Power equipment anomaly identification method and device, electronic device and medium

Abstract

This disclosure discloses a method, apparatus, electronic device, and medium for identifying power equipment anomalies. One specific implementation of the method includes: for each piece of power equipment data in a power equipment data acquisition set, performing the following anomaly identification steps: extracting features from each power equipment image in the power equipment image group included in the power equipment data acquisition to generate power equipment image features; inputting the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain power equipment appearance anomaly detection results; performing anomaly detection on the power equipment operation dataset included in the power equipment data acquisition to generate operation data anomaly detection results; and, in response to determining that the power equipment appearance anomaly detection results meet the appearance anomaly alarm conditions, sending equipment anomaly alarm information to the associated power equipment maintenance terminal. This implementation improves the efficiency and accuracy of power equipment operation data detection.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computers, and more specifically to methods, apparatus, electronic devices, and media for identifying anomalies in power equipment. Background Technology

[0002] With the continuous growth of global energy demand and the increasing severity of environmental problems, smart grid technology, as a key technology for achieving efficient energy utilization and optimized allocation, has become an important direction for the development of global power systems. Currently, the common method for monitoring the operation of power equipment is for technicians to periodically inspect the equipment within the power grid area. However, this method reduces the efficiency of power equipment monitoring, fails to promptly identify abnormal equipment, and can easily lead to potential power safety hazards.

[0003] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure provide methods, apparatus, electronic devices, and computer-readable media for identifying power equipment anomalies in order to solve one or more of the technical problems mentioned in the background section above.

[0006] In a first aspect, some embodiments of this disclosure provide a method for identifying power equipment anomalies. The method includes: in response to receiving a power equipment data anomaly identification request, acquiring a set of power equipment collected information sent by a power equipment information collection terminal, wherein the power equipment collected information in the power equipment collected information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset; for each power equipment collected information in the power equipment collected information set, performing the following anomaly identification steps: extracting features from each power equipment image in the power equipment image group included in the power equipment collected information to generate power equipment image features, obtaining a power equipment image feature group; inputting the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result; performing anomaly detection on the power equipment operation dataset included in the power equipment collected information to generate an operation data anomaly detection result; in response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, sending equipment anomaly alarm information to an associated power equipment maintenance terminal.

[0007] Secondly, some embodiments of this disclosure provide a power equipment anomaly identification device, which includes: an acquisition unit configured to acquire a power equipment collection information set sent by a power equipment information collection terminal in response to receiving a power equipment data anomaly identification request, wherein the power equipment collection information in the power equipment collection information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset; and an identification unit configured to perform the following anomaly identification steps for each power equipment collection information in the power equipment collection information set: extracting features from each power equipment image in the power equipment image group included in the power equipment collection information to generate power equipment image features, thereby obtaining a power equipment image feature group; inputting the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result; performing anomaly detection on the power equipment operation dataset included in the power equipment collection information to generate an operation data anomaly detection result; and sending equipment anomaly alarm information to an associated power equipment maintenance terminal in response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0010] The above embodiments of this disclosure have the following beneficial effects: the power equipment anomaly identification method of some embodiments of this disclosure improves the efficiency and accuracy of power equipment operation data detection. First, in response to receiving a power equipment data anomaly identification request, a power equipment collection information set sent by a power equipment information collection terminal is acquired. The power equipment collection information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset. Then, for each power equipment collection information in the power equipment collection information set, the following anomaly identification steps are performed: feature extraction is performed on each power equipment image in the power equipment image group included in the power equipment collection information to generate power equipment image features, resulting in a power equipment image feature group; the power equipment image feature group is input into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result. Thus, whether the appearance of the power equipment is abnormal is detected, thereby avoiding damage to internal hardware facilities. Finally, anomaly detection is performed on the power equipment operation dataset, which includes the collected information from the power equipment, to generate operation data anomaly detection results. In response to determining that the power equipment appearance anomaly detection results meet the appearance anomaly alarm conditions and / or the operation data anomaly detection results meet the data anomaly alarm conditions, equipment anomaly alarm information is sent to the associated power equipment maintenance terminal. Thus, the operation data anomaly detection model can detect the types of operation data anomalies for each power equipment. This reduces the amount of data processing during model prediction and minimizes the waste of computing resources. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of the power equipment anomaly identification method according to the present disclosure;

[0013] Figure 2 This is a flowchart of some embodiments of the power equipment anomaly identification device according to the present disclosure;

[0014] Figure 3This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0015] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0016] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0017] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0018] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0019] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0020] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] Figure 1 This is a flowchart of some embodiments of the power equipment anomaly identification method according to the present disclosure. A flowchart 100 of some embodiments of the power equipment anomaly identification method according to the present disclosure is shown. This power equipment anomaly identification method, applied to an intelligent robot, includes the following steps:

[0022] Step 101: In response to receiving the power equipment data anomaly identification request, obtain the power equipment collection information set sent by the power equipment information collection terminal.

[0023] In some embodiments, the executing entity of the power equipment anomaly identification method (e.g., a computing device) can, in response to receiving a power equipment data anomaly identification request, acquire a power equipment collection information set sent by a power equipment information acquisition terminal. The power equipment collection information set includes: power equipment name, power equipment location, power equipment image set, and power equipment operation dataset. The power equipment name can represent the name of the power equipment whose information is being collected. For example, the power equipment can be a transformer, circuit breaker, electricity meter, etc. The power equipment can also be a power generation device. Power generation devices mainly include power plant boilers, steam turbines, gas turbines, water turbines, generators, transformers, etc. The power equipment data anomaly identification request can refer to a request to identify anomalies in the power equipment collection information. That is, to determine whether there is abnormal data in the power equipment collection information, thereby determining whether the power equipment is abnormal (e.g., external damage, abnormal data operation). The power equipment information acquisition terminal can refer to an edge data acquisition terminal that collects the operation information of various power equipment. The power equipment location can represent the location of the power equipment whose information is being collected. The power equipment image set can represent various external images (front view, side view, etc.) of a power equipment. A power equipment operation dataset can represent various operational data of power equipment within a preset time period. For example, a power equipment operation dataset may include: electrical parameters, temperature parameters, hydrogen system parameters, and internal cooling water system parameters. Electrical parameters may include: active power, reactive power, stator voltage, stator current, rotor voltage, and rotor current. Temperature parameters may include: stator winding temperature, stator water inlet pipe temperature, stator core temperature, bearing temperature, and bearing return oil temperature. Hydrogen system parameters may include: cold hydrogen temperature, hot hydrogen temperature, hydrogen pressure, oil-hydrogen pressure difference, hydrogen-side oil and air-side oil pressure difference, and hydrogen quality (humidity, purity). Internal cooling water system parameters may include: internal cooling water inlet (outlet) temperature, internal cooling water inlet (outlet) pressure, water flow rate, and internal cooling water quality (pH value, conductivity).

[0024] Step 102: For each piece of power equipment data collected in the power equipment data collection set, perform the following anomaly identification steps:

[0025] Step 1021: Extract features from each power equipment image in the power equipment image group included in the power equipment acquisition information to generate power equipment image features, thereby obtaining a power equipment image feature group.

[0026] In some embodiments, the aforementioned executing entity may perform feature extraction on each power equipment image in the power equipment image group included in the power equipment acquisition information to generate power equipment image features, thereby obtaining a power equipment image feature group.

[0027] In practice, the aforementioned executing entity can extract features from each power equipment image in the power equipment image group included in the power equipment collection information through the following steps to generate power equipment image features:

[0028] The first step is to perform image denoising processing on the power equipment image to generate a denoised power equipment image. This can be done using an image denoising algorithm. For example, the image denoising algorithm could be mean filtering, statistical sorting filtering, or non-local mean denoising.

[0029] The second step involves inputting the denoised power equipment image into a pre-trained power equipment image feature extraction model to obtain the power equipment image features. The power equipment image feature extraction model can be a 2D convolutional neural network model. For example, the power equipment image feature extraction model can be a fully connected network model. The power equipment image features can characterize the semantic content of the power equipment in the power equipment image.

[0030] The power equipment image feature extraction model can be trained through the following steps:

[0031] The first step involves fine-tuning the initial power equipment image feature extraction model based on a pre-trained large model to generate a fine-tuned power equipment image feature extraction model. The pre-trained large model can be a large model specifically designed for power equipment image feature extraction. For example, the execution entity can first obtain an initial power equipment image sample set. Then, using the pre-trained large model, a set of power equipment feature information labels corresponding to the initial power equipment image sample set is generated. Next, using the power equipment feature information label set as the label set and the initial power equipment image sample set as the training sample set, the initial power equipment image feature extraction model is trained to generate the fine-tuned power equipment image feature extraction model.

[0032] The second step is to obtain a sample set of power equipment images. This sample set includes both sample power equipment images and real equipment images. The sample power equipment images can be input images related to the appearance of the power equipment. The real equipment images can be images of the actual surface of the power equipment, captured in photographs.

[0033] The third step is to select a target power equipment image sample from the power equipment image sample set. One power equipment image sample can be randomly selected from the power equipment image sample set as the target power equipment image sample.

[0034] Fourth, for the target power equipment image samples, perform the following training steps:

[0035] 1. Input the sample power equipment images, including the target power equipment image samples, into the fine-tuned power equipment image feature extraction model to generate sample power equipment image features.

[0036] 2. Input the features of the sample power equipment images into the initial power equipment image generation model to generate predicted power equipment images. The initial power equipment image generation model can be a neural network model for generating power equipment images that has not yet been fully trained. For example, the initial power equipment image generation model can be a multi-layered convolutional layer.

[0037] 3. Based on the predicted image of the power equipment and the corresponding real image of the equipment, determine whether the model training has ended. As an example, firstly, the execution entity can use the semantic content difference between the predicted image of the power equipment and the corresponding real image of the equipment as difference information. Then, the difference information is input into the loss function to generate image loss information. Next, in response to determining that the loss information in the image loss information sequence tends to a predetermined value, the model training is determined to have ended. In response to determining that the loss information in the image loss information sequence does not tend to a predetermined value, the model training is determined not to have ended.

[0038] 4. Upon completion of training, the fine-tuned power equipment image feature extraction model is identified as the fully trained power equipment image feature extraction model.

[0039] Fifth, in response to the training not being completed, the fine-tuned power equipment image feature extraction model and the initial power equipment image generation model are trained based on the predicted power equipment image and the corresponding real equipment image to generate the trained power equipment image feature extraction model and the trained power equipment image generation model. For example, the aforementioned execution entity can use backpropagation to train the fine-tuned power equipment image feature extraction model and the initial power equipment image generation model based on the image loss information of the predicted power equipment image and the corresponding real equipment image to generate the trained power equipment image feature extraction model and the trained power equipment image generation model.

[0040] The sixth step involves reselecting target power equipment image samples from the unselected power equipment image samples, using the trained power equipment image feature extraction model as the fine-tuned power equipment image feature extraction model, and using the trained power equipment image generation model as the power equipment image generation model, and then performing the training steps again.

[0041] Step 1022: Input the power equipment image feature group into the pre-trained power equipment appearance anomaly detection model to obtain the power equipment appearance anomaly detection result.

[0042] In some embodiments, the aforementioned executing entity can input the power equipment image feature set into a pre-trained power equipment appearance anomaly detection model to obtain the power equipment appearance anomaly detection result. The power equipment appearance anomaly detection model can refer to a pre-trained neural network model that takes the power equipment image feature set as input and outputs the power equipment appearance anomaly detection result. The power equipment appearance anomaly detection result can indicate whether the appearance of the power equipment corresponding to the power equipment image feature set is abnormal. For example, an appearance anomaly of power equipment can indicate rust / deformation / dentation on the surface of the power equipment. For example, the power equipment appearance anomaly detection model can refer to a convolutional neural network (CNN), a generative adversarial network (GAN), or a long short-term memory network (LSTM). Furthermore, the power equipment appearance anomaly detection model can refer to a model based on color feature analysis, a model based on texture feature analysis, or a model based on shape feature analysis.

[0043] Color feature analysis-based models: Color is one of the important features in object appearance detection. Common color feature analysis methods include color histograms, color filtering, and color segmentation. Color histograms generate a color distribution histogram by statistically analyzing the color values ​​of each pixel in an image. Color filtering detects specific colors in an image by defining different color filters. Color segmentation divides an image according to its color features.

[0044] Texture feature analysis-based models: Texture is another important feature in object appearance detection, used to describe the surface features of an object. Texture feature analysis methods are usually implemented by performing operations such as texture filtering, texture segmentation, and texture matching on the image. Texture filtering enhances or suppresses texture information in an image by defining texture filters, texture segmentation separates texture regions in an image, and texture matching performs object appearance detection and classification by comparing texture information in different images.

[0045] Shape feature analysis-based models: Shape is another important feature in object appearance detection, and it can be used for identification and classification by analyzing the shape features of objects. Shape feature analysis methods include edge detection, contour extraction, and shape description. Edge detection extracts the contours of objects by detecting edges in an image, contour extraction extracts the contours of objects in an image, and shape description mathematically describes the shape features of objects.

[0046] Step 1023: Perform anomaly detection on the power equipment operation dataset included in the power equipment collection information to generate anomaly detection results for operation data.

[0047] In some embodiments, the aforementioned executing entity can perform anomaly detection on the power equipment operation dataset included in the power equipment collection information to generate anomaly detection results. For example, the operation data of each power equipment can be monitored through a pre-defined power equipment operation index operation interval table to generate anomaly detection results. The anomaly detection results can indicate whether there is power equipment operation data in the power equipment operation dataset that is not within the corresponding power equipment operation index operation interval. One power equipment operation data corresponds to one power equipment operation index operation interval. The power equipment operation index operation interval can represent the normal operating parameter range of a certain power equipment operation index.

[0048] In practice, the aforementioned executing entity can input the power equipment operation dataset into a pre-trained power equipment operation data anomaly detection model to obtain power equipment operation data anomaly detection results.

[0049] The power equipment operation data anomaly detection model can be trained through the following steps:

[0050] The first step is to obtain a sample set of power equipment operation information. This sample includes sub-data sets of power equipment operation data. These sub-data sets can be data related to power operation collected by power consumption monitoring equipment within a preset time period. Power consumption monitoring equipment includes smart meters and power load collectors. This sub-data can be used to identify the types of operational anomalies.

[0051] The second step is to select a target power equipment operation information sample from the power equipment operation information sample set. One power equipment operation information sample can be randomly selected from the power equipment operation information sample set as the target power equipment operation information sample.

[0052] The third step involves identifying the operation anomaly types in the sub-data of each power equipment included in the target power equipment operation information sample, resulting in operation data anomaly type groups. For example, the above operation data anomaly types can represent abnormal operation data behavior types. These operation data anomaly types could be: operation data below a minimum threshold, or operation data above a maximum threshold.

[0053] The third step mentioned above may include the following sub-steps:

[0054] The first sub-step involves performing the following processing steps for each power equipment operation sub-data:

[0055] 1. Obtain the detection task information corresponding to the power equipment operation sub-data. For example, the above detection task information may be programming language code for detecting operational anomalies in the power equipment operation sub-data.

[0056] 2. Execute the detection task corresponding to the detection task information to perform type anomaly detection on the power equipment operation sub-data to generate a data anomaly type. For example, when the above-mentioned task detection information is a sequence of operational data correlation values, the above-mentioned execution entity can execute the detection task corresponding to the above-mentioned operational data correlation value sequence to perform power consumption anomaly detection on the above-mentioned power equipment operation sub-data and obtain the corresponding data anomaly type. The above-mentioned data anomaly type can be a representation of "power equipment operation sub-data anomaly" or a representation of "power equipment operation sub-data normal".

[0057] The second sub-step involves identifying data anomaly types that meet preset anomaly type conditions from among the various data anomaly types as operational data anomaly types, thus obtaining an operational data anomaly type group. Anomaly type conditions can be data anomaly types that characterize sub-data anomalies in the operation of power equipment.

[0058] The fourth step is to train the initial power equipment operation data anomaly detection model based on the power equipment operation sub-data group and the operation data anomaly type group included in the target power equipment operation information sample, so as to obtain the trained power equipment operation data anomaly detection model.

[0059] The fourth step mentioned above may include the following sub-steps:

[0060] The first sub-step involves performing the following processing steps for each power equipment operation sub-data in the power equipment operation sub-data group:

[0061] 1. The operation data anomaly type corresponding to the operation sub-data of the power equipment in the operation data anomaly type group is determined as the reference operation data anomaly type.

[0062] 2. The power equipment operation sub-data and the reference operation data anomaly types are merged into a power equipment operation sub-data sample.

[0063] The second sub-step involves selecting the target power equipment operation sub-data sample from the various power equipment operation sub-data samples. For example, one power equipment operation sub-data sample can be randomly selected from the various power equipment operation sub-data samples as the target power equipment operation sub-data sample.

[0064] The third sub-step involves inputting the power equipment operation sub-data included in the target power equipment operation sub-data sample into the initial power equipment operation data anomaly detection model to obtain the initial power equipment operation data anomaly detection result.

[0065] The fourth sub-step involves determining the loss value between the initial power equipment operating data anomaly detection result and the corresponding reference operating data anomaly type. For example, a preset damage function can be used to determine the loss value between the initial power equipment operating data anomaly detection result and the corresponding reference operating data anomaly type.

[0066] The fifth sub-step involves determining the initial power equipment operation data anomaly detection model as the trained power equipment operation data anomaly detection model in response to determining that the loss value is less than or equal to a preset loss value.

[0067] Therefore, the anomaly detection model for power equipment operation data can be used to detect the types of anomalies in the operation data of various power equipment. This reduces the amount of data processing during the model prediction process and reduces the waste of computing resources.

[0068] Step 1024: In response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, send equipment anomaly alarm information to the associated power equipment maintenance terminal.

[0069] In some embodiments, the aforementioned executing entity may, in response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operating data anomaly detection result meets the data anomaly alarm condition, send equipment anomaly alarm information to the associated power equipment maintenance terminal. For example, the appearance anomaly alarm condition may be: the power equipment appearance anomaly detection result indicates that the power equipment's appearance is abnormal. The data anomaly alarm condition may be: the operating data anomaly detection result indicates that the power equipment's operating data is abnormal. The power equipment maintenance terminal may refer to a terminal that performs maintenance monitoring of power equipment. The equipment anomaly alarm information may be information indicating that the power equipment is abnormal and requires maintenance.

[0070] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a power equipment anomaly identification device, which are similar to... Figure 1 Corresponding to the method embodiments shown, this power equipment anomaly identification device can be specifically applied to various electronic devices.

[0071] like Figure 2As shown, the power equipment anomaly identification device 200 in some embodiments includes: an acquisition unit 201 and an identification unit 202. The acquisition unit 201 is configured to, in response to receiving a power equipment data anomaly identification request, acquire a set of power equipment collected information sent by a power equipment information acquisition terminal, wherein the power equipment collected information in the power equipment collected information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset; the identification unit 202 is configured to, for each power equipment collected information in the power equipment collected information set, perform the following anomaly identification steps: extract features from each power equipment image in the power equipment image group included in the power equipment collected information to generate power equipment image features, thereby obtaining a power equipment image feature group; input the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result; perform anomaly detection on the power equipment operation dataset included in the power equipment collected information to generate an operation data anomaly detection result; and, in response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, send equipment anomaly alarm information to the associated power equipment maintenance terminal.

[0072] It is understandable that the units recorded in the power equipment anomaly identification device 200 are related to the reference. Figure 1 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to the power equipment anomaly identification device 200 and the units contained therein, and will not be repeated here.

[0073] The following is for reference. Figure 3 This document illustrates a structural schematic of an electronic device (e.g., a computing device) 300 suitable for implementing some embodiments of the present disclosure. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0074] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0075] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0076] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 309, or installed from a storage device 308, or installed from a ROM 302. When the computer program is executed by the processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0077] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0078] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0079] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs. When the electronic device executes one or more of these programs, the electronic device causes the following actions: In response to receiving a request for anomaly identification of power equipment data, it acquires a set of power equipment acquisition information sent by a power equipment information acquisition terminal, wherein the power equipment acquisition information in the power equipment acquisition information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset; for each piece of power equipment acquisition information in the power equipment acquisition information set, it performs the following anomaly identification steps: extracting features from each power equipment image in the power equipment image group included in the power equipment acquisition information to generate power equipment image features, thus obtaining a power equipment image feature group; inputting the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result; performing anomaly detection on the power equipment operation dataset included in the power equipment acquisition information to generate an operation data anomaly detection result; and in response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, it sends equipment anomaly alarm information to the associated power equipment maintenance terminal.

[0080] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0081] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0082] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit and an identification unit. The names of these units do not necessarily limit the specific unit; for example, the acquisition unit may also be described as "a unit that, in response to receiving a request for abnormal identification of power equipment data, acquires a set of power equipment data collected by a power equipment information acquisition terminal."

[0083] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0084] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for identifying abnormalities in power equipment, comprising: In response to receiving a request for abnormal identification of power equipment data, the system obtains a set of power equipment data collection information sent by the power equipment information collection terminal. The power equipment data collection information in the power equipment data collection set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset. The power equipment image group represents various appearance images of a power equipment, including a front view and a side view. For each piece of power equipment data collected in the power equipment data collection set, the following anomaly identification steps are performed: Feature extraction is performed on each power equipment image in the power equipment image group included in the power equipment acquisition information to generate power equipment image features, resulting in a power equipment image feature group. This includes: performing image denoising processing on the power equipment images to generate denoised power equipment images; inputting the denoised power equipment images into a pre-trained power equipment image feature extraction model to obtain power equipment image features. The power equipment image feature extraction model is trained through the following steps: fine-tuning the initial power equipment image feature extraction model based on a pre-trained large model to generate a fine-tuned power equipment image feature extraction model; and obtaining a power equipment image sample set. This includes sample power equipment images and real equipment images; target power equipment image samples are selected from the power equipment image sample set; for the target power equipment image samples, the following training steps are performed: inputting the sample power equipment images included in the target power equipment image samples into a fine-tuned power equipment image feature extraction model to generate sample power equipment image features; inputting the sample power equipment image features into an initial power equipment image generation model to generate predicted power equipment images; determining whether the model training is complete based on the predicted power equipment images and the corresponding real equipment images; in response to the end of training, determining the fine-tuned power equipment image feature extraction model as the trained power equipment image feature extraction model; The image feature set of the power equipment is input into a pre-trained power equipment appearance anomaly detection model to obtain the power equipment appearance anomaly detection result; Obtain a sample set of power equipment operation information, wherein the power equipment operation information sample includes: power equipment operation sub-data group; Select the target power equipment operation information sample from the power equipment operation information sample set; The operation anomaly type identification process is performed on each sub-data of the target power equipment operation information sample to obtain operation data anomaly type groups; Based on the power equipment operation sub-data group and the operation data anomaly type group included in the target power equipment operation information sample, an initial power equipment operation data anomaly detection model is trained to obtain a trained power equipment operation data anomaly detection model. For each power equipment operation sub-data in the power equipment operation sub-data group, the following processing steps are performed: The operation data anomaly type corresponding to the power equipment operation sub-data in the operation data anomaly type group is determined as a reference operation data anomaly type; the power equipment operation sub-data and the reference operation data anomaly type are merged into a power equipment operation sub-data sample; a target power equipment operation sub-data sample is selected from each power equipment operation sub-data sample; the power equipment operation sub-data included in the target power equipment operation sub-data sample is input into the initial power equipment operation data anomaly detection model to obtain an initial power equipment operation data anomaly detection result; the loss value between the initial power equipment operation data anomaly detection result and the corresponding reference operation data anomaly type is determined; in response to determining that the loss value is less than or equal to a preset loss value, the initial power equipment operation data anomaly detection model is determined as a trained power equipment operation data anomaly detection model. Anomaly detection is performed on the power equipment operation dataset, which includes the power equipment collected information, to generate operation data anomaly detection results; In response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, an equipment anomaly alarm message is sent to the associated power equipment maintenance terminal.

2. The method according to claim 1, wherein, The step of performing anomaly detection on the power equipment operation dataset, which includes the information collected from the power equipment, to generate anomaly detection results for the operation data includes: The power equipment operation dataset is input into a pre-trained power equipment operation data anomaly detection model to obtain power equipment operation data anomaly detection results.

3. The method according to claim 2, wherein, The process of identifying operational anomaly types in the operational sub-data of each power equipment included in the target power equipment operational information sample yields operational data anomaly type groups, including: For each power equipment operation sub-data, perform the following processing steps: Obtain detection task information corresponding to the operation sub-data of the power equipment; Execute the detection task corresponding to the detection task information to perform type anomaly detection on the power equipment operation sub-data to generate data anomaly types; The data anomaly types that meet the preset anomaly type conditions among the various data anomaly types are identified as the running data anomaly types, thus obtaining the running data anomaly type group.

4. A device for identifying abnormalities in power equipment, comprising: The acquisition unit is configured to, in response to receiving a power equipment data anomaly identification request, acquire a power equipment acquisition information set sent by the power equipment information acquisition terminal. The power equipment acquisition information set includes: power equipment name, power equipment location, power equipment image group, and power equipment operation dataset. The power equipment image group represents various appearance images of a power equipment, including a front view and a side view. The identification unit is configured to perform the following anomaly identification steps for each power equipment acquisition information in the power equipment acquisition information set: extracting features from each power equipment image in the power equipment image group included in the power equipment acquisition information to generate power equipment image features, thus obtaining a power equipment image feature group; inputting the power equipment image feature group into a pre-trained power equipment appearance anomaly detection model to obtain a power equipment appearance anomaly detection result; acquiring a power equipment operation information sample set, wherein the power equipment operation information sample includes: a power equipment operation sub-data group; selecting a target power equipment operation information sample from the power equipment operation information sample set; and performing anomaly identification on the target power equipment operation information sample. Each sub-data item of the power equipment operation is processed for operation anomaly type identification to obtain an operation data anomaly type group; based on the power equipment operation sub-data group and the operation data anomaly type group included in the target power equipment operation information sample, an initial power equipment operation data anomaly detection model is trained to obtain a trained power equipment operation data anomaly detection model. For each sub-data item of the power equipment operation sub-data group, the following processing steps are performed: the operation data anomaly type corresponding to the sub-data item in the operation data anomaly type group is determined as a reference operation data anomaly type; the sub-data item and the reference operation data anomaly type are merged into a single power equipment operation data anomaly detection model. Sub-data samples; selecting a target power equipment operation sub-data sample from various power equipment operation sub-data samples; inputting the power equipment operation sub-data included in the target power equipment operation sub-data sample into an initial power equipment operation data anomaly detection model to obtain an initial power equipment operation data anomaly detection result; determining the loss value between the initial power equipment operation data anomaly detection result and the corresponding reference operation data anomaly type; in response to determining that the loss value is less than or equal to a preset loss value, determining the initial power equipment operation data anomaly detection model as a trained power equipment operation data anomaly detection model; performing anomaly detection on the power equipment operation dataset included in the power equipment collected information to generate operation data anomaly detection. Result; In response to determining that the power equipment appearance anomaly detection result meets the appearance anomaly alarm condition and / or the operation data anomaly detection result meets the data anomaly alarm condition, an equipment anomaly alarm message is sent to the associated power equipment maintenance terminal, including: performing image denoising processing on the power equipment image to generate a denoised power equipment image; inputting the denoised power equipment image into a pre-trained power equipment image feature extraction model to obtain power equipment image features, wherein the power equipment image feature extraction model is trained through the following steps: fine-tuning the initial power equipment image feature extraction model based on the pre-trained large model to generate a fine-tuned power equipment image feature extraction model; obtaining a power equipment image sample set;The power equipment image samples include sample power equipment images and real equipment images. Target power equipment image samples are selected from the set of power equipment image samples. For the target power equipment image samples, the following training steps are performed: inputting the sample power equipment images included in the target power equipment image samples into a fine-tuned power equipment image feature extraction model to generate sample power equipment image features; inputting the sample power equipment image features into an initial power equipment image generation model to generate predicted power equipment images; determining whether the model training is complete based on the predicted power equipment images and the corresponding real equipment images; in response to the completion of training, determining the fine-tuned power equipment image feature extraction model as the successfully trained power equipment image feature extraction model.

5. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-3.

6. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1-3.

Images