Abnormal state monitoring method and device of power equipment and electronic equipment

Abstract

The invention discloses an abnormal state monitoring method and device of power equipment and electronic equipment. The abnormal state monitoring method of the power equipment in the embodiment of the invention comprises the following steps: acquiring operation data of to-be-monitored power equipment from at least one data source; performing collaborative extraction on the operation data of each data source to obtain fusion features of the operation data of each data source; the fusion features are input into an abnormal state detection model for abnormal state detection, a detection result of the abnormal state of the power equipment is obtained, the abnormal state detection model comprises a clustering model and a classification model, and the detection result of the abnormal state of the power equipment is a fusion value of an output result of the clustering model and an output result of the classification model. According to the invention, the safety and stability of the power equipment are improved.

Description

Technical Field

[0001] The present application belongs to the field of electric power technology, and in particular, relates to a method, device and electronic equipment for monitoring abnormal status of electric power equipment. Background Art

[0002] The power grid control system is an important component of the power system for monitoring, controlling and optimizing the operation of the power grid. It adjusts the operating parameters of the power equipment in each link of power generation, transmission and distribution by monitoring the state of the power grid in real time. Ensure the safe, stable and efficient operation of the power grid. Power equipment includes terminal power equipment, power transmission equipment, etc. Terminal power equipment includes generators, transformers and circuit breakers. Power transmission equipment includes transmission lines and substations. With the increasing number of types of power equipment, the monitoring data of these power equipment becomes massive data after aggregation. These massive data contain the status of power equipment. Therefore, how to use massive data to monitor the abnormal status of power equipment to improve the safety and stability of power equipment is a technical problem that technicians in this field need to solve. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a method, device and electronic device for monitoring the abnormal state of power equipment, which can monitor the abnormal state of power equipment and improve the safety and stability of the power equipment.

[0004] In a first aspect, a method for monitoring an abnormal state of an electric power device is provided, which is executed by a terminal, and the method comprises: acquiring operating data of the electric power device to be monitored from at least one data source; collaboratively extracting the operating data of each of the data sources to obtain fused features of the operating data of each of the data sources; inputting the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the electric power device, the abnormal state detection model comprises a clustering model and a classification model, and the detection result of the abnormal state of the electric power device is a fused value of an output result of the clustering model and an output result of the classification model.

[0005] In a second aspect, an abnormal state monitoring device for electric equipment is provided, comprising: an acquisition module for acquiring operating data of the electric equipment to be monitored from at least one data source; an extraction module for collaboratively extracting the operating data of each of the data sources to obtain fused features of the operating data of each of the data sources; a detection module for inputting the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the electric equipment, the abnormal state detection model comprising a clustering model and a classification model, and the detection result of the abnormal state of the electric equipment is a fused value of an output result of the clustering model and an output result of the classification model.

[0006] In a third aspect, an embodiment of the present application provides an electronic device comprising: a memory, a processor, and computer executable instructions stored in the memory and executable on the processor, wherein the computer executable instructions, when executed by the processor, implement the steps of executing the method described in the first aspect.

[0007] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which is used to store computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0008] In an embodiment of the present application, first, operating data of the power equipment to be monitored is acquired from at least one data source; then, the operating data of each of the data sources are collaboratively extracted to obtain fused features of the operating data of each of the data sources; finally, the fused features are input into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the power equipment, wherein the abnormal state detection model includes a clustering model and a classification model, and the detection result of the abnormal state of the power equipment is a fused value of an output result of the clustering model and an output result of the classification model.

[0009] In this way, the embodiment of the present application can obtain the operating data of the power equipment to be monitored from at least one data source, and collaboratively extract the operating data of each data source, so as to achieve the purpose of processing the massive data of the power equipment and extracting features. The extracted fusion features are further detected for abnormal states through the abnormal state detection model. Thus, the abnormal state of the power equipment is detected, thereby improving the safety and stability of the power equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0011] Figure 1 It is a flowchart of an abnormal state monitoring method of electric power equipment provided by an exemplary embodiment of the present application.

[0012] Figure 2 It is a structural schematic diagram of an abnormal state monitoring device for electric power equipment provided by an exemplary embodiment of the present application.

[0013] Figure 3 It is a schematic diagram of the structure of an electronic device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0014] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0015] The technical solution provided by the embodiments of the present application is described in detail below through some embodiments and their application scenarios in combination with the accompanying drawings.

[0016] Figure 1 A flow chart of a method for monitoring abnormal status of power equipment provided by an embodiment of the present application is shown. The method can be executed by an electronic device, such as a terminal device or a server device. In other words, the method can be executed by software or hardware installed in the terminal device or the server device. Figure 1 As shown, the method may include the following steps.

[0017] S101, obtaining operating data of a power device to be monitored from at least one data source.

[0018] Specifically, the data sources in the embodiments of the present application include but are not limited to the background monitoring data of the Energy Management System (EMS), alarm signals, operating status data of the power equipment, and external data. Among them, the external data includes but is not limited to weather data and grid load forecasts. The operating status data of the power equipment can be collected in real time through the Supervisory Control and Data Acquisition (SCADA) system. The collection frequency can be determined according to the actual situation. In the embodiments of the present application, the value is 1 minute to ensure the real-time and integrity of the data. In addition, historical operation data can also be exported in batches from the historical database to ensure the integrity of the data.

[0019] Furthermore, after obtaining the operating data of the power equipment, as an optional embodiment of the present application, the operating data of the power equipment of each data source is first preprocessed to obtain the target operating data, and the preprocessing includes data cleaning, data conversion and data normalization; then the target storage method of the target operating data is determined according to the type of the target operating data; finally, the target operating data is stored in the target storage method. Among them, the preprocessing of the data includes data cleaning, data conversion and data normalization, so as to ensure the quality and consistency of the data. Among them, data cleaning includes but is not limited to null value filling, outlier processing and duplicate data deletion. Data conversion is the conversion of non-numerical data into numerical data, such as converting the state of the power equipment into numerical form, to facilitate subsequent analysis and modeling. Data normalization is to eliminate the influence of dimension, thereby improving the training effect of the model.

[0020] Furthermore, in the data storage stage, the target storage method of the target operating data can be determined according to the type of the target operating data. Among them, the methods of storing data include but are not limited to distributed file systems (Hadoop Distributed File System, HDFS), relational databases (MySQL) and NoSQL databases (such as MongoDB). Among them, the HDFS distributed file system is suitable for the storage of large-scale data and supports high-concurrency read and write operations. The relational database (MySQL) is suitable for the storage and query of structured data. The NoSQL database is suitable for the storage of semi-structured and unstructured data. If the type of the target operating data is large-scale massive data, the distributed file system HDFS can be used. If the type of the target operating data is structured data, the relational database (MySQL) can be used. If the type of the target operating data is semi-structured or unstructured data, a NoSQL database can be used.

[0021] Furthermore, the architecture of the data warehouse can be designed, including data models, table structures, and indexes. The data model refers to an abstract description of data features, which is used to describe the data structure, the relationship between data, and the formal framework of data operation rules. The data model plays a core role in database design and information system development. It helps developers and users understand how data is organized, stored, and how to access and operate the data. The content described by the data model consists of three parts: data structure, data operation, and data constraints. The table structure refers to the design and organization of the table in the database, which defines the fields, data types, and constraints in the table. The table structure is a core component of the relational database and directly affects the storage, query, and management efficiency of data. The index is a decentralized storage structure created to speed up the retrieval of data rows in the table. The index is established for the table, which is composed of index pages other than the data page. Each row in the index page will have a logical pointer to speed up the retrieval of physical data. After the data warehouse is built, the preprocessed running data is loaded into the data warehouse to ensure the integrity and consistency of the data. The data quality in the data warehouse is regularly checked to detect and handle data quality problems in a timely manner.

[0022] Furthermore, the embodiment of the present application can also determine the computing platform according to the data volume of the operation data of the power equipment and the processing requirements of the operation data. As an optional embodiment of the present application, the computing platform is first determined according to the data volume of the operation data of the power equipment of each data source and the processing requirements of the operation data; then the cluster is configured on the computing platform, and the resource scheduling method of the cluster is determined; finally, the abnormal state detection model is deployed in the cluster, and the resources in the cluster are scheduled in the resource scheduling method so that the abnormal state detection model uses the resources in the cluster.

[0023] Specifically, the types of computing platforms include, but are not limited to, Apache Spark computing platform, Apache Flink computing platform, and TensorFlow computing platform. Among them, the Apache Spark computing platform is suitable for real-time computing and offline analysis of large-scale data, and supports multiple programming languages, such as Python, Scala, and Java. The Apache Flink computing platform is suitable for real-time streaming data processing, and supports low-latency and high-throughput data processing. The TensorFlow computing platform is suitable for the training and reasoning of deep learning models, and supports distributed computing of large-scale data. After selecting a suitable computing platform, configure the hardware and software environment of the computing cluster to ensure the high availability and scalability of the cluster. Then determine the resource scheduling method of the cluster. The embodiment of the present application can use the YARN core component or Kubernetes cluster in the Hadoop ecosystem for resource management and scheduling to ensure efficient execution of computing tasks. After determining the resource scheduling method of the cluster, data processing is performed, including writing data processing scripts, using tools such as MapReduce and Spark SQL for data processing and analysis, and the data processing script includes deploying an abnormal state detection model in the cluster, and scheduling the resources in the cluster in a resource scheduling manner so that the abnormal state detection model can use the resources in the cluster.

[0024] S102, collaboratively extracting the operating data of each data source to obtain fusion features of the operating data of each data source.

[0025] Specifically, in the stage of building a business-oriented analysis model, data collection features such as timestamp, voltage, current and temperature are extracted from the perspective of data mining. And based on the knowledge of electricity, features related to the formula, such as power factor, load rate, etc., are extracted. The collaborative extraction method is used to realize the fusion and feature extraction of multi-source heterogeneous data, and improve the representativeness of the extracted features and the accuracy of the model. Among them, the collaborative extraction method is a method of extracting data from multiple sources or different types of heterogeneous data, aiming to improve the quality and representativeness of the extracted features, thereby enhancing the performance of the model. In power system analysis or other types of complex data analysis, the collaborative extraction method can effectively integrate multi-dimensional information from different sensors, different time periods or different systems. By analyzing the inherent connections between these data, collaborative extraction can mine deep-level features that are difficult to display from a single data source, thereby providing more accurate and comprehensive support for subsequent analysis or prediction. Not only does it improve the representativeness of the features, but it also enhances the model's adaptability to complex scenarios.

[0026] S103, inputting the fused features into an abnormal state detection model to perform abnormal state detection, and obtaining a detection result of the abnormal state of the power equipment.

[0027] The abnormal state detection model includes a clustering model and a classification model, and the detection result of the abnormal state of the power equipment is a fusion value of the output result of the clustering model and the output result of the classification model.

[0028] Specifically, the clustering model in the embodiment of the present application can use the K-means clustering algorithm, which is used to preliminarily group the fusion features, select the appropriate number of clusters, and determine the optimal number of cluster centers by calculating the sum of squared errors and the sum of squared errors. Classification models include but are not limited to random forests, support vector machines, and deep learning models.

[0029] Furthermore, when performing abnormal state detection on the abnormal state detection model, as an optional embodiment of the present application, the fused features are first input into the clustering model and the classification model of the abnormal state detection model respectively; then the fused features output by the clustering model are obtained as the first prediction value of the abnormal state, and the fused features output by the classification model are obtained as the second prediction value of the abnormal state; the first prediction value and the second prediction value are weighted to obtain a third prediction value of the abnormal state detection model, and the third prediction value represents the probability that the power equipment is in an abnormal state; finally, when the third prediction value is greater than the first threshold value, it is determined that the state of the power equipment is an abnormal state.

[0030] Specifically, when cluster analysis is performed through the abnormal state detection model, the output results of the clustering model and the output results of the classification model can be fused. Result fusion refers to using the classification model and the clustering model to obtain the predicted value of the fusion characteristic abnormality of the power equipment, and then weighting the predicted values ​​output by the two models. Among them, the first prediction value and the second prediction value are the probabilities that the state of the power equipment is an abnormal state. Among them, in the embodiment of the present application, the weighting coefficient of the classification model is set to be greater than the weighting coefficient of the clustering model. The first threshold value can be determined according to the actual situation, and is not limited here in the embodiment of the present application.

[0031] Furthermore, when performing abnormal state detection on the abnormal state detection model, as an optional embodiment of the present application, the fused features are input into the clustering model for clustering to obtain the cluster centers of all clusters; the distance from each sample point to the cluster center is calculated, and the number of the cluster to which each sample point belongs is marked; the number of the cluster to which each sample point belongs and the distance from each sample point to the cluster center are input into the classification model; the fourth prediction value of each sample point output by the classification model is obtained, and the fourth prediction value represents the probability that each sample point is abnormal; when the fourth prediction value is greater than the second threshold, the state of the power equipment is determined to be an abnormal state.

[0032] Specifically, when abnormal state detection is performed through the abnormal state detection model, feature fusion can also be performed on the clustering model and the classification model. The fused features are input into the clustering model for clustering to obtain the cluster center of the cluster, and then the distance from each sample point to the cluster center is calculated, and the number of the cluster to which each sample point belongs is marked, that is, the classification to which each sample point belongs. The number of the cluster to which each sample point belongs and the distance from each sample point to the cluster center are then input into the classification model and used as two feature dimensions of the classification model to form a new set of sample attributes. The obtained samples are then classified to obtain the probability that each sample is abnormal. If it is greater than the second threshold, it means that the state of the power equipment is abnormal. The second threshold can be determined according to the actual situation, and the embodiments of the present application are not limited here.

[0033] Furthermore, before using the abnormal state detection model to detect the abnormal state, the initial abnormal state detection model can also be trained to obtain a trained abnormal state detection model. As an optional embodiment of the present application, before inputting the fusion feature into the abnormal state detection model to detect the abnormal state and obtain the detection result of the abnormal state of the power equipment, the method also includes: obtaining the historical operation data of the power equipment; marking the abnormal state data in the historical operation data to obtain negative samples, and determining the non-abnormal state data in the historical operation data as positive samples; inputting the positive samples and negative samples into the initial abnormal state detection model for iterative training to obtain the abnormal state detection model.

[0034] Specifically, the historical operation data acquired by the embodiment of the present application includes but is not limited to the historical background monitoring data of the Energy Management System (EMS), historical alarm signals, historical operation status data of power equipment, and historical external data. The abnormal state data in the historical operation data is then marked as a negative sample, and then the non-abnormal state data is determined as a positive sample. Finally, both the positive samples and the negative samples are input into the initial abnormal state detection model for iterative training to obtain the abnormal state detection model. The initial abnormal state detection model includes but is not limited to a classification model and a clustering model. In addition, the initial abnormal state detection model can also be of other types, which are not limited in the embodiment of the present application.

[0035] The embodiment of the present application can obtain the operating data of the power equipment to be monitored from at least one data source, and extract the operating data of each data source in a coordinated manner, so as to achieve the purpose of processing the massive data of the power equipment and extracting features. The extracted fusion features are further detected for abnormal states through an abnormal state detection model. Thus, the abnormal state of the power equipment is detected, thereby improving the safety and stability of the power equipment.

[0036] Furthermore, as an optional embodiment of the present application, after the positive samples and negative samples are input into the initial abnormal state detection model for iterative training to obtain the abnormal state detection model, the method also includes: evaluating the performance of the abnormal state detection model using a target evaluation method, and when the evaluation result of the performance of the abnormal state detection model does not meet the preset requirements, tuning the parameters of the abnormal state detection model; the target evaluation method includes at least one of the following: cross-validation evaluation method; AUC-ROC curve evaluation method; F1 score evaluation method.

[0037] Specifically, the embodiment of the present application can also evaluate the trained abnormal state detection model using cross-validation, AUC-ROC curve, F1 score, etc., and further improve the accuracy and robustness of the model through parameter tuning and model fusion.

[0038] Furthermore, in the abnormal state prediction and evaluation stage, data collection is performed every minute, and the prediction model is updated in real time to ensure the real-time prediction. Real-time rolling alarms and periodic analysis notifications are generated to timely discover and handle abnormal states. Information such as equipment name, abnormal occurrence time, specific abnormal location and description are automatically summarized to provide a detailed abnormal state report. Based on the generated analysis results, decision support is provided to power grid dispatchers to help them take timely measures to ensure the safe and stable operation of the power grid.

[0039] Figure 2 A structural schematic diagram provided by an embodiment of the present application is shown, and the device 200 includes: an acquisition module 201, which is used to acquire operating data of the power equipment to be monitored from at least one data source; an extraction module 202, which is used to collaboratively extract the operating data of each data source to obtain fused features of the operating data of each data source; a detection module 203, which is used to input the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the power equipment, the abnormal state detection model includes a clustering model and a classification model, and the detection result of the abnormal state of the power equipment is a fusion value of the output result of the clustering model and the output result of the classification model.

[0040] The embodiment of the present application can obtain the operating data of the power equipment to be monitored from at least one data source, and extract the operating data of each data source in a coordinated manner, so as to achieve the purpose of processing the massive data of the power equipment and extracting features. The extracted fusion features are further detected for abnormal states through an abnormal state detection model. Thus, the abnormal state of the power equipment is detected, thereby improving the safety and stability of the power equipment.

[0041] Optionally, it also includes: a processing module, used to preprocess the operating data of the power equipment from each data source to obtain target operating data, the preprocessing including data cleaning, data conversion and data normalization; a determination module, used to determine the target storage method of the target operating data according to the type of the target operating data; a storage module, used to store the target operating data in the target storage method.

[0042] Optionally, the determination module is also used to determine the computing platform based on the amount of operating data of the power equipment from each data source and the processing requirements of the operating data; configure the cluster on the computing platform and determine the resource scheduling method of the cluster; deploy the abnormal state detection model in the cluster, and schedule the resources in the cluster in a resource scheduling manner so that the abnormal state detection model uses the resources in the cluster.

[0043] Optionally, the acquisition module 201 is also used to obtain historical operating data of the power equipment; mark the abnormal state data in the historical operating data to obtain negative samples, and determine the non-abnormal state data in the historical operating data as positive samples; input the positive samples and negative samples into the initial abnormal state detection model for iterative training to obtain the abnormal state detection model.

[0044] Optionally, the processing module is also used to evaluate the performance of the abnormal state detection model using a target evaluation method, and to tune the parameters of the abnormal state detection model when the evaluation result of the performance of the abnormal state detection model does not meet the preset requirements; the target evaluation method includes at least one of the following: cross-validation evaluation method; AUC-ROC curve evaluation method; F1 score evaluation method.

[0045] Optionally, the detection module 203 is also used to input the fused features into the clustering model and the classification model of the abnormal state detection model respectively; obtain the first prediction value of the abnormal state when the fused features output by the clustering model are obtained, and obtain the second prediction value of the abnormal state when the fused features output by the classification model are obtained; weight the first prediction value and the second prediction value to obtain a third prediction value of the abnormal state detection model, the third prediction value represents the probability that the power equipment is in an abnormal state; when the third prediction value is greater than the first threshold value, determine that the state of the power equipment is an abnormal state.

[0046] Optionally, the detection module 203 is also used to input the fusion features into the clustering model for clustering to obtain the cluster centers of all clusters; calculate the distance from each sample point to the cluster center, and mark the number of the cluster to which each sample point belongs; input the number of the cluster to which each sample point belongs and the distance from each sample point to the cluster center into the classification model; obtain the fourth prediction value output by the classification model that each sample point is abnormal, the fourth prediction value represents the probability that each sample point is abnormal; when the fourth prediction value is greater than the second threshold value, determine that the state of the power equipment is an abnormal state.

[0047] The device 200 provided in the embodiment of the present application can execute each method in the foregoing method embodiment and realize the functions and beneficial effects of each method in the foregoing method embodiment, which will not be repeated here.

[0048] Figure 3 The hardware structure diagram of the electronic device provided by the embodiment of the present application is shown. Referring to the figure, at the hardware level, the electronic device includes a processor, and optionally, an internal bus, a network interface, and a memory. Among them, the memory may include a memory, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage, etc. Of course, the electronic device may also include hardware required for other services.

[0049] The processor, network interface and memory can be interconnected through an internal bus, which can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0050] The memory is used to store programs. Specifically, the program may include program code, and the program code includes computer operation instructions. The memory may include internal memory and non-volatile memory, and provides instructions and data to the processor.

[0051] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a device for locating the specified user at the logical level. The processor executes the program stored in the memory and is specifically used to execute: Figure 1 The method disclosed in the illustrated embodiment implements the functions and beneficial effects of each method in the previous method embodiments, which will not be described in detail here.

[0052] The above application Figure 1The method disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0053] The electronic device can also execute each method in the foregoing method embodiments and realize the functions and beneficial effects of each method in the foregoing method embodiments, which will not be described in detail here.

[0054] Of course, in addition to software implementation methods, the electronic device of the present application does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc. That is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0055] The present application also provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple application programs, the electronic device executes Figure 1 The method disclosed in the illustrated embodiment implements the functions and beneficial effects of each method in the previous method embodiments, which will not be described in detail here.

[0056] Among them, the computer readable storage medium includes a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0057] Furthermore, an embodiment of the present application also provides a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, the following process is implemented: Figure 1 The method disclosed in the illustrated embodiment implements the functions and beneficial effects of each method in the previous method embodiments, which will not be described in detail here.

[0058] In short, the above are only preferred embodiments of the present application and are not intended to limit the protection scope of the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0059] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0060] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0061] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0062] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

Claims

1. A method for monitoring abnormal state of electric power equipment, characterized in that: include: Acquire operation data of the electric power equipment to be monitored from at least one data source; Collaboratively extracting the operating data of each of the data sources to obtain fusion features of the operating data of each of the data sources; The fused features are input into an abnormal state detection model for abnormal state detection to obtain a detection result of the abnormal state of the power equipment. The abnormal state detection model includes a clustering model and a classification model. The detection result of the abnormal state of the power equipment is a fusion value of an output result of the clustering model and an output result of the classification model.

2. The abnormal state monitoring method of electric power equipment according to claim 1, characterized in that: After acquiring the operating data of the power equipment to be monitored from at least one data source, the method further includes: Preprocessing the operation data of the power equipment of each data source to obtain target operation data, wherein the preprocessing includes data cleaning, data conversion and data normalization; determining a target storage mode of the target operating data according to the type of the target operating data; The target operation data is stored in the target storage manner.

3. The abnormal state monitoring method of electric power equipment according to claim 1, characterized in that: Before inputting the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the electric power equipment, the method further includes: Determine a computing platform according to the amount of operation data of the power equipment of each data source and the processing requirements of the operation data; Configuring a cluster on the computing platform and determining a resource scheduling method for the cluster; The abnormal state detection model is deployed in the cluster, and resources in the cluster are scheduled in the resource scheduling manner so that the abnormal state detection model uses the resources in the cluster.

4. The abnormal state monitoring method of electric power equipment according to claim 1, characterized in that: Before inputting the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the electric power equipment, the method further includes: Obtain historical operating data of power equipment; Marking abnormal state data in the historical operation data to obtain negative samples, and determining non-abnormal state data in the historical operation data as positive samples; The positive samples and the negative samples are input into an initial abnormal state detection model for iterative training to obtain the abnormal state detection model.

5. The abnormal state monitoring method of electric power equipment according to claim 4, characterized in that: After inputting the positive samples and the negative samples into the initial abnormal state detection model for iterative training to obtain the abnormal state detection model, the method further includes: Evaluating the performance of the abnormal state detection model using a target evaluation method, and if the evaluation result of the performance of the abnormal state detection model does not meet the preset requirements, optimizing the parameters of the abnormal state detection model; The target assessment method includes at least one of the following: Cross-validation evaluation method; AUC-ROC curve evaluation method; F1 score evaluation method.

6. The abnormal state monitoring method of electric power equipment according to any one of claims 1 to 5, characterized in that: The step of inputting the fusion feature into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the power equipment includes: Inputting the fusion features into the clustering model and the classification model of the abnormal state detection model respectively; Acquire the first prediction value of the abnormal state of the fused feature output by the clustering model, and acquire the second prediction value of the abnormal state of the fused feature output by the classification model; Weighting the first prediction value and the second prediction value to obtain a third prediction value of the abnormal state detection model, wherein the third prediction value represents the probability that the power device is in an abnormal state; When the third prediction value is greater than the first threshold, it is determined that the state of the electric power equipment is an abnormal state.

7. The abnormal state monitoring method of electric power equipment according to any one of claims 1 to 5, characterized in that: The step of inputting the fusion feature into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the power equipment includes: Inputting the fused features into the clustering model for clustering to obtain cluster centers of all clusters; Calculate the distance from each sample point to the center of the cluster, and mark the number of the cluster to which each sample point belongs; Inputting the number of the cluster to which each sample point belongs and the distance from each sample point to the center of the cluster into the classification model; Obtaining a fourth prediction value output by the classification model that each of the sample points is abnormal, wherein the fourth prediction value represents a probability that each of the sample points is abnormal; When the fourth prediction value is greater than the second threshold, it is determined that the state of the power equipment is an abnormal state.

8. An abnormal state monitoring device for electric power equipment, characterized in that: include: An acquisition module, used for acquiring operation data of the power equipment to be monitored from at least one data source; An extraction module, used to collaboratively extract the operation data of each of the data sources to obtain fusion features of the operation data of each of the data sources; A detection module is used to input the fused features into an abnormal state detection model to perform abnormal state detection to obtain a detection result of the abnormal state of the power equipment. The abnormal state detection model includes a clustering model and a classification model. The detection result of the abnormal state of the power equipment is a fusion value of the output result of the clustering model and the output result of the classification model.

9. The abnormal state monitoring device for electric power equipment according to claim 8, characterized in that: Also includes: A processing module, used for preprocessing the operation data of the power equipment of each data source to obtain target operation data, wherein the preprocessing includes data cleaning, data conversion and data normalization; A determination module, configured to determine a target storage mode of the target operation data according to a type of the target operation data; A storage module is used to store the target operation data in the target storage mode.

10. An electronic device, characterized in that: The abnormal state monitoring device of the electric power equipment comprises: a processor and a memory; wherein the memory is used to store a computer program that can be run on the processor; the processor is used to execute the program stored in the memory to implement the steps of the abnormal state monitoring method of the electric power equipment as described in any one of claims 1-7.

Images