provided in an embodiment of this application.

[0033] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0034] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or modules is not limited to the listed steps or modules, but in one possible example includes steps or modules not listed, or in one possible example includes other steps or modules inherent to these processes, methods, products, or devices.

[0035] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0036] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0037] In this embodiment, the intelligent switchgear may include at least one of the following: high-voltage switchgear, low-voltage switchgear, ultra-high-voltage switchgear, etc., without limitation. Specifically, it enables real-time monitoring of the switchgear's operating status and environmental information. For example, the intelligent switchgear can collect data such as switch contact and cable joint temperatures, ambient temperature and humidity, mechanical characteristics of the operating mechanism, and partial discharge conditions. Specifically, based on the monitored data and trends, intelligent analysis can be used to assess the equipment's operating status in real time, predict its remaining lifespan, provide early warnings of equipment anomalies, and guide maintenance work. The system in this embodiment allows maintenance personnel to understand the equipment status before an accident occurs and take preventative measures, significantly reducing the probability of equipment accidents and providing a basis for implementing condition-based maintenance.

[0038] In this embodiment, a knowledge graph describes concepts, entities, and their relationships in the objective world in a structured form, expressing internet information in a way that is closer to human cognition, and providing a better ability to organize, manage, and understand the massive amounts of information on the internet. Knowledge graphs have revitalized internet semantic search and demonstrated powerful capabilities in intelligent question answering, becoming the infrastructure for knowledge-driven intelligent applications on the internet. Correspondingly, the knowledge graph of the switchgear intelligent monitoring system can include at least one knowledge graph. For example, different locations can correspond to different knowledge graphs, or the switchgear itself can correspond to a single knowledge graph, with each location corresponding to a node in that knowledge graph. In specific implementations, structured and unstructured data such as switchgear operation data, historical anomaly cases, and maintenance procedures can be collected, and knowledge extraction and relationship mapping can be performed to construct a switchgear knowledge graph that includes equipment component hierarchy, functional dependencies, and anomaly pattern associations.

[0039] In this embodiment, a switchgear knowledge graph, including equipment operation data and historical anomaly cases, can be constructed. Through knowledge extraction and relational reasoning, in-depth mining and cause tracing of equipment family anomalies can be achieved, improving the effectiveness of anomaly prediction. The knowledge graph can also be used to assist in decision-making and control model applications. For example, a model and strategy library including equipment status assessment, anomaly identification, and maintenance decision-making can be built. Based on the knowledge graph, the model can be continuously optimized and refined, ultimately achieving automated maintenance suggestion output and execution control.

[0040] In this application embodiment, the electronic device may include at least one of the following: switch cabinet (intelligent switch cabinet), cloud server, edge server, etc., which are not limited here.

[0041] Figure 1 This is a functional module block diagram of a knowledge graph-based intelligent monitoring system for switchgear, as described in this application embodiment. The knowledge graph-based intelligent monitoring system for switchgear includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, a reasoning module, and a feedback adjustment module.

[0042] The acquisition module is used to acquire the first working data of the preset position of the intelligent switch cabinet in the first time period.

[0043] The anomaly diagnosis module is used to input the first working data into the first anomaly diagnosis model to obtain the first diagnosis result;

[0044] The acquisition module is used to acquire the first attribute information of the preset position and the second attribute information of the first working data;

[0045] The reasoning module is used to locate nodes in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, obtain target nodes, and acquire the first knowledge graph corresponding to the target nodes; and to reason about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result.

[0046] The feedback adjustment module is used to adjust the first anomaly diagnosis model according to the first inference result to obtain a second anomaly diagnosis model.

[0047] The acquisition module is also used to acquire second working data of the preset position of the intelligent switch cabinet in a second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the second time period.

[0048] The anomaly diagnosis module is also used to input the second working data into the second anomaly diagnosis model to obtain a second diagnosis result.

[0049] The first time period can be preset or set by the system default. The second time period can also be preset or set by the system default, but its start time is later than the start time of the first time period, and its end time is later than the end time of the first time period. The first and second time periods may or may not overlap. The durations of the first and second time periods can be the same or different.

[0050] The first operating data may include at least one of the following: operating mode, operating current, operating voltage, operating power, operating temperature, operating pressure, operating humidity, etc., without limitation. The first operating data may include one or more types of data.

[0051] Different sensors can be installed at different locations within the switchgear to collect corresponding operational data.

[0052] The first anomaly diagnosis model can include any model used to realize the anomaly diagnosis of the switch cabinet. For example, the first anomaly diagnosis model can include at least one of the following: neural network model (such as convolutional neural network model, recurrent neural network model, etc.), linear regression model, large model (such as ChatGpt, deepseek, etc.), etc., without limitation.

[0053] For example, the data acquisition module may include at least one of the following: a temperature measurement terminal, a mechanical characteristic sensor, a cabinet temperature and humidity monitoring module, a data acquisition unit, a partial discharge monitoring module, a partial discharge acquisition module, a comprehensive status data concentrator, etc., without limitation. In this way, automatic online, real-time, multi-dimensional status information acquisition of the switchgear can be realized, constructing a comprehensive digital status sensing system driven by heterogeneous data such as temperature, partial discharge, mechanical characteristics, and temperature and humidity, replacing traditional manual status detection and inspection.

[0054] In practice, key parameters of the switchgear can be monitored and evaluated online. Specifically, various sensor devices can be used to automatically collect data on the switchgear's temperature, partial discharge, mechanical characteristics, humidity, and other conditions online. Combined with expert experience and risk models (the first anomaly diagnosis model), real-time health evaluation and anomaly warning of the switchgear can be achieved.

[0055] The preset location can be pre-set or be a system default, and can include one or more locations. The first attribute information of the preset location can include at least one of the following: location, sensor type, data type, environmental parameters, importance, sensitivity, accuracy, etc., which are not limited here.

[0056] The environmental parameters may include at least one of the following: ambient temperature, ambient humidity, atmospheric pressure, magnetic field interference intensity, etc., which are not limited here.

[0057] The data acquisition module can acquire the first working data of a preset location in the smart switch cabinet at preset time intervals within a first time period. The preset time interval can be pre-set or a system default. The preset time interval can be related to the first attribute information of the preset location. For example, a mapping relationship between the preset location's attribute information and the time interval can be pre-stored, and then the preset time interval corresponding to the first attribute information can be determined based on this mapping relationship.

[0058] The anomaly diagnosis module can input the first working data into the first anomaly diagnosis model to obtain a first diagnosis result. The first diagnosis result may include at least one of the following: a first anomaly type, a first probability value corresponding to the first anomaly type, an anomaly severity value corresponding to the first anomaly type, etc., without limitation. The first anomaly type may include at least one anomaly type, and different positions may correspond to different anomaly types. For example, when the first anomaly diagnosis model outputs multiple anomaly types, and each anomaly type includes a probability value, the largest probability value can be selected as the first probability value, and its corresponding anomaly type is the first anomaly type.

[0059] Accordingly, the second diagnostic result may include at least one of the following: target anomaly type, target probability value corresponding to the target anomaly type, anomaly severity value corresponding to the target anomaly type, etc., without limitation. The target anomaly type may include at least one anomaly type, and different positions may correspond to different anomaly types. For example, when the target anomaly diagnostic model outputs multiple anomaly types, and each anomaly type includes a probability value, the largest probability value can be selected as the target probability value, and the anomaly type corresponding to it is the target anomaly type.

[0060] The second attribute information of the first working data may include at least one of the following: data type, data structure, data stability, data source, data memory size, data change range, etc., without limitation.

[0061] The acquisition module can acquire first attribute information of a preset location. The first attribute information can reflect the physical characteristics of the preset location or the actual environmental characteristics of the preset location. Correspondingly, it can also acquire second attribute information of the first working data. The second attribute information reflects the data characteristics and characterizes the changes in the working status of the switchgear at the preset location.

[0062] In specific implementation, the reasoning module can locate nodes in the corresponding knowledge graph of the switchgear intelligent monitoring system based on the first attribute information and the second attribute information to obtain the target node and acquire the first knowledge graph corresponding to the target node. Since the first attribute information can reflect the physical characteristics of the preset location or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics and characterizes the changes in the working status of the switchgear at the preset location, the relevant nodes can be quickly located based on the relevant physical characteristics, environmental characteristics and relevant data characteristics, and the relevant knowledge graph can be accurately acquired, ensuring the accuracy of knowledge graph acquisition and helping to ensure the efficiency of subsequent reasoning.

[0063] Next, the first diagnostic result can be inferred based on the first knowledge graph to obtain the first reasoning result. Due to the accurate acquisition of the relevant knowledge graph, the accuracy and efficiency of the reasoning of the diagnostic result can be guaranteed.

[0064] Next, the feedback adjustment module can adjust the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model. This deep optimization of the anomaly diagnosis model based on the inference result further ensures the diagnostic accuracy of the anomaly diagnosis model. The acquisition module then obtains the second working data of the preset location of the intelligent switchgear in the second time period. The anomaly diagnosis module inputs the second working data into the second anomaly diagnosis model to obtain the second diagnostic result. On the one hand, this not only optimizes the anomaly diagnosis model using the inference result to ensure diagnostic accuracy, but also, by re-acquiring working data after the first time period, considering the continuity and gradual change of anomalies, it not only prevents the randomness of anomaly detection but also accurately captures anomalies when they actually occur, further ensuring and improving the accuracy of anomaly diagnosis for the switchgear, and also ensuring the intelligence of the anomaly diagnosis.

[0065] Furthermore, such as Figure 2As shown, the knowledge graph-based intelligent monitoring system for switchgear can also include an early warning module. When the second diagnostic result meets preset conditions, the early warning module can perform an early warning operation. For example, if the second diagnostic result includes: target anomaly type and target probability value corresponding to the target anomaly type, then the target early warning method corresponding to the target anomaly type can be determined. For example, a preset mapping relationship between anomaly types and early warning methods can be stored in advance. The early warning method can include at least one of the following: vibration mode, ringing mode, light emission mode, voice mode, video mode, call mode, etc., without limitation. Then, based on this mapping, the target early warning method corresponding to the target anomaly type and the target early warning parameter corresponding to the target probability value can be determined. A preset mapping relationship between the probability value and the early warning parameter of the target early warning method can also be stored in advance. Then, based on this mapping relationship, the target early warning parameter corresponding to the target probability value can be determined. Finally, the early warning module performs the corresponding early warning operation based on the target early warning method and the target early warning parameter. In this way, the accuracy and vigilance of the early warning can be improved, which helps to detect and resolve anomalies in a timely manner during maintenance.

[0066] In this embodiment of the application, a health auxiliary decision-making system for switchgear equipment can also be constructed based on knowledge graphs. This system can automatically combine actual on-site operating data, continuously perform deep learning based on correlation analysis to reason about family-based anomalies of power equipment, automatically match expert rules and maintenance auxiliary decision-making models, and achieve auxiliary decision-making.

[0067] In specific implementation, such as Figure 3 As shown in the embodiments of this application, the intelligent monitoring system for switchgear based on knowledge graphs may also include an upgrade module. This upgrade module can be used to realize the automatic maintenance and updating of the knowledge graph driven by real-time data, and to carry out family anomaly reasoning based on association rule mining. At the same time, the auxiliary decision model library is also continuously optimized through continuous learning. Of course, it can also realize the dual-loop collaboration and unification of knowledge graph and decision model.

[0068] In the embodiments of this application, such as Figure 4 As shown, the electronic device may include a smart switch cabinet, which can communicate with a server, which may include a cloud server, and the knowledge graph may be located on the cloud server.

[0069] In the embodiments of this application, such as Figure 5As shown, when electronic devices include edge servers, the edge servers can establish communication connections with switch cabinets and cloud servers. The knowledge graph can be located on the cloud server or the edge server, which can realize multi-dimensional joint analysis of health status evaluation applications based on edge computing. Specifically, it can realize the use of lightweight intelligent edge computing methods to achieve abnormal diagnosis and related early warning, support the joint analysis and evaluation of multi-dimensional heterogeneous status data at the switch cabinet end, reduce data upload latency, improve real-time response capabilities, and realize local status assessment and prediction early warning.

[0070] Of course, in practical applications, the effectiveness can also be evaluated through simulation test data and actual system integration, which may include indicators such as diagnostic accuracy and strategy standardization, and the decision model can be further refined and verified based on this.

[0071] Optionally, in the step of locating the target node by analyzing the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, the reasoning module is specifically used for:

[0072] Obtain the corresponding knowledge graph library of the intelligent monitoring system for the switchgear. The knowledge graph library includes at least one knowledge graph, each knowledge graph includes at least one node, and each node corresponds to a node tag set and a node index; the node tag set includes multiple node tags.

[0073] The first node tag set is determined based on the first attribute information;

[0074] The second node tag set is determined based on the second attribute information;

[0075] The first node tag set is matched with the node tag set of the knowledge graph in the at least one knowledge graph to obtain the target node tag set that is successfully matched.

[0076] Obtain the target node index corresponding to the target node tag set, and obtain the target node based on the target node index.

[0077] The knowledge graph library can include at least one knowledge graph. For example, it can include knowledge graphs corresponding to different models of switchgear, or it can include knowledge graphs corresponding to at least one switchgear. In specific implementations, each knowledge graph can include at least one node, and each node corresponds to a node tag set and a node index. The node index is used to uniquely identify a node; for example, the node index can be a node number. The node tag set can include multiple node tags, which can describe the relevant characteristics of the node. For example, a node tag can be temperature, or it can be pressure, etc., without limitation.

[0078] The first node label set may include at least one node label, and the second node label set may include at least one node label.

[0079] In specific implementation, the corresponding knowledge graph library of the intelligent monitoring system for switchgear can be obtained, and the mapping relationship between preset attribute information and node tags can be stored in advance. Then, the first node tag set corresponding to the first attribute information can be determined based on the mapping relationship. Similarly, the mapping relationship between the attribute information of preset working data and node tags can be stored in advance, and the second node tag set corresponding to the second attribute information can be determined based on the mapping relationship. Then, the union between the first node tag set and the second node tag set can be determined. Based on this union, it is matched with the node tag set of the knowledge graph in at least one knowledge graph to obtain the successfully matched target node tag set. Then, the target node index corresponding to the target node tag set is obtained, and the target node is obtained based on the target node index. Since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics and characterizes the changes in the working status of the switchgear at the preset location, the relevant nodes can be quickly locked based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics, which helps to ensure the efficiency of subsequent reasoning.

[0080] Optionally, in obtaining the first knowledge graph corresponding to the target node, the reasoning module is specifically used for:

[0081] Obtain the reference knowledge graph corresponding to the target node;

[0082] Determine the first keyword corresponding to the first attribute information;

[0083] Determine the second keyword corresponding to the second attribute information;

[0084] The reference knowledge graph is searched based on the first keyword to obtain the first part of the knowledge graph;

[0085] The second part of the knowledge graph is obtained by searching the reference knowledge graph based on the second keyword;

[0086] The first knowledge graph is determined based on the first part of the knowledge graph and the second part of the knowledge graph.

[0087] The first keyword may include at least one keyword, and the second keyword may include at least one keyword. The keywords may include at least one of the following: strings, Chinese characters, patterns, etc., without limitation.

[0088] In practice, a reference knowledge graph corresponding to the target node can be obtained, thus acquiring a relatively large knowledge graph. Then, keywords are extracted from the first attribute information to obtain the first keyword, and keywords are extracted from the second attribute information to obtain the second keyword. The first and second keywords reflect the actual needs. The reference knowledge graph is then searched based on the first keyword to obtain the first part of the knowledge graph, and the reference knowledge graph is searched based on the second keyword to obtain the second part of the knowledge graph. Finally, the first knowledge graph can be determined based on the first and second parts of the knowledge graph. For example, the first and second parts of the knowledge graph can be concatenated or merged to obtain the first knowledge graph. In this way, relevant knowledge graphs can be accurately obtained, ensuring the accuracy of knowledge graph acquisition.

[0089] Optionally, the first diagnostic result includes: a first abnormality type and a first probability value corresponding to the first abnormality type; in terms of reasoning about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result, the reasoning module is specifically used for:

[0090] The first knowledge graph is used to obtain a set of historical anomaly diagnosis results corresponding to the first attribute information and the second attribute information. The set of historical anomaly diagnosis results includes multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a diagnosis result authenticity identifier. The diagnosis result authenticity identifier includes a diagnosis result true identifier or a diagnosis result false identifier.

[0091] Determine the number of true diagnostic results among the multiple historical diagnostic results to obtain a first number;

[0092] Determine the number of false diagnostic results among the multiple historical diagnostic results to obtain a second number;

[0093] The first average value is obtained by averaging the second probability value corresponding to the first quantity;

[0094] The average value of the second probability value corresponding to the second quantity is calculated to obtain the second average value;

[0095] The first inference result is obtained by reasoning based on the first average value and the second average value.

[0096] The first diagnostic result may include: a first abnormality type and a first probability value corresponding to the first abnormality type.

[0097] The historical anomaly diagnosis result set can include multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a true / false identifier for the diagnosis result. The true / false identifier can include a true identifier or a false identifier. The true / false identifier can be automatically marked or manually marked (for example, after each anomaly warning, manual verification is performed to determine if it is truly normal or abnormal, and appropriate marking is done). Each historical diagnosis result can correspond to corresponding historical first attribute information and historical second attribute information. That is, the first attribute information and second attribute information can be matched with the corresponding historical first attribute information and historical second attribute information to obtain the successfully matched historical diagnosis results.

[0098] In specific implementation, the historical anomaly diagnosis result set may include the historical anomaly diagnosis result set of this intelligent switch cabinet, or it may include the historical anomaly diagnosis result set of other intelligent switch cabinets. For example, the third attribute information of the intelligent switch cabinet can be obtained, and the historical anomaly diagnosis result set can be obtained based on the third attribute information. The third attribute information may include at least one of the following: the model of the intelligent switch cabinet, the manufacturer of the intelligent switch cabinet, the purpose of the intelligent switch cabinet, the function of the intelligent switch cabinet, the performance of the intelligent switch cabinet, the usage environment of the intelligent switch cabinet, etc., which are not limited here.

[0099] Specifically, when matching the first attribute information and the second attribute information with their corresponding historical first attribute information and historical second attribute information, the first attribute information is matched with the historical first attribute information to obtain a first matching value, and the second attribute information is matched with the historical second attribute information to obtain a second matching value. Then, a weighted calculation is performed on the first and second matching values ​​to obtain the target matching value. If the target matching value is greater than a set threshold, the two are considered to have matched successfully, and vice versa. Alternatively, a first threshold and a second threshold can be set. If both the first and second matching values ​​are greater than the first threshold, the weighted calculation is performed to obtain the target matching value. Conversely, if the first matching value is less than or equal to the first threshold, or the second matching value is less than or equal to the second threshold, the two fail to match. The thresholds (first threshold, second threshold, etc.) can be preset or set by system default.

[0100] The first matching value corresponds to the first weight, the second matching value corresponds to the second weight, and the sum of the first weight and the second weight is 1. The first weight and the second weight can be preset or set by the system default.

[0101] A true diagnostic result means that the corresponding first abnormality type actually exists. A false diagnostic result means that the corresponding first abnormality type was misidentified, and the corresponding abnormality type does not exist.

[0102] In specific implementation, a set of historical abnormal diagnosis results corresponding to the first attribute information and the second attribute information can be obtained through a first knowledge graph. Then, the number of true diagnostic results among multiple historical diagnostic results is determined to obtain a first number, and the number of false diagnostic results among multiple historical diagnostic results is determined to obtain a second number. Then, an average value is calculated based on the second probability value corresponding to the first number, that is, the average value of the second probability value corresponding to the first number is calculated to obtain a first average value. Similarly, an average value is calculated based on the second probability value corresponding to the second number, that is, the average value of the second probability value corresponding to the second number is calculated to obtain a second average value. Finally, the first diagnostic result is inferred based on the first average value and the second average value to obtain a first inference result. Since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics and characterizes the changes in the working status of the switchgear at the preset location, the corresponding historical diagnostic situation can be determined based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics. Based on the historical diagnostic situation and the authenticity of the abnormal diagnosis, the probability screening criteria (first average value and second average value) are accurately determined, thereby ensuring the accuracy and efficiency of the diagnostic result inference.

[0103] Optionally, in the step of averaging the second probability value corresponding to the first quantity to obtain the first average value, the inference module is specifically used for:

[0104] The first reference average value is obtained by averaging the second probability value corresponding to the first quantity.

[0105] A first proportion value for identifying true diagnostic results is determined based on the first quantity and the second quantity;

[0106] A first optimization factor corresponding to the first ratio value is determined, and the value range of the first optimization factor is 0 to 0.1;

[0107] The first reference average value is obtained by optimizing the first optimization factor, specifically: First average value = (1 + first optimization factor) * first reference average value.

[0108] The value range of the first optimization factor can be preset or set by the system default. For example, the value range of the first optimization factor is 0 to 0.1. Of course, the value range of the first optimization factor can also be 0 to 0.2, which is not limited here.

[0109] In the specific implementation, the average value can be calculated based on the second probability value corresponding to the first quantity, that is, the mean of the second probability value corresponding to the first quantity is calculated to obtain the first reference average value. Then, the first proportion value of the diagnostic result as true is determined based on the first quantity and the second quantity. The first proportion value = first quantity / (first quantity + second quantity).

[0110] Next, a pre-stored mapping relationship between preset ratio values ​​and optimization factors can be stored. Then, a first optimization factor corresponding to the first ratio value can be determined based on this mapping relationship. The larger the first ratio value, the smaller the first optimization factor, and vice versa. Then, the first reference average value can be optimized based on the first optimization factor to obtain the first average value, specifically: first average value = (1 + first optimization factor) * first reference average value. In this way, the first ratio value reflects the historical accuracy of the diagnostic results. Determining the first average value based on historical diagnostic situations and the historical accuracy of diagnostic results, i.e., raising the threshold of the upper limit (first average value), can further prevent the randomness of anomaly detection, thereby ensuring the accuracy and efficiency of diagnostic result reasoning.

[0111] Optionally, in the operation of determining the second probability value corresponding to the second quantity to obtain the second average value, the inference module is specifically used for:

[0112] The second reference average value is obtained by averaging the second probability value corresponding to the second quantity.

[0113] The second reference average value is obtained by optimizing the first optimization factor, specifically: first average value = (1 - first optimization factor) * second reference average value.

[0114] In specific implementation, the average value can be calculated based on the second probability value corresponding to the second quantity, that is, the mean of the second probability value corresponding to the second quantity is calculated to obtain the second reference average value. Then, the second reference average value is optimized based on the first optimization factor to obtain the second average value. Specifically, the first average value = (1 - first optimization factor) * second reference average value. In this way, the first ratio value reflects the historical accuracy of the diagnostic results. The second average value is determined based on the historical diagnostic situation and the historical accuracy of the diagnostic results, that is, the threshold of the lower limit (second average value) is lowered, which can further prevent the anomaly detection from being missed. Thus, the accuracy and efficiency of the diagnostic result reasoning can be guaranteed.

[0115] Optionally, in the step of reasoning about the first diagnostic result based on the first average value and the second average value to obtain the first reasoning result, the reasoning module is specifically used for:

[0116] When the first probability value is greater than or equal to the second average value and less than or equal to the first average value, a first difference between the first probability value and the first average value is determined; the first inference result is determined based on the first difference.

[0117] When the first probability value is less than the second average value, the diagnostic result authenticity identifier of the first diagnostic result is determined to be a diagnostic result false identifier, and the first inference result is determined based on the diagnostic result false identifier;

[0118] When the first probability value is greater than the first average value, the diagnostic result authenticity identifier of the first diagnostic result is determined as the diagnostic result is true identifier, and the first inference result is determined based on the diagnostic result is true identifier.

[0119] In practical implementation, when the first probability value is greater than or equal to the second average value and less than or equal to the first average value, it indicates that the abnormal diagnosis result may be a misdiagnosis. Therefore, a first difference between the first probability value and the first average value can be determined: First difference = First probability value - First average value. The first average value (upper limit threshold) reflects the average probability of a true diagnosis among multiple historical diagnostic results, while the first difference reflects the degree of difference between the actual probability value and the upper limit threshold. Next, the first inference result is determined based on the first difference, i.e., the first inference result is the first difference. Furthermore, the inference result can be used to optimize the abnormal diagnosis model, ensuring diagnostic accuracy.

[0120] In specific implementation, when the first probability value is less than the second average value, it indicates that the abnormal diagnosis result is likely to be a misdiagnosis. Therefore, the true or false identifier of the first diagnosis result can be determined as the false identifier of the diagnosis result. For example, it can be determined that the first abnormal type in the first diagnosis result does not exist, and the first inference result can be determined based on the false identifier of the diagnosis result. That is, the false identifier of the diagnosis result is taken as the first inference result. Then, the step of the acquisition module re-acquiring the first working data of the preset position of the intelligent switch cabinet in the first time period can be carried out, thereby ensuring the accuracy of the diagnosis.

[0121] In specific implementation, when the first probability value is greater than the first average value, it means that the abnormal diagnosis result is highly likely to be a misdiagnosis. Therefore, the true or false identifier of the first diagnosis result can be determined as the true identifier of the diagnosis result, and the first inference result can be determined based on the true identifier of the diagnosis result. In other words, the true identifier of the diagnosis result can be used as the first inference result, thereby ensuring the accuracy of the diagnosis.

[0122] Optionally, when the first probability value is less than the second average value, the feedback adjustment module does not perform the step of adjusting the first anomaly diagnosis model according to the first inference result to obtain the second anomaly diagnosis model.

[0123] or,

[0124] When the first probability value is greater than the first average value, the feedback adjustment module does not perform the step of adjusting the first anomaly diagnosis model according to the first inference result to obtain the second anomaly diagnosis model.

[0125] In this embodiment of the application, when the first probability value is less than the second average value, the feedback adjustment module does not perform the step of adjusting the first abnormal diagnosis model according to the first reasoning result to obtain the second abnormal diagnosis model. This indicates that the abnormal diagnosis result is likely to be a misdiagnosis, that is, not optimizing the abnormal diagnosis model can improve the efficiency of abnormal diagnosis.

[0126] In this embodiment of the application, when the first probability value is greater than the first average value, the feedback adjustment module does not perform the step of adjusting the first abnormal diagnosis model according to the first reasoning result to obtain the second abnormal diagnosis model. This indicates that the abnormal diagnosis result is highly likely to be free of misdiagnosis, that is, without optimizing the abnormal diagnosis model, the efficiency of abnormal diagnosis can be improved.

[0127] Optionally, in adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model, the feedback adjustment module is specifically used for:

[0128] Obtain the first algorithm control parameters of the first anomaly diagnosis model;

[0129] Determine the first adjustment parameter corresponding to the first difference;

[0130] The first algorithm control parameters are adjusted according to the first adjustment parameters to obtain the second algorithm control parameters;

[0131] The second anomaly diagnosis model is determined based on the second algorithm control parameters and the first anomaly diagnosis model.

[0132] The first algorithm control parameter of the first anomaly diagnosis model is used to control the anomaly diagnosis effect, which may include at least one of the following: anomaly diagnosis accuracy, anomaly diagnosis speed, etc., which are not limited here.

[0133] In specific implementation, a pre-stored mapping relationship between preset differences and adjustment parameters can be used. Then, based on this mapping relationship, a first adjustment parameter corresponding to the first difference can be determined. The first algorithm control parameter is then adjusted according to the first adjustment parameter to obtain the second algorithm control parameter, i.e., the second algorithm control parameter = (1 + first adjustment parameter) * first algorithm control parameter. The second anomaly diagnosis model is then determined based on the second algorithm control parameter and the first anomaly diagnosis model, i.e., the model parameters of the first anomaly diagnosis model are adapted to the second algorithm control parameter. The first difference reflects the degree of difference between the actual probability value and the upper limit threshold. In this way, the anomaly diagnosis model can be optimized using the inference results. That is, the anomaly diagnosis model can be dynamically optimized based on the differences between historical anomaly diagnosis cases and actual anomaly diagnosis cases to ensure diagnostic accuracy.

[0134] In this embodiment, an auxiliary decision-making model library can be established for fields such as anomaly diagnosis and status assessment, maintenance strategy optimization, and spare parts management. These models can take the form of rule-based expert systems, instance-based case reasoning, and deep learning-based intelligent models. Each model can be deeply optimized to improve the accuracy and efficiency of anomaly diagnosis.

[0135] Optionally, the anomaly diagnosis module is further specifically used for:

[0136] Obtain a reference anomaly diagnostic model corresponding to the preset position;

[0137] Based on the first working data, a first fitted straight line and a second fitted curve are obtained;

[0138] Determine the absolute value of the slope of the first fitted line to obtain the first absolute value;

[0139] Determine the target mean and first standard deviation of the second fitted curve;

[0140] A first stability evaluation value is determined based on the first absolute value, the target average value, and the first standard deviation.

[0141] Determine the first algorithm control parameters corresponding to the first stability evaluation value;

[0142] The first anomaly diagnosis model is determined based on the first algorithm control parameters and the reference anomaly diagnosis model.

[0143] In this embodiment, a mapping relationship between a preset location and an anomaly diagnosis model can be stored in advance. Then, a reference anomaly diagnosis model corresponding to the preset location can be determined based on the mapping relationship. Since different locations have different anomalies, different anomaly diagnosis models can be configured for different locations.

[0144] In specific implementation, since the first working data can include multiple working data points, and each working data point can correspond to a sampling time, the multiple working data points and their corresponding sampling times can be mapped to a coordinate system. The horizontal axis of this coordinate system represents time, and the vertical axis represents the working data. Therefore, the multiple working data points and their corresponding sampling times can be mapped to multiple coordinate points. Based on these multiple coordinate points, fitting can be performed to obtain a first fitted straight line and a second fitted curve. The curve segment of the second fitted curve located in the first time period can be extracted. Next, the absolute value of the slope of the first fitted straight line can be determined to obtain the first absolute value. The first absolute value reflects the working stability of the switchgear at the preset position to a certain extent. Of course, the target average value and the first standard deviation of the second fitted curve can also be determined. That is, the average value of the curve segment is determined to obtain the target average value, and the standard deviation of the curve segment is determined to obtain the first standard deviation. The target average value reflects the working state of the switchgear at the preset position, and the first standard deviation also reflects the working stability of the switchgear at the preset position. Therefore, a first stability evaluation value can be determined based on the first absolute value, the target average value, and the first standard deviation. In this way, the working stability of the switchgear at the preset position corresponding to the actual working state can be obtained.

[0145] Next, a pre-stored mapping relationship between preset stability evaluation values ​​and algorithm control parameters can be stored. Then, based on this mapping relationship, the first algorithm control parameters corresponding to the first stability evaluation value can be determined. Then, based on the first algorithm control parameters and the reference anomaly diagnosis model, the first anomaly diagnosis model can be determined. In this way, on the one hand, the corresponding anomaly diagnosis model can be adapted based on the characteristics of the preset location, and on the other hand, the corresponding algorithm control parameters can be adapted based on the working stability of the switch cabinet at the preset location. Thus, the accuracy of anomaly diagnosis can be significantly improved to a certain extent.

[0146] Optionally, in determining the first stability evaluation value based on the first absolute value, the target average value, and the first standard deviation, the anomaly diagnosis module is specifically used for:

[0147] According to the preset mapping relationship between absolute values ​​and stability evaluation values, determine the first reference stability evaluation value corresponding to the first absolute value;

[0148] Based on the preset mapping relationship between standard deviation and stability evaluation value, determine the second reference stability evaluation value corresponding to the first standard deviation;

[0149] Obtain the target weight pair corresponding to the target average value;

[0150] The first stability evaluation value is obtained by performing a weighted operation on the target weight pair, the first reference stability evaluation value, and the second reference stability evaluation value.

[0151] In a specific implementation, a pre-defined mapping relationship between absolute values ​​and stability evaluation values ​​can be stored in advance, and then a first reference stability evaluation value corresponding to the first absolute value can be determined based on the mapping relationship. Furthermore, a pre-defined mapping relationship between standard deviations and stability evaluation values ​​can be stored in advance, and a second reference stability evaluation value corresponding to the first standard deviation can be determined based on the mapping relationship.

[0152] Next, a pre-stored mapping relationship between the average value and the weight pair can be stored. Then, based on this mapping relationship, the target weight pair corresponding to the target average value can be determined. The weight pair can include two weights, the sum of which is 1. The two weights correspond to the first reference stability evaluation value and the second reference stability evaluation value, respectively. Finally, a weighted operation can be performed on the target weight pair, the first reference stability evaluation value, and the second reference stability evaluation value to obtain the first stability evaluation value. In this way, not only can the stability of the amplitude dimension (standard deviation) and the stability of the trend direction (absolute value of the slope) be determined, but also the corresponding weight pair can be adapted based on the working stability of the preset position. The stability of the switch cabinet at the preset position can be accurately determined, which greatly improves the accuracy of anomaly diagnosis to a certain extent.

[0153] As can be seen, the knowledge graph-based intelligent monitoring system for switchgear described in this application includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The data acquisition module acquires first working data of the preset location of the intelligent switchgear in a first time period; the anomaly diagnosis module inputs the first working data into a first anomaly diagnosis model to obtain a first diagnosis result; the acquisition module acquires first attribute information of the preset location and second attribute information of the first working data; the inference module performs node positioning on the corresponding knowledge graph of the intelligent switchgear monitoring system based on the first and second attribute information to obtain a target node and acquires the first knowledge graph corresponding to the target node; it infers the first diagnosis result based on the first knowledge graph to obtain a first inference result; the feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model; the data acquisition module acquires second working data of the preset location of the intelligent switchgear in a second time period; the start time of the second time period is later than the start time of the first time period, and... The end time of the second time period is later than the end time of the second time period. The anomaly diagnosis module inputs the second working data into the second anomaly diagnosis model to obtain the second diagnosis result. Firstly, since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics, it characterizes the changes in the working status of the switchgear at the preset location. Therefore, it can quickly lock the relevant nodes based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics, and accurately obtain the relevant knowledge graph, ensuring the accuracy of knowledge graph acquisition, which helps to ensure the efficiency of subsequent reasoning. Moreover, since the relevant knowledge graph is accurately obtained, the accuracy and efficiency of the reasoning of the diagnosis result can be guaranteed. Secondly, it can not only use the reasoning result to optimize the anomaly diagnosis model and ensure the accuracy of diagnosis, but also re-collect the working data after the first time period. Considering the continuity and gradual change of the anomaly, it can prevent the randomness of anomaly detection, and can also accurately capture the anomaly when it actually occurs, further ensuring and improving the accuracy of the anomaly diagnosis of the switchgear, and also ensuring the intelligence of the anomaly diagnosis of the switchgear.

[0154] It is understood that the functions of each program module of the knowledge graph-based intelligent monitoring system for switchgear in this embodiment can be specifically implemented according to the methods in the above method embodiments. The specific implementation process can be referred to the relevant descriptions in the above method embodiments, and will not be repeated here.

[0155] Please see Figure 6 , Figure 6This is a flowchart illustrating a knowledge graph-based intelligent monitoring method for switchgear, provided in an embodiment of this application. The method is applied to a knowledge graph-based intelligent monitoring system for switchgear, which includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The knowledge graph-based intelligent monitoring method for switchgear includes:

[0156] 601. The acquisition module obtains the first working data of the preset position of the intelligent switch cabinet in the first time period.

[0157] 602. The first working data is input into the first anomaly diagnosis model through the anomaly diagnosis module to obtain the first diagnosis result.

[0158] 603. Obtain the first attribute information of the preset position through the acquisition module, and obtain the second attribute information of the first working data.

[0159] 604. The reasoning module performs node localization on the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information to obtain the target node and acquire the first knowledge graph corresponding to the target node; and performs reasoning on the first diagnostic result based on the first knowledge graph to obtain the first reasoning result.

[0160] 605. The feedback adjustment module adjusts the first anomaly diagnosis model according to the first inference result to obtain a second anomaly diagnosis model.

[0161] 606. The acquisition module obtains the second working data of the preset position of the intelligent switch cabinet in the second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the second time period.

[0162] 607. The second working data is input into the second anomaly diagnosis model through the anomaly diagnosis module to obtain the second diagnosis result.

[0163] The specific descriptions of steps 601-607 can be found in the corresponding descriptions of the knowledge graph-based intelligent monitoring system for switchgear, and will not be repeated here.

[0164] Optionally, the above steps, which involve locating the target node in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, can be implemented in the following manner:

[0165] Obtain the corresponding knowledge graph library of the intelligent monitoring system for the switchgear. The knowledge graph library includes at least one knowledge graph, each knowledge graph includes at least one node, and each node corresponds to a node tag set and a node index; the node tag set includes multiple node tags.

[0166] The first node tag set is determined based on the first attribute information;

[0167] The second node tag set is determined based on the second attribute information;

[0168] The first node tag set is matched with the node tag set of the knowledge graph in the at least one knowledge graph to obtain the target node tag set that is successfully matched.

[0169] Obtain the target node index corresponding to the target node tag set, and obtain the target node based on the target node index.

[0170] Optionally, regarding the step above of obtaining the first knowledge graph corresponding to the target node, the reasoning module can be implemented in the following manner:

[0171] Obtain the reference knowledge graph corresponding to the target node;

[0172] Determine the first keyword corresponding to the first attribute information;

[0173] Determine the second keyword corresponding to the second attribute information;

[0174] The reference knowledge graph is searched based on the first keyword to obtain the first part of the knowledge graph;

[0175] The second part of the knowledge graph is obtained by searching the reference knowledge graph based on the second keyword;

[0176] The first knowledge graph is determined based on the first part of the knowledge graph and the second part of the knowledge graph.

[0177] Optionally, the first diagnostic result includes: a first abnormality type and a first probability value corresponding to the first abnormality type. The above steps, which involve reasoning about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result, can be implemented in the following manner:

[0178] The first knowledge graph is used to obtain a set of historical anomaly diagnosis results corresponding to the first attribute information and the second attribute information. The set of historical anomaly diagnosis results includes multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a diagnosis result authenticity identifier. The diagnosis result authenticity identifier includes a diagnosis result true identifier or a diagnosis result false identifier.

[0179] Determine the number of true diagnostic results among the multiple historical diagnostic results to obtain a first number;

[0180] Determine the number of false diagnostic results among the multiple historical diagnostic results to obtain a second number;

[0181] The first average value is obtained by performing a calculation based on the second probability value corresponding to the first quantity.

[0182] The second average value is obtained by performing calculations based on the second probability value corresponding to the second quantity.

[0183] The first inference result is obtained by reasoning based on the first average value and the second average value.

[0184] Optionally, the above step of averaging the second probability value corresponding to the first quantity to obtain the first average value can be implemented in the following manner:

[0185] The first reference average value is obtained by averaging the second probability value corresponding to the first quantity.

[0186] A first proportion value for identifying true diagnostic results is determined based on the first quantity and the second quantity;

[0187] A first optimization factor corresponding to the first ratio value is determined, and the value range of the first optimization factor is 0 to 0.1;

[0188] The first reference average value is obtained by optimizing the first optimization factor, specifically: First average value = (1 + first optimization factor) * first reference average value.

[0189] Optionally, the above steps, determining the second probability value corresponding to the second quantity to obtain the second average value, can be implemented as follows:

[0190] The second reference average value is obtained by averaging the second probability value corresponding to the second quantity.

[0191] The second reference average value is obtained by optimizing the first optimization factor, specifically: first average value = (1 - first optimization factor) * second reference average value.

[0192] Optionally, the above steps, inferring the first diagnostic result based on the first average value and the second average value to obtain the first inference result, can be implemented in the following manner:

[0193] When the first probability value is greater than or equal to the second average value and less than or equal to the first average value, a first difference between the first probability value and the first average value is determined; the first inference result is determined based on the first difference.

[0194] When the first probability value is less than the second average value, the diagnostic result authenticity identifier of the first diagnostic result is determined to be a diagnostic result false identifier, and the first inference result is determined based on the diagnostic result false identifier;

[0195] When the first probability value is greater than the first average value, the diagnostic result authenticity identifier of the first diagnostic result is determined as the diagnostic result is true identifier, and the first inference result is determined based on the diagnostic result is true identifier.

[0196] Optionally, the following steps may also be included:

[0197] When the first probability value is less than the second average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed through the feedback adjustment module.

[0198] or,

[0199] When the first probability value is greater than the first average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed by the feedback adjustment module.

[0200] Optionally, the above steps, adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model, can be implemented in the following manner:

[0201] Obtain the first algorithm control parameters of the first anomaly diagnosis model;

[0202] Determine the first adjustment parameter corresponding to the first difference;

[0203] The first algorithm control parameters are adjusted according to the first adjustment parameters to obtain the second algorithm control parameters;

[0204] The second anomaly diagnosis model is determined based on the second algorithm control parameters and the first anomaly diagnosis model.

[0205] Optionally, the following steps may also be included:

[0206] The anomaly diagnosis module obtains a reference anomaly diagnosis model corresponding to the preset location.

[0207] Based on the first working data, a first fitted straight line and a second fitted curve are obtained;

[0208] Determine the absolute value of the slope of the first fitted line to obtain the first absolute value;

[0209] Determine the target mean and first standard deviation of the second fitted curve;

[0210] A first stability evaluation value is determined based on the first absolute value, the target average value, and the first standard deviation.

[0211] Determine the first algorithm control parameters corresponding to the first stability evaluation value;

[0212] The first anomaly diagnosis model is determined based on the first algorithm control parameters and the reference anomaly diagnosis model.

[0213] The specific descriptions of the steps in the above-mentioned knowledge graph-based intelligent monitoring method for switchgear can be found in the corresponding descriptions in the above-mentioned knowledge graph-based intelligent monitoring system for switchgear, and will not be repeated here.

[0214] As can be seen, the knowledge graph-based intelligent monitoring method for switchgear described in this application embodiment is applied to a knowledge graph-based intelligent monitoring system for switchgear. This system includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The data acquisition module acquires first working data of the preset location of the intelligent switchgear in a first time period. The anomaly diagnosis module inputs the first working data into a first anomaly diagnosis model to obtain a first diagnosis result. The acquisition module acquires first attribute information of the preset location and second attribute information of the first working data. The inference module locates nodes in the corresponding knowledge graph of the intelligent switchgear monitoring system based on the first and second attribute information to obtain target nodes and acquires the first knowledge graph corresponding to the target nodes. The first diagnosis result is inferred based on the first knowledge graph to obtain a first inference result. The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model. The data acquisition module acquires second working data of the preset location of the intelligent switchgear in a second time period. The start time of the second time period is later than... The first time period begins at the start time, and the end time of the second time period ends later than the end time of the second time period. The anomaly diagnosis module inputs the second working data into the second anomaly diagnosis model to obtain the second diagnosis result. Firstly, since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics, it characterizes the changes in the working status of the switchgear at the preset location. Therefore, based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics, relevant nodes can be quickly located, and relevant knowledge graphs can be accurately obtained, ensuring the accuracy of knowledge graph acquisition and helping to ensure the efficiency of subsequent reasoning. Furthermore, due to the accurate acquisition of relevant knowledge graphs, the accuracy and efficiency of the reasoning of the diagnosis results can be guaranteed. Secondly, not only can the anomaly diagnosis model be optimized using the reasoning results to ensure the accuracy of diagnosis, but also the working data after the first time period can be re-collected. Considering the continuity and gradual change of anomalies, the randomness of anomaly detection can be prevented, and anomalies can be accurately captured when they actually occur, further ensuring and improving the accuracy of anomaly diagnosis of the switchgear, and also ensuring the intelligence of anomaly diagnosis of the switchgear.

[0215] Consistent with the above embodiments, please refer to Figure 7 , Figure 7This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device includes a processor, a memory, a communication interface, and one or more programs. The one or more programs are stored in the memory and configured to be executed by the processor. In this embodiment, the electronic device includes a knowledge graph-based intelligent monitoring system for switchgear. The knowledge graph-based intelligent monitoring system for switchgear includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The program includes instructions for performing the following steps:

[0216] The acquisition module obtains the first working data of the preset location of the intelligent switch cabinet in the first time period.

[0217] The first working data is input into the first anomaly diagnosis model through the anomaly diagnosis module to obtain the first diagnosis result;

[0218] The acquisition module obtains the first attribute information of the preset location and the second attribute information of the first working data.

[0219] The reasoning module locates the target node in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, and obtains the first knowledge graph corresponding to the target node; it then infers the first diagnostic result based on the first knowledge graph to obtain the first reasoning result.

[0220] The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model.

[0221] The acquisition module obtains the second working data of the preset position of the intelligent switch cabinet in the second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the second time period.

[0222] The second working data is input into the second anomaly diagnosis model through the anomaly diagnosis module to obtain the second diagnosis result.

[0223] Optionally, in the step of locating the target node by analyzing the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, the above program includes instructions for performing the following steps:

[0224] Obtain the corresponding knowledge graph library of the intelligent monitoring system for the switchgear. The knowledge graph library includes at least one knowledge graph, each knowledge graph includes at least one node, and each node corresponds to a node tag set and a node index; the node tag set includes multiple node tags.

[0225] The first node tag set is determined based on the first attribute information;

[0226] The second node tag set is determined based on the second attribute information;

[0227] The first node tag set is matched with the node tag set of the knowledge graph in the at least one knowledge graph to obtain the target node tag set that is successfully matched.

[0228] Obtain the target node index corresponding to the target node tag set, and obtain the target node based on the target node index.

[0229] Optionally, in obtaining the first knowledge graph corresponding to the target node, the above procedure includes instructions for performing the following steps:

[0230] Obtain the reference knowledge graph corresponding to the target node;

[0231] Determine the first keyword corresponding to the first attribute information;

[0232] Determine the second keyword corresponding to the second attribute information;

[0233] The reference knowledge graph is searched based on the first keyword to obtain the first part of the knowledge graph;

[0234] The second part of the knowledge graph is obtained by searching the reference knowledge graph based on the second keyword;

[0235] The first knowledge graph is determined based on the first part of the knowledge graph and the second part of the knowledge graph.

[0236] Optionally, the first diagnostic result includes: a first abnormality type and a first probability value corresponding to the first abnormality type; in terms of reasoning about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result, the above procedure includes instructions for performing the following steps:

[0237] The first knowledge graph is used to obtain a set of historical anomaly diagnosis results corresponding to the first attribute information and the second attribute information. The set of historical anomaly diagnosis results includes multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a diagnosis result authenticity identifier. The diagnosis result authenticity identifier includes a diagnosis result true identifier or a diagnosis result false identifier.

[0238] Determine the number of true diagnostic results among the multiple historical diagnostic results to obtain a first number;

[0239] Determine the number of false diagnostic results among the multiple historical diagnostic results to obtain a second number;

[0240] The first average value is obtained by performing a calculation based on the second probability value corresponding to the first quantity.

[0241] The second average value is obtained by performing a calculation based on the second probability value corresponding to the second quantity.

[0242] The first inference result is obtained by reasoning based on the first average value and the second average value.

[0243] Optionally, in the step of averaging the second probability value corresponding to the first quantity to obtain the first average, the above procedure includes instructions for performing the following steps:

[0244] The first reference average value is obtained by averaging the second probability value corresponding to the first quantity.

[0245] A first proportion value for identifying true diagnostic results is determined based on the first quantity and the second quantity;

[0246] A first optimization factor corresponding to the first ratio value is determined, and the value range of the first optimization factor is 0 to 0.1;

[0247] The first reference average value is obtained by optimizing the first optimization factor, specifically: First average value = (1 + first optimization factor) * first reference average value.

[0248] Optionally, in the operation of determining the second probability value corresponding to the second quantity to obtain the second average value, the above procedure includes instructions for performing the following steps:

[0249] The second reference average value is obtained by averaging the second probability value corresponding to the second quantity.

[0250] The second reference average value is obtained by optimizing the first optimization factor, specifically: first average value = (1 - first optimization factor) * second reference average value.

[0251] Optionally, in the step of reasoning about the first diagnostic result based on the first average value and the second average value to obtain the first reasoning result, the above procedure includes instructions for performing the following steps:

[0252] When the first probability value is greater than or equal to the second average value and less than or equal to the first average value, a first difference between the first probability value and the first average value is determined; the first inference result is determined based on the first difference.

[0253] When the first probability value is less than the second average value, the diagnostic result authenticity identifier of the first diagnostic result is determined to be a diagnostic result false identifier, and the first inference result is determined based on the diagnostic result false identifier;

[0254] When the first probability value is greater than the first average value, the diagnostic result authenticity identifier of the first diagnostic result is determined as the diagnostic result is true identifier, and the first inference result is determined based on the diagnostic result is true identifier.

[0255] Optionally, the above procedure may also include instructions for performing the following steps:

[0256] When the first probability value is less than the second average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed through the feedback adjustment module.

[0257] or,

[0258] When the first probability value is greater than the first average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed by the feedback adjustment module.

[0259] Optionally, in adjusting the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model, the above procedure includes instructions for performing the following steps:

[0260] Obtain the first algorithm control parameters of the first anomaly diagnosis model;

[0261] Determine the first adjustment parameter corresponding to the first difference;

[0262] The first algorithm control parameters are adjusted according to the first adjustment parameters to obtain the second algorithm control parameters;

[0263] The second anomaly diagnosis model is determined based on the second algorithm control parameters and the first anomaly diagnosis model.

[0264] Optionally, the above procedure may also include instructions for performing the following steps:

[0265] The anomaly diagnosis module obtains a reference anomaly diagnosis model corresponding to the preset location.

[0266] Based on the first working data, a first fitted straight line and a second fitted curve are obtained;

[0267] Determine the absolute value of the slope of the first fitted line to obtain the first absolute value;

[0268] Determine the target mean and first standard deviation of the second fitted curve;

[0269] A first stability evaluation value is determined based on the first absolute value, the target average value, and the first standard deviation.

[0270] Determine the first algorithm control parameters corresponding to the first stability evaluation value;

[0271] The first anomaly diagnosis model is determined based on the first algorithm control parameters and the reference anomaly diagnosis model.

[0272] As can be seen, the electronic device described in this application embodiment includes a knowledge graph-based intelligent monitoring system for switchgear. This system comprises: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The data acquisition module acquires first working data of a preset location of the intelligent switchgear within a first time period. The anomaly diagnosis module inputs the first working data into a first anomaly diagnosis model to obtain a first diagnosis result. The acquisition module acquires first attribute information of the preset location and second attribute information of the first working data. The inference module locates nodes in the corresponding knowledge graph of the intelligent switchgear monitoring system based on the first and second attribute information to obtain target nodes and acquires the first knowledge graph corresponding to the target nodes. It then infers the first diagnosis result based on the first knowledge graph to obtain a first inference result. The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model. The data acquisition module acquires second working data of the preset location of the intelligent switchgear within a second time period. The start time of the second time period is later than that of the first time period. The start time is later than the end time of the second time period; the anomaly diagnosis module inputs the second working data into the second anomaly diagnosis model to obtain the second diagnosis result. Firstly, since the first attribute information can reflect the physical characteristics of the preset location, or the actual environmental characteristics of the preset location, and the second attribute information reflects the data characteristics, characterizing the changes in the working status of the switchgear at the preset location, the relevant nodes can be quickly located based on the relevant physical characteristics, environmental characteristics, and relevant data characteristics, and the relevant knowledge graph can be accurately obtained, ensuring the accuracy of knowledge graph acquisition, which helps to ensure the efficiency of subsequent reasoning. Moreover, since the relevant knowledge graph is accurately obtained, the accuracy and efficiency of the reasoning of the diagnosis result can be guaranteed. Secondly, not only can the anomaly diagnosis model be optimized using the reasoning result to ensure the accuracy of diagnosis, but also the working data after the first time period can be re-collected. Considering the continuity and gradual change of the anomaly, the randomness of anomaly detection can be prevented, and the anomaly can be accurately captured when it actually occurs, further ensuring and improving the accuracy of anomaly diagnosis of the switchgear, and also ensuring the intelligence of anomaly diagnosis of the switchgear.

[0273] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0274] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may include an electronic device.

[0275] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0276] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0277] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or modules may be electrical or other forms.

[0278] The modules described above as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0279] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0280] If the integrated modules described above are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0281] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0282] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A knowledge graph-based intelligent monitoring system for switchgear, characterized in that, The knowledge graph-based intelligent monitoring system for switchgear includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, an inference module, and a feedback adjustment module. The acquisition module is used to acquire the first working data of the preset position of the intelligent switch cabinet in the first time period. The anomaly diagnosis module is used to input the first working data into the first anomaly diagnosis model to obtain the first diagnosis result; The acquisition module is used to acquire the first attribute information of the preset position and the second attribute information of the first working data; The reasoning module is used to locate nodes in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, obtain target nodes, and acquire the first knowledge graph corresponding to the target nodes; and to reason about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result. The feedback adjustment module is used to adjust the first anomaly diagnosis model according to the first inference result to obtain a second anomaly diagnosis model. The acquisition module is further configured to acquire second working data of the preset position of the intelligent switch cabinet in a second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the first time period. The anomaly diagnosis module is also used to input the second working data into the second anomaly diagnosis model to obtain a second diagnosis result; Specifically, regarding the acquisition of the first knowledge graph corresponding to the target node, the reasoning module is used for: Obtain the reference knowledge graph corresponding to the target node; Determine the first keyword corresponding to the first attribute information; Determine the second keyword corresponding to the second attribute information; The reference knowledge graph is searched based on the first keyword to obtain the first part of the knowledge graph; The second part of the knowledge graph is obtained by searching the reference knowledge graph based on the second keyword; The first knowledge graph is determined based on the first part of the knowledge graph and the second part of the knowledge graph; Wherein, the first diagnostic result includes: a first abnormality type and a first probability value corresponding to the first abnormality type; in terms of reasoning about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result, the reasoning module is specifically used for: The first knowledge graph is used to obtain a set of historical anomaly diagnosis results corresponding to the first attribute information and the second attribute information. The set of historical anomaly diagnosis results includes multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a diagnosis result authenticity identifier. The diagnosis result authenticity identifier includes a diagnosis result true identifier or a diagnosis result false identifier. Determine the number of true diagnostic results among the multiple historical diagnostic results to obtain a first number; Determine the number of false diagnostic results among the multiple historical diagnostic results to obtain a second number; The first average value is obtained by performing a calculation based on the second probability value corresponding to the first quantity. The second average value is obtained by performing a calculation based on the second probability value corresponding to the second quantity. The first inference result is obtained by reasoning based on the first average value and the second average value.

2. The system according to claim 1, characterized in that, In the process of locating the target node based on the first attribute information and the second attribute information in the knowledge graph of the intelligent monitoring system for the switchgear, the reasoning module is specifically used for: Obtain the corresponding knowledge graph library of the intelligent monitoring system for the switchgear. The knowledge graph library includes at least one knowledge graph, each knowledge graph includes at least one node, and each node corresponds to a node tag set and a node index; the node tag set includes multiple node tags. The first node tag set is determined based on the first attribute information; The second node tag set is determined based on the second attribute information; The first node tag set is matched with the node tag set of the knowledge graph in the at least one knowledge graph to obtain the target node tag set that is successfully matched. Obtain the target node index corresponding to the target node tag set, and obtain the target node based on the target node index.

3. The system according to claim 1, characterized in that, In the process of averaging the second probability value corresponding to the first quantity to obtain the first average value, the reasoning module is specifically used for: The first reference average value is obtained by averaging the second probability value corresponding to the first quantity. A first proportion value for identifying true diagnostic results is determined based on the first quantity and the second quantity; Determine a first optimization factor corresponding to the first ratio value, wherein the value of the first optimization factor ranges from 0 to 0.1; The first reference average value is obtained by optimizing the first optimization factor, specifically: First average value = (1 + first optimization factor) * first reference average value.

4. The system according to claim 3, characterized in that, In the operation of determining the second probability value corresponding to the second quantity and obtaining the second average value, the reasoning module is specifically used for: The second reference average value is obtained by averaging the second probability value corresponding to the second quantity. The second reference average value is obtained by optimizing the second reference average value based on the first optimization factor, specifically: first average value = (1 - first optimization factor) * second reference average value.

5. The system according to any one of claims 1-4, characterized in that, In the aspect of reasoning about the first diagnostic result based on the first average value and the second average value to obtain the first reasoning result, the reasoning module is specifically used for: When the first probability value is greater than or equal to the second average value and less than or equal to the first average value, a first difference between the first probability value and the first average value is determined; the first inference result is determined based on the first difference. When the first probability value is less than the second average value, the diagnostic result authenticity identifier of the first diagnostic result is determined to be a diagnostic result false identifier, and the first inference result is determined based on the diagnostic result false identifier; When the first probability value is greater than the first average value, the diagnostic result authenticity identifier of the first diagnostic result is determined as the diagnostic result is true identifier, and the first inference result is determined based on the diagnostic result is true identifier.

6. The system according to claim 5, characterized in that, When the first probability value is less than the second average value, the feedback adjustment module does not perform the step of adjusting the first anomaly diagnosis model according to the first inference result to obtain the second anomaly diagnosis model. or, When the first probability value is greater than the first average value, the feedback adjustment module does not perform the step of adjusting the first anomaly diagnosis model according to the first inference result to obtain the second anomaly diagnosis model.

7. The system according to claim 5, characterized in that, In adjusting the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model, the feedback adjustment module is specifically used for: Obtain the first algorithm control parameters of the first anomaly diagnosis model; Determine the first adjustment parameter corresponding to the first difference; The first algorithm control parameters are adjusted according to the first adjustment parameters to obtain the second algorithm control parameters; The second anomaly diagnosis model is determined based on the second algorithm control parameters and the first anomaly diagnosis model.

8. The system according to claim 7, characterized in that, The anomaly diagnosis module is also specifically used for: Obtain a reference anomaly diagnostic model corresponding to the preset position; Based on the first working data, a first fitted straight line and a second fitted curve are obtained; Determine the absolute value of the slope of the first fitted line to obtain the first absolute value; Determine the target mean and first standard deviation of the second fitted curve; A first stability evaluation value is determined based on the first absolute value, the target average value, and the first standard deviation. Determine the first algorithm control parameters corresponding to the first stability evaluation value; The first anomaly diagnosis model is determined based on the first algorithm control parameters and the reference anomaly diagnosis model.

9. A knowledge graph-based intelligent monitoring method for switchgear, characterized in that, This method is applied to a knowledge graph-based intelligent monitoring system for switchgear. The knowledge graph-based intelligent monitoring system for switchgear includes: a data acquisition module, an anomaly diagnosis module, an acquisition module, a reasoning module, and a feedback adjustment module. The method includes: The acquisition module obtains the first working data of the preset location of the intelligent switch cabinet in the first time period. The first working data is input into the first anomaly diagnosis model through the anomaly diagnosis module to obtain the first diagnosis result; The acquisition module obtains the first attribute information of the preset location and the second attribute information of the first working data. The reasoning module locates the target node in the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information, and obtains the first knowledge graph corresponding to the target node; it then infers the first diagnostic result based on the first knowledge graph to obtain the first reasoning result. The feedback adjustment module adjusts the first anomaly diagnosis model based on the first inference result to obtain a second anomaly diagnosis model. The acquisition module obtains second working data of the preset position of the intelligent switch cabinet in a second time period; the start time of the second time period is later than the start time of the first time period, and the end time of the second time period is later than the end time of the first time period. The second working data is input into the second anomaly diagnosis model through the anomaly diagnosis module to obtain the second diagnosis result; Wherein, obtaining the first knowledge graph corresponding to the target node includes: Obtain the reference knowledge graph corresponding to the target node; Determine the first keyword corresponding to the first attribute information; Determine the second keyword corresponding to the second attribute information; The reference knowledge graph is searched based on the first keyword to obtain the first part of the knowledge graph; The second part of the knowledge graph is obtained by searching the reference knowledge graph based on the second keyword; The first knowledge graph is determined based on the first part of the knowledge graph and the second part of the knowledge graph; The first diagnostic result includes: a first abnormality type and a first probability value corresponding to the first abnormality type; the step of reasoning about the first diagnostic result based on the first knowledge graph to obtain a first reasoning result includes: The first knowledge graph is used to obtain a set of historical anomaly diagnosis results corresponding to the first attribute information and the second attribute information. The set of historical anomaly diagnosis results includes multiple historical diagnosis results. Each historical diagnosis result includes a first anomaly type, a second probability value, and a diagnosis result authenticity identifier. The diagnosis result authenticity identifier includes a diagnosis result true identifier or a diagnosis result false identifier. Determine the number of true diagnostic results among the multiple historical diagnostic results to obtain a first number; Determine the number of false diagnostic results among the multiple historical diagnostic results to obtain a second number; The first average value is obtained by performing a calculation based on the second probability value corresponding to the first quantity. The second average value is obtained by performing a calculation based on the second probability value corresponding to the second quantity. The first inference result is obtained by reasoning based on the first average value and the second average value.

10. The method according to claim 9, characterized in that, The step of locating the target node by analyzing the knowledge graph of the intelligent monitoring system for the switchgear based on the first attribute information and the second attribute information includes: Obtain the corresponding knowledge graph library of the intelligent monitoring system for the switchgear. The knowledge graph library includes at least one knowledge graph, each knowledge graph includes at least one node, and each node corresponds to a node tag set and a node index; the node tag set includes multiple node tags. The first node tag set is determined based on the first attribute information; The second node tag set is determined based on the second attribute information; The first node tag set is matched with the node tag set of the knowledge graph in the at least one knowledge graph to obtain the target node tag set that is successfully matched. Obtain the target node index corresponding to the target node tag set, and obtain the target node based on the target node index.

11. The method according to claim 10, characterized in that, The step of calculating the average value based on the second probability value corresponding to the first quantity to obtain the first average value includes: The first reference average value is obtained by averaging the second probability value corresponding to the first quantity. A first proportion value for identifying true diagnostic results is determined based on the first quantity and the second quantity; Determine a first optimization factor corresponding to the first ratio value, wherein the value of the first optimization factor ranges from 0 to 0.1; The first reference average value is obtained by optimizing the first optimization factor, specifically: First average value = (1 + first optimization factor) * first reference average value.

12. The method according to claim 11, characterized in that, The operation of determining the second probability value corresponding to the second quantity to obtain the second average value includes: The second reference average value is obtained by averaging the second probability value corresponding to the second quantity. The second reference average value is obtained by optimizing the second reference average value based on the first optimization factor, specifically: first average value = (1 - first optimization factor) * second reference average value.

13. The method according to any one of claims 9-12, characterized in that, The step of reasoning about the first diagnostic result based on the first average value and the second average value to obtain the first reasoning result includes: When the first probability value is greater than or equal to the second average value and less than or equal to the first average value, a first difference between the first probability value and the first average value is determined; the first inference result is determined based on the first difference. When the first probability value is less than the second average value, the diagnostic result authenticity identifier of the first diagnostic result is determined to be a diagnostic result false identifier, and the first inference result is determined based on the diagnostic result false identifier; When the first probability value is greater than the first average value, the diagnostic result authenticity identifier of the first diagnostic result is determined as the diagnostic result is true identifier, and the first inference result is determined based on the diagnostic result is true identifier.

14. The method according to claim 13, characterized in that, The method further includes: When the first probability value is less than the second average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed through the feedback adjustment module. or, When the first probability value is greater than the first average value, the step of adjusting the first anomaly diagnosis model based on the first inference result to obtain the second anomaly diagnosis model is not performed by the feedback adjustment module.

15. The method according to claim 13, characterized in that, The step of adjusting the first anomaly diagnosis model based on the first reasoning result to obtain a second anomaly diagnosis model includes: Obtain the first algorithm control parameters of the first anomaly diagnosis model; Determine the first adjustment parameter corresponding to the first difference; The first algorithm control parameters are adjusted according to the first adjustment parameters to obtain the second algorithm control parameters; The second anomaly diagnosis model is determined based on the second algorithm control parameters and the first anomaly diagnosis model.

16. The method according to claim 15, characterized in that, The method further includes: The anomaly diagnosis module obtains a reference anomaly diagnosis model corresponding to the preset location. Based on the first working data, a first fitted straight line and a second fitted curve are obtained; Determine the absolute value of the slope of the first fitted line to obtain the first absolute value; Determine the target mean and first standard deviation of the second fitted curve; A first stability evaluation value is determined based on the first absolute value, the target average value, and the first standard deviation. Determine the first algorithm control parameters corresponding to the first stability evaluation value; The first anomaly diagnosis model is determined based on the first algorithm control parameters and the reference anomaly diagnosis model.

Images