State detection method and device of power equipment, electronic equipment and storage medium

Abstract

The invention discloses a state detection method and device of power equipment, electronic equipment and a storage medium. Comprising the steps that multi-dimensional data, including equipment data, detection data and external data, of power equipment is acquired, and the external data comprises one or more of video data and image data; performing feature extraction on the external data to obtain feature data, and associating the feature data, the equipment data and the detection data to obtain multi-dimensional feature data; determining associated feature data based on a pre-constructed knowledge graph; and performing state detection based on the multi-dimensional feature data and the associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment. According to the scheme, feature extraction of the multi-dimensional data of the power system is realized in combination with the knowledge graph, and the extracted multi-dimensional feature data and the associated feature data are processed through the state detection model to obtain the state detection result, so that the accuracy and the efficiency of the state detection result are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and the Internet of Things, and in particular to a state detection method, device, electronic equipment and storage medium for electric power equipment. Background Art

[0002] In the power system, for key facilities such as substations, the status of the equipment is directly related to the safe and stable operation of the power grid. Traditional equipment status detection mainly relies on manual regular inspections and passive response mechanisms after failures. This method is not only inefficient, but also difficult to detect potential problems in a timely manner, which can easily lead to the expansion of failures.

[0003] With the development of Internet of Things technology, more and more intelligent individual devices have been applied to the daily maintenance of substations to detect the status of equipment. However, there are problems such as the inability to conduct effective data mining on the collected data and the inability to accurately and efficiently detect the status of equipment. Summary of the invention

[0004] The present invention provides a method and device for detecting the state of electric equipment, an electronic device and a storage medium, so as to solve the problems of inaccurate and low efficiency in detecting the state of electric equipment.

[0005] According to one aspect of the present invention, there is provided a method for detecting a state of an electric device, comprising:

[0006] Acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data;

[0007] Extract features from external data to obtain feature data, associate the feature data, equipment data and detection data to obtain multi-dimensional feature data of the power equipment;

[0008] Determine the associated characteristic data corresponding to the electric power equipment based on a pre-built knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and association relationships between the plurality of knowledge entities;

[0009] By using at least one state detection model that has been pre-trained, state detection is performed based on multi-dimensional feature data and associated feature data to obtain at least one state detection result of the electric equipment.

[0010] Optionally, feature extraction is performed on external data to obtain feature data, including: taking key video frames in image data or video data as images to be processed; performing target detection on the images to be processed to obtain feature points of the images to be processed, and determining descriptive features of the images to be processed based on the feature points; performing clustering processing on the images to be processed based on the descriptive features of the images to be processed to determine the image class to which the images to be processed belong, and determining feature data based on the image class to which the images to be processed belong, wherein the feature data includes one or more of device appearance features and operational behavior features.

[0011] Optionally, determining feature data based on the image class to which the image to be processed belongs includes: determining the feature data corresponding to the image class to which the image to be processed belongs as the feature data corresponding to the image to be processed; or; classifying the descriptive features of the image to be processed and / or the clustering centers of the image class through a classification model to obtain the feature data.

[0012] Optionally, the feature data, equipment data and detection data are associated to obtain multidimensional feature data of the power equipment, including: extracting identification information of the power equipment from text data in the multidimensional data; and associating the feature data, equipment data and detection data based on the identification information of the power equipment.

[0013] Optionally, a method for constructing a knowledge graph includes: obtaining historical multidimensional data, wherein the historical multidimensional data includes source data from multiple data sources; extracting knowledge entities from the source data of each data source, extracting attribute information of each knowledge entity and the associations between the knowledge entities, and constructing a local knowledge graph corresponding to each data source based on the attribute information and associations of each knowledge entity; performing consistency processing on the knowledge entities of the local knowledge graphs of multiple data sources, and merging the local knowledge graphs of multiple data sources to obtain a knowledge graph.

[0014] Optionally, based on a pre-constructed knowledge graph, the associated characteristic data corresponding to the power equipment is determined, including: matching the identification information or characteristic data of the power equipment in the knowledge graph to determine the successfully matched knowledge entity; based on the association relationship between the knowledge entities in the knowledge graph, determining the associated knowledge entity corresponding to the successfully matched knowledge entity, and determining the associated characteristic data based on the attribute information of the associated knowledge entity.

[0015] Optionally, the status detection result includes one or more of the following: the health status and failure mode of the power equipment at the current moment; the predicted health status and predicted failure mode of the power equipment in a future time period; and the remaining life prediction result of the power equipment.

[0016] According to another aspect of the present invention, there is provided a state detection device for electric power equipment, comprising:

[0017] A multi-dimensional data acquisition module, used to acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data;

[0018] A multi-dimensional characteristic data determination module is used to extract characteristics of external data to obtain characteristic data, and associate the characteristic data, equipment data and detection data to obtain multi-dimensional characteristic data of the power equipment;

[0019] An associated characteristic data determination module is used to determine the associated characteristic data corresponding to the power equipment based on a pre-built knowledge graph, wherein the knowledge graph includes multiple knowledge entities, attribute information of the knowledge entities, and association relationships between the multiple knowledge entities;

[0020] The state detection result determination module is used to perform state detection based on multidimensional feature data and associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment.

[0021] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0022] at least one processor; and

[0023] a memory communicatively connected to at least one processor; wherein,

[0024] The memory stores a computer program that can be executed by at least one processor. The computer program is executed by at least one processor so that the at least one processor can execute the state detection method of the power equipment according to any embodiment of the present invention.

[0025] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the state detection method of the power equipment according to any embodiment of the present invention when executed.

[0026] The technical solution of the embodiment of the present invention is to obtain multidimensional data of power equipment, wherein the multidimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; extract features from the external data to obtain feature data, associate the feature data, equipment data and detection data to obtain multidimensional feature data of the power equipment; determine the associated feature data corresponding to the power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes multiple knowledge entities, attribute information of knowledge entities and the association relationship between multiple knowledge entities; perform state detection based on the multidimensional feature data and the associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment. This solution realizes deep mining of the multidimensional data of the power equipment by extracting features from the multidimensional data and determining the associated feature data in combination with the knowledge graph, and uses the prediction model to process the multidimensional feature data and the associated feature data to obtain the state detection result, solves the problem of inaccurate and low efficiency of state detection of the power equipment, helps to warn potential risk points in advance, and provides decision support for preventive maintenance of the power equipment, while reducing the probability of errors or accidents caused by human factors, thereby improving the operating efficiency and safety of the power equipment.

[0027] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 is a flow chart of a method for detecting the state of an electric power device provided in Embodiment 1 of the present invention;

[0030] Figure 2 is a flow chart of a method for detecting the state of an electric power device provided in Embodiment 2 of the present invention;

[0031] Figure 3 It is a structural schematic diagram of a state detection device for electric power equipment provided in Embodiment 3 of the present invention;

[0032] Figure 4 It is a structural schematic diagram of an electronic device for implementing the state detection method of the electric power equipment according to the embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0034] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] Embodiment 1

[0036] Figure 1 This is a flow chart of a method for detecting the status of an electric device provided in the first embodiment of the present invention. This embodiment is applicable to the case of detecting the status of an electric device. The method can be executed by a status detection device for an electric device. The status detection device for an electric device can be implemented in the form of hardware and / or software. The status detection device for an electric device can be configured in electronic devices such as smart terminals and individual devices used in electric power systems. Figure 1 As shown, the method includes:

[0037] S110. Acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data, and external data, and the external data includes one or more of video data and image data.

[0038] Among them, power equipment refers to equipment that performs power transmission and distribution, voltage conversion, protection and control in the power system, including but not limited to generators, motors, switchgear, line cables and transformers. The multidimensional data of power equipment refers to data associated with power equipment, including but not limited to equipment data, detection data and external data. The equipment data specifically represents the equipment ledger data, including but not limited to the basic information, specification parameters, installation location, equipment manual and equipment internal structure and connection method of the power equipment. After each power equipment is installed, the corresponding information will be stored in the equipment ledger data. In this embodiment, it also includes inspection data. The equipment data also includes inspection data. The inspection data includes but is not limited to equipment inspection records, historical defect records and maintenance records; detection data refers to various data about the operating status of the equipment collected by various sensors and detection equipment; external data is specifically data collected by image acquisition equipment, and external data includes one or more of video data and image data. In this embodiment, image or video data can be collected by individual equipment, wherein individual equipment refers to equipment that can be used to collect image data and video data, and can be an intelligent terminal device with data such as image data, video data and audio data.

[0039] Specifically, external data can be collected through the camera of individual equipment. The external data includes but is not limited to video data and image data. Equipment data and detection data are obtained by reading data from a server or importing it offline, thereby obtaining multi-dimensional data such as equipment data, detection data and external data corresponding to the power equipment.

[0040] S120, extract features from external data to obtain feature data, associate the feature data, equipment data, and detection data to obtain multi-dimensional feature data of the power equipment.

[0041] Among them, the feature data can be specifically understood as the feature extracted based on the external data, which can be the data obtained by extracting features from one or more of the video data and image data corresponding to the external data. The feature data corresponding to the external data can be obtained by extracting features from the external data through the target detection algorithm, and the target detection algorithm includes but is not limited to YOLO, SSD, and Faster R-CNN. Exemplarily, the feature data can be the equipment open state data, the equipment closed state data, the equipment corner point information, the equipment edge and texture information, which can be determined according to the actual extracted information, and is not limited here. It can be understood that for the same power equipment, there is a certain correlation between the feature data, equipment data and detection data corresponding to the power equipment, for example, the identification information of the power equipment corresponding to the feature data, equipment data and detection data is the same, and the identification information can be used as an association feature to associate the feature data, equipment data and detection data corresponding to the power equipment. The identification information can be the unique coding information that characterizes the power equipment. The association of the feature data, equipment data and detection data can be specifically understood as integrating the feature data, equipment data and detection data corresponding to the power equipment through the unique identification information to obtain the multi-dimensional feature data of the power equipment.

[0042] Specifically, the target detection algorithm is called to perform feature extraction processing on one or more of the image data and video data in the external data to obtain feature data corresponding to the external data, and correlation information is extracted from the feature data, device data and detection data. Exemplarily, the correlation information may be information such as the device unique number and device model that can characterize the association relationship between devices. Preferably, the feature data, device data and detection data corresponding to the power equipment can be integrated through the identification information of the power equipment to obtain multi-dimensional feature data of the power equipment.

[0043] In this embodiment, image feature extraction is performed from one or more of the image data and video data in the external data to obtain feature data of the image dimension, and multi-dimensional feature data such as feature data, equipment data and detection data are associated to obtain multi-dimensional feature data of the power equipment. This implements deep data mining of the multi-dimensional data of the power equipment to obtain more feature data of the power equipment, provide data support for subsequent detection of the equipment status, and help improve the accuracy of equipment status detection.

[0044] Optionally, the feature data, equipment data and detection data are associated to obtain multidimensional feature data of the power equipment, including: extracting identification information of the power equipment from text data in the multidimensional data; and associating the feature data, equipment data and detection data based on the identification information of the power equipment.

[0045] The identification information can be specifically understood as information used to characterize the power equipment, which can be the unique coding information of the power equipment. Feature extraction can be performed from the text information of the power equipment to obtain the identification information of the power equipment. The text data can be equipment data and detection data, or text data extracted from external data.

[0046] Specifically, identification information may be extracted from text data corresponding to dimensions such as equipment data, detection data, and external data, and feature data, equipment data, and detection data having the same identification information may be associated to obtain multi-dimensional feature data of the power equipment.

[0047] S130. Determine the associated characteristic data corresponding to the electric power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and association relationships between the plurality of knowledge entities.

[0048] Among them, the knowledge graph specifically represents the data based on the multidimensional data corresponding to the power system and its association relationship. The knowledge graph includes multiple knowledge entities, attribute information of knowledge entities and association relationships between multiple knowledge entities. The knowledge entity can be specifically understood as a data individual or data set that represents a certain object or equipment operating status in the multidimensional data of the power equipment. A variety of data entities can be identified in the multidimensional data of the power equipment through the entity recognition method. For example, the object can be an equipment indicator light package and an equipment switch. The knowledge entity includes but is not limited to data entities in the equipment ledger data, inspection operation data, picture data and video data. The value corresponding to each data entity can be used as the attribute information of the knowledge entity. For example, the identification information of the power equipment can be used as a knowledge entity, and the specific value or string corresponding to the identification information is used as the attribute information corresponding to the identification information. The association relationship between multiple knowledge entities specifically refers to the dependency relationship or integration relationship between entities. For example, if multiple knowledge entities can jointly represent a certain operating state of the power equipment, it can be determined that there is a certain association relationship between the multiple knowledge entities, and the corresponding association relationship can be set according to the dependency relationship or integration relationship between the knowledge entities. It should be noted that building a knowledge graph requires processing a large amount of data. Therefore, you can obtain massive data in advance, build the knowledge graph in advance according to the knowledge graph construction method, and store the pre-built knowledge graph in the target storage space to facilitate subsequent calls.

[0049] Specifically, the knowledge graph includes multiple knowledge entities, attribute information of knowledge entities and the association relationships between multiple knowledge entities. By calling the pre-built knowledge graph, matching can be performed in the knowledge graph according to the identification information corresponding to the power equipment to obtain the knowledge entity corresponding to the matching identification information. Matching can also be performed in the knowledge graph according to the feature data corresponding to the power equipment to obtain the knowledge entity corresponding to the matching identification information, and then the attribute information of the knowledge entity corresponding to the successfully matched knowledge entity and the association relationships between multiple knowledge entities are used as the associated feature data corresponding to the power equipment.

[0050] In this embodiment, by combining the knowledge graph to obtain the associated feature data corresponding to the power equipment, deeper feature data associated with the power equipment can be obtained, which helps to improve the integrity and comprehensiveness of the feature data obtained for the power equipment, and further helps to improve the accuracy of equipment status detection.

[0051] Optionally, a method for constructing a knowledge graph includes: obtaining historical multidimensional data, wherein the historical multidimensional data includes source data from multiple data sources; extracting knowledge entities from the source data of each data source, extracting attribute information of each knowledge entity and the associations between the knowledge entities, and constructing a local knowledge graph corresponding to each data source based on the attribute information and associations of each knowledge entity; performing consistency processing on the knowledge entities of the local knowledge graphs of multiple data sources, and merging the local knowledge graphs of multiple data sources to obtain a knowledge graph.

[0052] Among them, multiple data sources can be source data of power equipment in the power system collected from multiple data sources such as servers, equipment detection systems, individual operation equipment and external imported equipment. The source data of multiple data sources include but are not limited to video data, image text data, inspection data, equipment circuit panoramic data, equipment ledger data, version information, defect records, operation and maintenance records, setting sheets and equipment manuals. Video data can be data collected by the camera of an individual operation instrument. Image text data includes images, drawings, images and text information in documents such as maintenance manuals of the equipment, as well as image data captured and collected by the camera of an individual operation instrument. Inspection data is data recorded during manual inspections, such as environmental parameters such as temperature, humidity, and pressure. Equipment circuit panoramic data is data related to the internal structure of the equipment and its connection method. Equipment ledger data includes basic information, specification parameters, installation location and other data of the equipment. Version information is information such as the version number of the software or firmware. Defect records are records of faults that have occurred in the past. Operation and maintenance records are historical records of maintenance and repair. Setting sheets are specific numerical settings for equipment configurations. Equipment manuals are guides for the use and maintenance of equipment.

[0053] Specifically, data satisfying multiple dimensions such as equipment data, detection data and external data are obtained from the source data within the historical time period to obtain historical multi-dimensional data, for example: equipment information data includes but is not limited to ledger data, nameplate information, historical operation data, historical inspection data, defect / elimination records and fault / maintenance records; detection data includes but is not limited to equipment production signals, measurement data and control instruction data; external data includes but is not limited to audio and video data. It should be noted that external data also includes meteorological data, power environment data, etc., which can be obtained from the information published by the Meteorological Bureau. Power environment data is data that characterizes whether there is external power data applied to the equipment or power data generated inside the equipment. Meteorological data and power environment data can be used as auxiliary information, which can be used to construct knowledge graphs and can also be used for subsequent status detection of power equipment. Historical multi-dimensional data includes source data from multiple data sources. Knowledge entities are extracted from source data from each data source. Corresponding knowledge entity extraction rules and constraints can be set in advance so that the extracted knowledge entities can meet the needs of power system application scenarios. Knowledge entity extraction rules and constraints can be set according to actual needs and are not limited here. After extracting the knowledge entities from the source data of each data source, attribute information matching the knowledge entities and related knowledge entities, as well as the association relationship between the related knowledge entities, are extracted from the corresponding source data according to the knowledge entities. The corresponding knowledge entities and the attribute information of the knowledge entities are associated according to the association relationship between the knowledge entities to obtain the local knowledge graph corresponding to each data source. After obtaining the local knowledge graphs corresponding to the multiple data sources included in the historical multidimensional data, the local knowledge graphs of the multiple data sources are processed for consistency of knowledge entities. The consistency processing includes but is not limited to the consistency processing of entity names and the consistency processing of data formats to resolve conflicts and ambiguities between knowledge entities from different data sources and determine the consistency of knowledge entities and their attribute information; after merging the local knowledge graphs of multiple data sources after consistency processing, duplicate knowledge entities, attribute information of knowledge entities and association relationships between knowledge entities can be deleted to avoid duplicate information in the knowledge graph, thereby obtaining an effective knowledge graph.

[0054] Exemplarily, building a knowledge graph can also be understood as a data model obtained by integrating and structuring multidimensional data, obtaining a large amount of historical multidimensional data, performing data preprocessing on the historical multidimensional data, and performing entity recognition on the preprocessed historical multidimensional data. The entity recognition technology (Named Entity Recognition, NER) can be used to identify multiple data entities in the preprocessed data, that is, to identify knowledge entities, such as equipment inventory data, inspection operation data, image data, video data, etc. Extract the attribute information of multiple knowledge entities. For each identified knowledge entity, extract its attributes and attribute values, and form structured data. Then extract the relationship between knowledge entities, identify the dependency or integration relationship between knowledge entities, such as extracting the unique coding information of power equipment to identify the dependency or integration relationship between equipment ledger data entities and other entities. Then, construct the ontology of the knowledge graph based on the identified knowledge entities, the attribute information of knowledge entities and the association relationship between knowledge entities. By designing the model layer of the knowledge graph, define the knowledge entity type, relationship type and attributes, as well as the constraints and rules between the knowledge entity type, association relationship type and attributes, and then perform knowledge fusion. By merging knowledge entities and their corresponding information from different data sources, and processing the knowledge entities and corresponding information corresponding to different data sources for consistency, the conflict and ambiguity of knowledge entities can be resolved, the consistency of knowledge entities can be ensured, and the fused knowledge entities, the relationship between knowledge entities and the attributes of knowledge entities can be stored in the database to form a knowledge graph library, so that the state detection model can better understand the correlation and dependency between power equipment, perform retrieval and reasoning, which is helpful to improve the accuracy of state detection of power equipment.

[0055] Optionally, after acquiring historical multidimensional data, data preprocessing is performed on the source data of multiple data sources in the historical multidimensional data. The source data of multiple data sources are processed by data cleaning operations such as noise removal, missing value filling, and outlier detection to obtain relatively complete and accurate source data. The source data after data cleaning is annotated. For example, for image data, it is necessary to mark specific objects in the image data to ensure that knowledge entities can be accurately extracted. This enables more accurate knowledge extraction and improves the accuracy and reliability of the constructed knowledge graph. It also includes linearly combining the preprocessed data and its data features into a set of new features as the main components of the data through PCA algorithm technology, and determining whether there is a mutually orthogonal relationship between the main components. If they are mutually orthogonal, it means that there is no correlation between the corresponding multidimensional data, and the corresponding data can be deleted. Retaining the data with correlation can reduce the dimension of the multidimensional data set, which provides assistance for subsequent data association mining and avoids irrelevant data interfering with the mining of data associations.

[0056] On the basis of the above-mentioned embodiment, based on the pre-constructed knowledge graph, the associated characteristic data corresponding to the power equipment is determined, including: matching the identification information or characteristic data of the power equipment in the knowledge graph to determine the successfully matched knowledge entity; based on the association relationship between the knowledge entities in the knowledge graph, determining the associated knowledge entity corresponding to the successfully matched knowledge entity, and determining the associated characteristic data based on the attribute information of the associated knowledge entity.

[0057] Specifically, in the process of determining the associated characteristic data corresponding to the power equipment in combination with the pre-built knowledge graph, matching can be performed in the knowledge graph according to the identification information of the power equipment, specifically matching the identification information with the knowledge entity in the knowledge graph or the attribute information of the knowledge entity to obtain a successfully matched knowledge entity, or matching can be performed in the knowledge graph according to the characteristic data of the power equipment, specifically matching the characteristic data of the power equipment with the knowledge entity in the knowledge graph or the attribute information of the knowledge entity to obtain a successfully matched knowledge entity. According to the successfully matched knowledge entity, the corresponding association relationship is obtained, and the knowledge entity associated with the successfully matched knowledge entity is obtained according to the association relationship, and the associated knowledge entity corresponding to the successfully matched knowledge entity is obtained, and the attribute information corresponding to the associated knowledge entity is extracted, and the attribute information corresponding to the associated knowledge entity can be used as the associated characteristic data.

[0058] S140. Perform state detection based on multidimensional feature data and associated feature data using at least one pre-trained state detection model to obtain at least one state detection result of the electric power equipment.

[0059] Among them, the state detection model specifically refers to a deep learning model used for state detection of power equipment. Model training can be performed according to different state detection requirements to obtain at least one state detection model. The state detection result can be specifically understood as the output result of the state detection model. Each state detection model corresponds to a set of detection results. Optionally, the state detection result includes one or more of the following: the health status and failure mode of the power equipment at the current moment; the predicted health status and predicted failure mode of the power equipment in the future time period; the remaining life prediction result of the power equipment. The training of the state detection model can use the extracted feature data to train the deep learning model to obtain a model that can meet the state detection requirements.

[0060] Specifically, different state detection models are called according to actual needs, and one or more state detection models can be called at the same time. The obtained multidimensional feature data and associated feature data are used as input data of the state detection model. The state detection model processes the multidimensional feature data and the associated feature data to obtain the state detection result of the corresponding state detection model. If one state detection model is called, one state detection result of the power equipment is obtained; if multiple state detection models are called, multiple state detection results of the power equipment are obtained, that is, each state detection model corresponds to one state detection result.

[0061] In this embodiment, by calling one or more state detection models according to actual needs, state detection can be performed based on multidimensional feature data and associated feature data at the same time, thereby obtaining a state detection result corresponding to each state detection model, which can intelligently judge the operating state of the power equipment, provide a basis for power equipment fault diagnosis and equipment state prediction, improve the accuracy and efficiency of the state detection of the power equipment, and can also meet the needs of different power equipment for state detection in different scenarios, thereby improving the applicability and scalability of the state detection method.

[0062] In a specific embodiment, the processing flow of the detection status of power equipment is as follows: Step 1, data collection: specifically collect multidimensional data from multiple sources, where equipment information data includes ledgers, nameplates, operation history, inspection history, defect / elimination records, fault / maintenance records; detection data includes real-time signals, measurements, and control instruction data of equipment production; external data includes audio and video data, meteorological data, and power environment data. Step 2, data preprocessing: first clean the collected multidimensional data, and ensure that the data set does not contain duplicate records by deleting duplicate values ​​to avoid bias in the analysis results, then use the Z-score method to identify and process outliers, and fill missing values ​​with the mean or median, fuse the cleaned data, and integrate data from different data sources to ensure data consistency and integrity. Classify and annotate the data that needs further explanation to make it structured data that can be understood by the deep learning model, and store the processed data in the database as a model training set. Step 3, data feature extraction and association: Useful information that helps deep learning model learning and prediction is extracted from the preprocessed data using the yolov5 algorithm as feature data. The algorithm associates the feature data according to certain rules. For example, the closed state in the equipment inspection picture can be used as a feature data and associated with the corresponding substation. Step 4, build a knowledge graph: Based on the preprocessed data and the feature data extracted by the yolov5 algorithm, the semantic understanding, context perception and autonomous dynamic learning capabilities of the AI ​​large language model LLM are used to build a knowledge graph; support vectorization processing, and use the Elasticsearch intelligent search engine to establish a full-text retrieval index to achieve fast and intelligent retrieval of multidimensional data information. Step 5, model training: Based on the preprocessed data and the feature data extracted by the yolov5 algorithm, decision trees and support vector machines are used as machine learning algorithms, and convolutional neural network optimization algorithms are used to perform unsupervised feature deep learning to train the model, continuously improving the accuracy and efficiency of data analysis. Step 6, output of multidimensional data analysis results. After conducting model intelligent analysis on the collected multi-dimensional data, comprehensive analysis results were obtained, such as conducting intelligent inspections on power equipment and automatically generating inspection reports; conducting intelligent analysis and fault diagnosis on the multi-dimensional operating status data of power equipment; and conducting intelligent predictions on the multi-dimensional operating status data of current power equipment to obtain the detection status results of power equipment.

[0063] The technical solution of this embodiment is to obtain multidimensional data of power equipment, wherein the multidimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; extract features from external data to obtain feature data, associate feature data, equipment data and detection data to obtain multidimensional feature data of power equipment; determine associated feature data corresponding to power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes multiple knowledge entities, attribute information of knowledge entities and associations between multiple knowledge entities; perform state detection based on multidimensional feature data and associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of power equipment. This solution realizes deep mining of multidimensional data of power equipment by extracting features from multidimensional data and determining associated feature data in combination with knowledge graph, and uses prediction model to process multidimensional feature data and associated feature data to obtain state detection result, solves the problem of inaccurate and inefficient state detection of power equipment, helps to warn potential risk points in advance, and provides decision support for preventive maintenance of power equipment, while reducing the probability of errors or accidents caused by human factors, thereby improving the operating efficiency and safety of power equipment.

[0064] Embodiment 2

[0065] Figure 2 This is a flow chart of a method for detecting the state of an electric power device provided in the second embodiment of the present invention. This embodiment is a further optimization of the method of the above embodiment. Optionally, a key video frame in the image data or video data is used as the image to be processed; target detection is performed on the image to be processed to obtain feature points of the image to be processed, and the descriptive features of the image to be processed are determined based on the feature points; clustering is performed on the image to be processed based on the descriptive features of the image to be processed to determine the image class to which the image to be processed belongs, and feature data is determined based on the image class to which the image to be processed belongs, wherein the feature data includes one or more of the device appearance features and the operational behavior features. Figure 2 As shown, the method includes:

[0066] S210. Acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data, and external data, and the external data includes one or more of video data and image data.

[0067] S220: Use the key video frame in the image data or video data as an image to be processed.

[0068] Specifically, if the acquired external data includes image data, any image data can be used as the image to be processed. If the acquired external data includes video data, it is necessary to extract key video frames from the video data and use the extracted key video frames as the image to be processed. The key video frame extraction can be performed by calling an image similarity algorithm to calculate the similarity between any frame image and the corresponding previous frame image to obtain a similarity calculation result. If the similarity result is lower than a preset similarity threshold, the image of the corresponding frame can be used as the key video frame.

[0069] S230 , performing target detection on the image to be processed to obtain feature points of the image to be processed, and determining description features of the image to be processed based on the feature points.

[0070] Among them, feature points can be specifically understood as points where the grayscale value in the image changes dramatically or points with large curvature on the edge of the image. Feature points can be obvious edges, texture areas or spots of the image, usually with strong grayscale changes, texture information or local structures, and are different from other points in adjacent areas. Feature point detection algorithms include SIFT (Scale-invariant feature transform) algorithm, SURF (Speeded-UpRobust Features) algorithm and ORB (Oriented FAST and RotatedBRIEF) algorithm, which detect feature points by finding stable key points at different scales.

[0071] Specifically, the feature point detection algorithm is called to perform target detection on the image to be processed, the feature points in the image to be processed are obtained, and the extracted feature points are described. For example, the ORB algorithm is called to perform target detection on the image to be processed. The ORB algorithm mainly includes two parts: feature point extraction and feature point description. The feature point extraction uses the FAST algorithm to find special areas from the image, which are called key points. Key points are small prominent areas in the image, such as corners, with the feature that the pixel value changes sharply from light to dark. The feature point description is based on the BRIEF algorithm to calculate the corresponding feature vector for each key point. The feature vector created by the ORB algorithm only contains 1 and 0, which is called a binary feature vector. A binary feature vector is created through a set of key points to obtain the corresponding binary description feature. The description features corresponding to the feature points of the image to be processed constitute the description features of the image to be processed.

[0072] S240, clustering the image to be processed based on the descriptive features of the image to be processed, determining the image class to which the image to be processed belongs, and determining feature data based on the image class to which the image to be processed belongs, wherein the feature data includes one or more of device appearance features and operation behavior features.

[0073] Specifically, clustering is performed on the descriptive features of the image to be processed by a clustering algorithm to obtain at least one cluster, each cluster can be used as an image class, and the image class to which the image to be processed belongs is determined based on the clustering result. Each image class includes one or more features of device appearance features and operation behavior features, and feature data corresponding to the image to be processed can be determined based on the features corresponding to the image class to which the image to be processed belongs. Exemplarily, the descriptive features of the image to be processed specifically represent the display status of the indicator light of the device, and clustering is performed on the descriptive features of the image to be processed, and the feature data of the image class to which the image to be processed belongs can be obtained as the device appearance features.

[0074] Optionally, determining feature data based on the image class to which the image to be processed belongs includes: determining the feature data corresponding to the image class to which the image to be processed belongs as the feature data corresponding to the image to be processed; or; classifying the descriptive features of the image to be processed and / or the clustering centers of the image class through a classification model to obtain the feature data.

[0075] It is understandable that any cluster has corresponding feature data, and the feature data corresponding to each cluster can be obtained according to the clustering result, that is, the feature data corresponding to each image class can be obtained from the clustering result, and the feature data corresponding to each image class can be directly set as the feature data corresponding to the image to be processed. Alternatively, the classification model is called to classify the descriptive features of the image to be processed, so as to obtain the feature data of the image to be processed, or the classification model is called to classify the cluster center of the image class of the image to be processed, so as to obtain the feature data of the image to be processed, or the classification model is used to classify the descriptive features of the image to be processed and the cluster center of the image class, so as to obtain the feature data of the image to be processed.

[0076] In this embodiment, the descriptive features of the image to be processed and / or the cluster centers of the image classes are classified through a classification model, and the feature data corresponding to the image to be processed can be determined through further classification processing, thereby improving the accuracy of the classification results of the feature data in the image to be processed and facilitating the accuracy of feature data extraction.

[0077] S250, associating the characteristic data, the equipment data and the detection data to obtain multi-dimensional characteristic data of the power equipment.

[0078] S260. Determine the associated characteristic data corresponding to the electric power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and association relationships between the plurality of knowledge entities.

[0079] S270. Perform state detection based on multidimensional feature data and associated feature data using at least one pre-trained state detection model to obtain at least one state detection result of the electric power equipment.

[0080] The technical solution of this embodiment is to obtain multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; take the key video frame in the image data or video data as the image to be processed; perform target detection on the image to be processed to obtain the feature points of the image to be processed, and determine the descriptive features of the image to be processed based on the feature points; perform clustering processing on the image to be processed based on the descriptive features of the image to be processed to determine the image class to which the image to be processed belongs, and determine the feature data based on the image class to which the image to be processed belongs, wherein the feature data includes one or more of equipment appearance features and operation behavior features; associate the feature data, equipment data and detection data to obtain the multi-dimensional feature data of the power equipment; determine the associated feature data corresponding to the power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes multiple knowledge entities, attribute information of the knowledge entities and the association relationship between the multiple knowledge entities; perform state detection based on the multi-dimensional feature data and the associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment. This solution achieves in-depth mining of multidimensional data of power equipment by extracting features from multidimensional data and determining associated feature data in combination with knowledge graphs. It also uses prediction models to process multidimensional feature data and associated feature data to obtain status detection results, solving the problems of inaccurate and inefficient status detection of power equipment, helping to provide early warning of potential risk points, and providing decision support for preventive maintenance of power equipment. It also reduces the probability of errors or accidents caused by human factors, thereby improving the operating efficiency and safety of power equipment.

[0081] Embodiment 3

[0082] Figure 3 Schematic diagram of the structure of a state detection device for electric power equipment provided by the third embodiment of the present invention. Figure 3 As shown, the device comprises:

[0083] A multi-dimensional data acquisition module 310, used to acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data;

[0084] The multi-dimensional characteristic data determination module 320 is used to extract characteristics of external data to obtain characteristic data, and associate the characteristic data, equipment data and detection data to obtain multi-dimensional characteristic data of the power equipment;

[0085] The associated characteristic data determination module 330 is used to determine the associated characteristic data corresponding to the power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and associations between the plurality of knowledge entities;

[0086] The state detection result determination module 340 is used to perform state detection based on multi-dimensional feature data and associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment.

[0087] The technical solution of this embodiment is to obtain multi-dimensional data of the power equipment through a multi-dimensional data acquisition module, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; the multi-dimensional feature data determination module extracts features from the external data to obtain feature data, and associates the feature data, equipment data and detection data to obtain multi-dimensional feature data of the power equipment; the associated feature data determination module determines the associated feature data corresponding to the power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes multiple knowledge entities, attribute information of the knowledge entities and associations between the multiple knowledge entities; the state detection result determination module is used to perform state detection through multi-dimensional feature data and associated feature data of at least one state detection model trained in advance, and obtain at least one state detection result of the power equipment. This solution achieves in-depth mining of multidimensional data of power equipment by extracting features from multidimensional data and determining associated feature data in combination with knowledge graphs. It also uses prediction models to process multidimensional feature data and associated feature data to obtain status detection results, solving the problems of inaccurate and inefficient status detection of power equipment, helping to provide early warning of potential risk points, and providing decision support for preventive maintenance of power equipment. It also reduces the probability of errors or accidents caused by human factors, thereby improving the operating efficiency and safety of power equipment.

[0088] On the basis of the above embodiment, optionally, the multi-dimensional feature data determination module 320 includes a unit for determining an image to be processed, a unit for determining a description feature of the image to be processed, and a unit for determining feature data. The unit for determining an image to be processed is used to use a key video frame in the image data or video data as an image to be processed; the unit for determining a description feature of the image to be processed is used to perform target detection on the image to be processed, obtain feature points of the image to be processed, and determine the description features of the image to be processed based on the feature points; the feature data determination unit is used to perform clustering processing on the image to be processed based on the description features of the image to be processed, determine the image class to which the image to be processed belongs, and determine feature data based on the image class to which the image to be processed belongs, wherein the feature data includes one or more of device appearance features and operation behavior features.

[0089] Optionally, the feature data determination unit is specifically used to determine the feature data corresponding to the image class to which the image to be processed belongs as the feature data corresponding to the image to be processed; or; classify the descriptive features of the image to be processed and / or the clustering center of the image class through a classification model to obtain the feature data.

[0090] Optionally, the multi-dimensional feature data determination module 320 is specifically used to extract identification information of the electric power equipment from the text data in the multi-dimensional data; and associate the feature data, the equipment data and the detection data based on the identification information of the electric power equipment.

[0091] Optionally, the associated feature data determination module 330 includes a knowledge graph construction unit. The knowledge graph construction unit is used to obtain historical multidimensional data, wherein the historical multidimensional data includes source data from multiple data sources; extract knowledge entities from the source data of each data source, extract the attribute information of each knowledge entity and the association relationship between each knowledge entity, and construct a local knowledge graph corresponding to each data source based on the attribute information and association relationship of each knowledge entity; perform consistency processing on the local knowledge graphs of multiple data sources, and merge the local knowledge graphs of multiple data sources to obtain a knowledge graph.

[0092] Optionally, the associated characteristic data determination module 330 is specifically used to match the identification information or characteristic data of the power equipment in the knowledge graph to determine the successfully matched knowledge entity; based on the association relationship between the knowledge entities in the knowledge graph, determine the associated knowledge entity corresponding to the successfully matched knowledge entity, and determine the associated characteristic data based on the attribute information of the associated knowledge entity.

[0093] Optionally, the status detection result includes one or more of the following: the health status and failure mode of the power equipment at the current moment; the predicted health status and predicted failure mode of the power equipment in a future time period; and the remaining life prediction result of the power equipment.

[0094] The state detection device of the electric equipment provided in the embodiment of the present invention can execute the state detection method of the electric equipment provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0095] Embodiment 4

[0096] Figure 4 1 is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0097] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0098] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0099] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a state detection method for power equipment.

[0100] In some embodiments, the state detection method of the power device may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the state detection method of the power device described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the state detection method of the power device in any other appropriate manner (e.g., by means of firmware).

[0101] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0102] The computer program for implementing the state detection method of the power equipment of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program can be executed entirely on the machine, partially on the machine, partially on the machine as an independent software package and partially on a remote machine, or completely on a remote machine or server.

[0103] Embodiment 5

[0104] Embodiment 5 of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to execute a state detection method of an electric power device, the method comprising:

[0105] Acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data;

[0106] Extract features from external data to obtain feature data, associate the feature data, equipment data and detection data to obtain multi-dimensional feature data of the power equipment;

[0107] Determine the associated characteristic data corresponding to the electric power equipment based on a pre-built knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and association relationships between the plurality of knowledge entities;

[0108] By using at least one state detection model that has been pre-trained, state detection is performed based on multi-dimensional feature data and associated feature data to obtain at least one state detection result of the electric equipment.

[0109] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0110] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0111] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0112] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0113] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0114] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting the state of an electric power device, characterized in that: include: Acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; Extracting features from the external data to obtain feature data, and associating the feature data, the device data, and the detection data to obtain multi-dimensional feature data of the power device; Determine the associated feature data corresponding to the electric power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities, and association relationships between the plurality of knowledge entities; By using at least one pre-trained state detection model, state detection is performed based on the multi-dimensional feature data and the associated feature data to obtain at least one state detection result of the electric power equipment.

2. The method according to claim 1, characterized in that Extracting features from the external data to obtain feature data includes: Using the image data or a key video frame in the video data as an image to be processed; Performing target detection on the image to be processed to obtain feature points of the image to be processed, and determining description features of the image to be processed based on the feature points; The image to be processed is clustered based on the descriptive features of the image to be processed to determine the image class to which the image to be processed belongs, and the feature data is determined based on the image class to which the image to be processed belongs, wherein the feature data includes one or more of device appearance features and operation behavior features.

3. The method according to claim 2, characterized in that The determining the feature data based on the image class to which the image to be processed belongs includes: Determine the feature data corresponding to the image class to which the image to be processed belongs as the feature data corresponding to the image to be processed; or; The feature data is obtained by classifying the descriptive features of the image to be processed and / or the cluster centers of the image classes through a classification model.

4. The method according to claim 1, characterized in that: The step of associating the characteristic data, the equipment data and the detection data to obtain multi-dimensional characteristic data of the electric power equipment includes: Extracting identification information of the electric power equipment from text data in the multidimensional data; The characteristic data, the device data, and the detection data are associated based on the identification information of the electric power device.

5. The method according to claim 1, characterized in that: The method for constructing the knowledge graph includes: Acquiring historical multidimensional data, wherein the historical multidimensional data includes source data from multiple data sources; Extracting knowledge entities from source data of each of the data sources, extracting attribute information of each of the knowledge entities and the association relationship between the knowledge entities, and constructing a local knowledge graph corresponding to each of the data sources based on the attribute information and the association relationship of each of the knowledge entities; The local knowledge graphs of the multiple data sources are processed for consistency of knowledge entities, and the local knowledge graphs of the multiple data sources are merged to obtain the knowledge graph.

6. The method according to claim 1 or 5, characterized in that: Based on the pre-built knowledge graph, determining the associated feature data corresponding to the electric power equipment includes: Matching the identification information or the characteristic data of the electric power equipment in the knowledge graph to determine a successfully matched knowledge entity; Based on the association relationship between each of the knowledge entities in the knowledge graph, the associated knowledge entity corresponding to the successfully matched knowledge entity is determined, and the associated feature data is determined based on the attribute information of the associated knowledge entity.

7. The method according to claim 1, characterized in that The status detection result includes one or more of the following: The health status and failure mode of the power equipment at the current moment; The predicted health status and predicted failure mode of the power equipment in a future time period; The remaining life prediction result of the electric equipment.

8. A state detection device for electric power equipment, characterized in that: include: A multi-dimensional data acquisition module, used to acquire multi-dimensional data of the power equipment, wherein the multi-dimensional data includes equipment data, detection data and external data, and the external data includes one or more of video data and image data; A multi-dimensional characteristic data determination module, used for extracting characteristics of the external data to obtain characteristic data, and associating the characteristic data, the equipment data and the detection data to obtain multi-dimensional characteristic data of the electric equipment; An associated characteristic data determination module, used to determine the associated characteristic data corresponding to the electric power equipment based on a pre-constructed knowledge graph, wherein the knowledge graph includes a plurality of knowledge entities, attribute information of the knowledge entities and association relationships between the plurality of the knowledge entities; The state detection result determination module is used to perform state detection based on the multidimensional feature data and the associated feature data through at least one pre-trained state detection model to obtain at least one state detection result of the power equipment.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the state detection method of the electric power equipment according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the state detection method of the power equipment according to any one of claims 1 to 7 when executed.

Images