A power equipment diagnosis method and system based on artificial intelligence

Abstract

The present invention discloses an artificial intelligence-based power equipment diagnosis method and system, which belongs to the technical field of power equipment fault diagnosis, including collecting operation data and image data of target power equipment, and performing preprocessing operations on the operation data and image data to generate target operation data and target image data respectively, constructing an artificial intelligence-based power equipment fault diagnosis model, and obtaining historical fault data and normal operation data of the target power equipment to train the power equipment fault diagnosis model to obtain the target power equipment fault diagnosis model, inputting the target operation data and target image data into the target power equipment fault diagnosis model for fault judgment, outputting the fault judgment result, and providing corresponding fault handling suggestions according to the fault judgment result, thereby improving the efficiency and accuracy of power equipment fault diagnosis, handling fault situations in a timely manner, and having a high level of intelligence, achieving rapid response to faults while ensuring the diagnosis effect.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment fault diagnosis, and in particular to an artificial intelligence-based power equipment fault diagnosis method and system thereof. Background Art

[0002] As the infrastructure of modern society, the stable operation of the power system is crucial to ensuring social and economic activities. With the continuous expansion of the scale of the power system and the increasing complexity of its structure, the fault diagnosis of power equipment faces unprecedented challenges. In recent years, with the rapid development of artificial intelligence technology, its application in the field of power equipment fault diagnosis has gradually become a research hotspot. Artificial intelligence, especially deep learning and machine learning technology, can automatically extract features from massive equipment operation data, establish the association between fault mode and equipment status, and realize intelligent diagnosis of power equipment faults. This method can not only improve the accuracy and efficiency of diagnosis, but also monitor the equipment status in real time, predict potential faults, and provide a scientific basis for the maintenance and management of power systems.

[0003] In the prior art, traditional fault diagnosis methods mainly rely on manual inspections and expert experience. This method is not only inefficient, but also often difficult to guarantee accuracy and timeliness when facing complex faults. Especially in large-scale power systems, there are many types of equipment and the operating environment is complex, and the limitations of manual diagnosis are more obvious. Since the operating data of power equipment often has the characteristics of high dimension, nonlinearity, and time-varying, the real-time and accuracy requirements of power equipment fault diagnosis are high, and the operating environment of the power system is complex and changeable, resulting in the poor adaptability and generalization ability of existing fault diagnosis methods, and the low level of intelligence, which cannot achieve rapid response to faults while ensuring the diagnosis effect.

[0004] Therefore, how to improve the efficiency and accuracy of power equipment fault diagnosis, handle fault conditions in a timely manner, achieve better adaptability and generalization capabilities, and have a high level of intelligence, while ensuring the diagnostic effect and achieving rapid response to faults has become a problem that needs to be solved urgently. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes an artificial intelligence-based power equipment diagnosis method and system, which selects target power equipment, collects equipment data of the target power equipment, the equipment data includes operation data and image data, obtains the operation data and image data, and performs preprocessing operations on the operation data and image data to generate target operation data and target image data respectively, constructs an artificial intelligence-based power equipment fault diagnosis model, and obtains historical fault data and normal operation data of the target power equipment to train the power equipment fault diagnosis model to obtain the target power equipment fault diagnosis model, obtains target operation data and target image data, and inputs the target operation data and target image data into the target power equipment fault diagnosis model for fault judgment to obtain a fault judgment result, outputs the fault judgment result, and provides corresponding fault handling suggestions based on the fault judgment result, thereby improving the efficiency and accuracy of power equipment fault diagnosis, handling fault conditions in a timely manner, achieving good adaptability and generalization ability, and having a high level of intelligence, while ensuring the diagnosis effect while achieving rapid response to faults.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solutions:

[0007] An artificial intelligence-based power equipment diagnosis method of the present invention comprises:

[0008] Selecting a target electric power device and collecting device data of the target electric power device, wherein the device data includes operation data and image data;

[0009] Acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively;

[0010] Constructing an artificial intelligence-based power equipment fault diagnosis model, and acquiring historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model, thereby obtaining a target power equipment fault diagnosis model;

[0011] Acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model to perform fault diagnosis, and obtain a fault diagnosis result;

[0012] The fault judgment result is output, and corresponding fault handling suggestions are provided according to the fault judgment result.

[0013] A further improvement of the present invention is that: the equipment parameters of the target electric equipment are monitored in real time by the target sensor on the target electric equipment to obtain corresponding operation data.

[0014] A further improvement of the present invention is that: the external device image data and the operating environment image data of the target power equipment are collected by a target camera device to obtain corresponding image data, wherein the image data includes the external device image data and the operating environment image data.

[0015] A further improvement of the present invention is that the preprocessing operation includes data cleaning, normalization, dimensionality reduction processing and enhancement processing.

[0016] A further improvement of the present invention is that the construction of an artificial intelligence-based power equipment fault diagnosis model, and obtaining the historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model to obtain the target power equipment fault diagnosis model, includes the following operations:

[0017] The historical fault data and normal operation data of the target power equipment are subjected to feature extraction by the following feature extraction algorithm to obtain historical fault features and normal operation features:

[0018] The extraction process of the historical fault features is as follows:

[0019] Feature extraction is performed on the historical fault data to obtain a historical fault matrix, and the expression of the historical fault matrix is:

[0020]

[0021] In the formula, represents the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of

[0022] Calculate the extreme value of the historical fault matrix, the expression of the extreme value of the historical fault matrix is:

[0023]

[0024] In the formula, represents the historical fault matrix; Represents the extreme value of the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of

[0025] Select the extreme value of historical fault matrix The corresponding pixel point is used as a historical fault feature point, and a stable historical fault feature point is determined. The historical fault feature direction is determined according to the wavelet feature of the circular neighborhood of the stable historical fault feature point, and a 4*4 rectangular area around the stable historical fault feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable historical fault feature point are counted to obtain the historical fault feature;

[0026] The extraction process of the normal operation characteristics is as follows:

[0027] Feature extraction is performed on the normal operation data to obtain a normal operation matrix, and the expression of the normal operation matrix is:

[0028]

[0029] In the formula, represents the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of

[0030] Calculate the extreme value of the normal operation matrix, the expression of the extreme value of the normal operation matrix is:

[0031]

[0032] In the formula, represents the normal operation matrix; Indicates the extreme value of the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of

[0033] Select the extreme value of normal operation matrix The corresponding pixel point is used as a normal operation feature point, and a stable normal operation feature point is determined. The normal operation feature direction is determined according to the wavelet feature of the circular neighborhood of the stable normal operation feature point, and a 4*4 rectangular area around the stable normal operation feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable normal operation feature point are counted to obtain the normal operation feature;

[0034] The historical fault data and normal operation data of the target power equipment are subjected to feature extraction by a feature extraction algorithm to generate the historical fault features and normal operation features.

[0035] A further improvement of the present invention is that: the difference between the historical fault characteristics and the normal operating characteristics is compared by the following difference verification algorithm to obtain a difference comparison result, that is, the power equipment fault diagnosis model is trained according to the historical fault data and normal operating data of the target power equipment to obtain the target power equipment fault diagnosis model:

[0036] Set the historical fault characteristics to And the normal operating characteristics are ;

[0037] The historical fault characteristics and normal operating characteristics After L2 regularization, the expression of the dot product is:

[0038]

[0039] In the formula, Indicates historical fault characteristics; Indicates normal operating characteristics; Indicates historical fault characteristics and normal operating characteristics The dot product after L2 regularization; Indicates historical fault characteristics The result of the transpose operation; Indicates historical fault characteristics The L2 norm of ; Indicates normal operating characteristics The L2 norm of ; Indicates the division sign;

[0040] By performing L2 regularization on the historical fault features and the normal operating features, the difference comparison results of the historical fault features and the normal operating features are generated according to the dot product, that is, the training process of the power equipment fault diagnosis model is completed according to the historical fault data and the normal operating data of the target power equipment, and the target power equipment fault diagnosis model is obtained.

[0041] A further improvement of the present invention lies in that: when the target operating data and the target image data are input into the target power equipment fault diagnosis model for fault judgment, the operating state of the target operating data is judged by a deep learning algorithm to obtain an operating state judgment result; and image information recognition is performed on the target image data by the image recognition algorithm to obtain an image information recognition result, the fault judgment result is obtained based on the operating state judgment result and the image information recognition result, and the fault type and fault location corresponding to the target power equipment are determined based on the fault judgment result.

[0042] The present invention provides an artificial intelligence-based power equipment diagnostic system, comprising:

[0043] A collection module, used for selecting a target electric power device and collecting device data of the target electric power device, wherein the device data includes operation data and image data;

[0044] A preprocessing module, used for acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively;

[0045] A construction module is used to construct an artificial intelligence-based power equipment fault diagnosis model, and obtain historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model to obtain a target power equipment fault diagnosis model;

[0046] A judgment module, used for acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model for fault judgment to obtain a fault judgment result;

[0047] The output module is used to output the fault judgment result and provide corresponding fault handling suggestions according to the fault judgment result.

[0048] The beneficial effect of the present invention is as follows: the present invention proposes an artificial intelligence-based power equipment diagnosis method and system, which selects target power equipment, collects equipment data of the target power equipment, the equipment data includes operation data and image data, obtains the operation data and image data, and performs preprocessing operations on the operation data and image data to generate target operation data and target image data respectively, constructs an artificial intelligence-based power equipment fault diagnosis model, obtains historical fault data and normal operation data of the target power equipment to train the power equipment fault diagnosis model, obtains the target operation data and target image data, and inputs the target operation data and target image data into the target power equipment fault diagnosis model for fault judgment, obtains the fault judgment result, outputs the fault judgment result, and provides corresponding fault handling suggestions based on the fault judgment result, thereby improving the efficiency and accuracy of power equipment fault diagnosis, handling fault conditions in a timely manner, achieving good adaptability and generalization capabilities, and having a high level of intelligence, achieving rapid response to faults while ensuring the diagnosis effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic diagram of a flow chart of an artificial intelligence-based power equipment diagnosis method provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of the structure of an artificial intelligence-based power equipment diagnostic system provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] See also Figure 1 , is a flow chart of an artificial intelligence-based power equipment diagnosis method provided by an embodiment of the present invention, Figure 1The execution subject of the method shown may be a software and / or hardware device. The execution subject of the present application may include but is not limited to at least one of the following: user equipment, network equipment, etc. Among them, user equipment may include but is not limited to computers, smart phones, personal digital assistants (PDA) and the electronic devices mentioned above. Network equipment may include but is not limited to a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. It includes steps S1 to S5, as follows:

[0054] S1, selecting a target power device and collecting device data of the target power device, wherein the device data includes operation data and image data;

[0055] It is understandable that, according to the actual needs of the power system and the importance assessment of the equipment, specific power equipment can be selected as the diagnosis object, such as transformers, circuit breakers, generators and other key components. Using modern sensor technology and image acquisition equipment, the operation data and image data of these power equipment can be fully collected.

[0056] Specifically, operating data includes but is not limited to real-time equipment parameters such as current, voltage, temperature, and vibration frequency, which can directly reflect the operating status and performance changes of the equipment. The high-definition camera on the drone can capture visual information such as the appearance, wiring conditions, and oil level indication of the power equipment, and obtain corresponding image data, thereby improving the efficiency and accuracy of data collection.

[0057] S2, acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively;

[0058] Among them, since the collected raw data often contains noise, redundant information or inconsistent formats, it is usually necessary to pre-process these raw data. Specifically, for the operation data, the present invention adopts filtering, denoising and normalization methods to improve the data quality and ensure the accuracy of subsequent data analysis and processing; for the image data, image processing technologies such as image enhancement, edge detection and feature extraction are used to extract image features that are valuable for fault diagnosis, such as cracks, oil leakage and foreign matter attachment, and then generate target image data.

[0059] Through the above methods, not only the complexity of data processing can be reduced, but also the interpretability of data and the learning ability of artificial intelligence models can be enhanced.

[0060] S3, constructing an artificial intelligence-based power equipment fault diagnosis model, and acquiring historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model, thereby obtaining a target power equipment fault diagnosis model;

[0061] Specifically, based on the preprocessed data, the present invention can construct a power equipment fault diagnosis model that integrates advanced artificial intelligence technologies such as deep learning and machine learning. The model can automatically learn the data characteristics of the equipment under normal operation and fault conditions, and identify potential fault modes through complex algorithm analysis.

[0062] It should be noted that model design usually needs to consider multi-source data fusion, that is, the combination of operating data and image data, so as to make full use of the complementary information between different data types and improve the accuracy and comprehensiveness of fault diagnosis.

[0063] Furthermore, the present invention trains the constructed power equipment fault diagnosis model based on the historical fault data and normal operation data of the target power equipment. Through repeated iterations and continuous adjustment of model parameters, the structure of the power equipment fault diagnosis model can be optimized until the power equipment fault diagnosis model shows good generalization ability and high diagnostic accuracy on the validation set. In addition, the present invention also adopts methods and strategies such as transfer learning and ensemble learning to further improve the performance of the power equipment fault diagnosis model, shorten its training cycle, and reduce the amount of data processing.

[0064] S4, acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model to perform fault diagnosis, and obtain a fault diagnosis result;

[0065] It is understandable that after the training of the power equipment fault diagnosis model is completed, the target power equipment fault diagnosis model can be obtained, and then the target operation data and target image data collected in real time can be input into the target power equipment fault diagnosis model for fault judgment. Specifically, the model can automatically analyze and output fault judgment results based on the input data. These fault judgment results include information such as the fault type, fault location, and fault severity of the power equipment. These fault judgment results can not only provide clear fault indications for operation and maintenance personnel, but also greatly shorten the troubleshooting time of power equipment and improve the efficiency of power equipment fault handling.

[0066] S5, outputting the fault judgment result and providing corresponding fault handling suggestions according to the fault judgment result.

[0067] Through the above method, according to the fault judgment results of the power equipment, corresponding fault handling suggestions can be automatically generated. These fault handling suggestions specifically include emergency handling measures, maintenance plans, preventive measures, etc., so as to provide decision support for operation and maintenance personnel. At the same time, the process and results of each fault diagnosis can be recorded to form a fault diagnosis case library, providing corresponding data support for subsequent model optimization and fault diagnosis, and then through continuous data accumulation and model iteration, the intelligent level and accuracy of power equipment fault diagnosis can be continuously improved to ensure the safe and stable operation of the power system.

[0068] The target sensors on the target power equipment monitor the equipment parameters of the target power equipment in real time to obtain corresponding operation data.

[0069] In practical applications, by integrating sensor technology and using high-precision sensors, real-time monitoring of target power equipment can be achieved, and equipment parameters of the target power equipment can be collected. These equipment parameters can be the temperature, voltage, current, vibration and acoustic signals of the power equipment, etc. According to these equipment parameters, the corresponding operating data of the target power equipment can be obtained, thereby providing rich data support for fault diagnosis of power equipment and ensuring the accuracy and real-time nature of the collected data.

[0070] The target camera device collects external device image data and operating environment image data of the target power equipment to obtain corresponding image data, wherein the image data includes the external device image data and the operating environment image data.

[0071] It can be understood that the target camera device refers to an image capture device preset by the staff in advance. These image capture devices can be drones, etc. The camera on the drone can be used to capture the outer surface image and the surrounding operating environment of the target power equipment, thereby obtaining corresponding image data.

[0072] Through the above method, integrating sensor technology and image capture technology can realize the comprehensive collection of power equipment operation data and image data, ensure the accuracy and comprehensiveness of the collected data, improve the collection efficiency, and provide rich data support for power equipment fault diagnosis.

[0073] The preprocessing operations include data cleaning, normalization, dimensionality reduction and enhancement.

[0074] In practical applications, the purpose of data cleaning operations is to identify and correct errors in collected data, including removing outliers, duplicate values, missing values ​​or unreasonable data, so as to ensure the accuracy and consistency of collected data, including deleting duplicate data, filling missing values, correcting erroneous data, etc. Since power equipment data may come from different sources and formats, data conversion is required to unify data formats and types. Data conversion may include data type conversion, data format conversion, data unit unification, etc., which will help eliminate differences and inconsistencies between data and facilitate subsequent data analysis and processing. And since power equipment data may be scattered in different systems, databases or files, data integration is required to form a complete data set. Data integration may include data extraction, data merging, data connection and other operations to ensure the integrity and consistency of data. Through data conversion and data integration operations, complete data normalization processing can be achieved.

[0075] Furthermore, since power equipment data may contain a large number of features or variables, these features may be redundant or correlated. The purpose of data dimensionality reduction is to reduce the dimension of the data and extract key features to simplify the data structure and improve the analysis efficiency, including feature selection, principal component analysis, cluster analysis and other methods. Finally, the model performance can be improved by increasing data samples and diversity. Specifically, methods such as Mixup and AutoAugment can be used to increase data samples of small categories to solve the problem of data imbalance.

[0076] The construction of an artificial intelligence-based power equipment fault diagnosis model, and obtaining the historical fault data and normal operation data of the target power equipment to train the power equipment fault diagnosis model to obtain the target power equipment fault diagnosis model, includes the following operations:

[0077] The historical fault data and normal operation data of the target power equipment are subjected to feature extraction by the following feature extraction algorithm to obtain historical fault features and normal operation features:

[0078] The extraction process of the historical fault features is as follows:

[0079] Feature extraction is performed on the historical fault data to obtain a historical fault matrix, and the expression of the historical fault matrix is:

[0080]

[0081] In the formula, represents the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of

[0082] Calculate the extreme value of the historical fault matrix, the expression of the extreme value of the historical fault matrix is:

[0083]

[0084] In the formula, represents the historical fault matrix; Represents the extreme value of the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of

[0085] Select the extreme value of historical fault matrix The corresponding pixel point is used as a historical fault feature point, and a stable historical fault feature point is determined. The historical fault feature direction is determined according to the wavelet feature of the circular neighborhood of the stable historical fault feature point, and a 4*4 rectangular area around the stable historical fault feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable historical fault feature point are counted to obtain the historical fault feature;

[0086] The extraction process of the normal operation characteristics is as follows:

[0087] Feature extraction is performed on the normal operation data to obtain a normal operation matrix, and the expression of the normal operation matrix is:

[0088]

[0089] In the formula, represents the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of

[0090] Calculate the extreme value of the normal operation matrix, the expression of the extreme value of the normal operation matrix is:

[0091]

[0092] In the formula, represents the normal operation matrix; Indicates the extreme value of the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of

[0093] Select the extreme value of normal operation matrix The corresponding pixel point is used as a normal operation feature point, and a stable normal operation feature point is determined. The normal operation feature direction is determined according to the wavelet feature of the circular neighborhood of the stable normal operation feature point, and a 4*4 rectangular area around the stable normal operation feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable normal operation feature point are counted to obtain the normal operation feature;

[0094] The historical fault data and normal operation data of the target power equipment are subjected to feature extraction by a feature extraction algorithm to generate the historical fault features and normal operation features.

[0095] The difference between the historical fault characteristics and the normal operating characteristics is compared by the following difference verification algorithm to obtain a difference comparison result, that is, the power equipment fault diagnosis model is trained according to the historical fault data and normal operating data of the target power equipment to obtain the target power equipment fault diagnosis model:

[0096] Set the historical fault characteristics to The normal operating characteristics are ;

[0097] The historical fault characteristics and normal operating characteristics After L2 regularization, the expression of the dot product is:

[0098]

[0099] In the formula, Indicates historical fault characteristics; Indicates normal operating characteristics; Indicates historical fault characteristics and normal operating characteristics The dot product after L2 regularization; Indicates historical fault characteristics The result of the transpose operation; Indicates historical fault characteristics The L2 norm of ; Indicates normal operating characteristics The L2 norm of ; Indicates the division sign;

[0100] By performing L2 regularization on the historical fault features and the normal operating features, the difference comparison results of the historical fault features and the normal operating features are generated according to the dot product, that is, according to the historical fault data and normal operating data of the target power equipment, the training process of the power equipment fault diagnosis model is completed based on the image recognition algorithm and the deep learning model framework to obtain the target power equipment fault diagnosis model.

[0101] Among them, when the target operating data and the target image data are input into the target power equipment fault diagnosis model for fault judgment, the operating state of the target operating data is judged by a deep learning algorithm to obtain an operating state judgment result; and the image information recognition result is obtained by performing image information recognition on the target image data through the image recognition algorithm, and the fault judgment result is obtained according to the operating state judgment result and the image information recognition result, and the fault type and fault location corresponding to the target power equipment are determined according to the fault judgment result.

[0102] It is understandable that the target operation data and target image data are input into the trained target power equipment fault diagnosis model for fault judgment. The target power equipment fault diagnosis model can identify potential fault modes from the data based on machine learning or deep learning algorithms. Combined with the operating principle of the equipment in the store and the knowledge base of the fault mode, the fault type of the power equipment can be judged. Specifically, it is necessary to use methods such as expert systems or rule reasoning. After determining the type of power equipment fault, the use of fault location technology, such as gradual elimination method and step-by-step method, can further narrow the scope of the fault and determine the specific location of the fault. Specifically, this step usually requires a comprehensive analysis of information such as the design drawings and maintenance records of the power equipment.

[0103] It should be noted that the use of image recognition technology, such as convolutional neural networks (CNN), can analyze the target image data, identify abnormal conditions such as damage, cracks, corrosion, etc. on the power equipment, generate image information recognition results, and then output the appearance condition and damage report of the power equipment based on the image information recognition results, providing support for subsequent power equipment fault diagnosis.

[0104] In practical applications, according to the fault diagnosis results of power equipment, corresponding maintenance suggestions and treatment measures can be put forward, including replacing damaged parts, adjusting equipment parameters, and optimizing the operating environment. Then, the fault diagnosis process, results, and maintenance suggestions can be recorded in detail and fed back to relevant departments and personnel, which will help accumulate fault diagnosis experience and improve the efficiency and accuracy of subsequent diagnosis. Finally, according to the practical feedback of power equipment fault diagnosis, the power equipment fault diagnosis model, image recognition algorithm, and fault location technology are continuously optimized, so as to improve the performance and reliability of the overall fault diagnosis system.

[0105] See also Figure 2 , is an artificial intelligence-based power equipment diagnosis system provided by an embodiment of the present invention, the artificial intelligence-based power equipment diagnosis system comprising:

[0106] A collection module, used for selecting a target electric power device and collecting device data of the target electric power device, wherein the device data includes operation data and image data;

[0107] A preprocessing module, used for acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively;

[0108] A construction module is used to construct an artificial intelligence-based power equipment fault diagnosis model, and obtain historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model to obtain a target power equipment fault diagnosis model;

[0109] A judgment module, used for acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model for fault judgment to obtain a fault judgment result;

[0110] The output module is used to output the fault judgment result and provide corresponding fault handling suggestions according to the fault judgment result.

[0111] Figure 2 The apparatus of the embodiment shown can be used to perform Figure 1 The implementation principles and technical effects of the steps in the method embodiment shown are similar and will not be repeated here.

[0112] See also Figure 3 , is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention, the electronic device 30 includes: a processor 31, a memory 32 and a computer program; wherein

[0113] The memory 32 is used to store the computer program, which may also be a flash memory. The computer program is, for example, an application program, a functional module, etc. for implementing the above method.

[0114] The processor 31 is used to execute the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the relevant description in the above method embodiment.

[0115] Optionally, the memory 32 may be independent or integrated with the processor 31 .

[0116] When the memory 32 is a device independent of the processor 31, the device may further include:

[0117] The bus 33 is used to connect the memory 32 and the processor 31 .

[0118] The present invention also provides a readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the various embodiments described above.

[0119] Among them, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application-specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). In addition, the ASIC can be located in a user device. Of course, the processor and the readable storage medium can also exist in a communication device as discrete components. The readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0120] The present invention also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of a device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the device implements the methods provided in the above various embodiments.

[0121] In the embodiments of the above-mentioned devices, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0122] Through the introduction of the above embodiments, the present invention adopts an artificial intelligence-based power equipment diagnosis method and system, which selects target power equipment, collects equipment data of the target power equipment, and the equipment data includes operation data and image data. The operation data and image data are obtained, and the operation data and image data are preprocessed to generate target operation data and target image data respectively, and a power equipment fault diagnosis model based on artificial intelligence is constructed, and the historical fault data and normal operation data of the target power equipment are obtained to train the power equipment fault diagnosis model to obtain the target power equipment fault diagnosis model, and the target operation data and target image data are obtained, and the target operation data and target image data are input into the target power equipment fault diagnosis model for fault judgment, and the fault judgment result is obtained, and the fault judgment result is output, and corresponding fault handling suggestions are provided according to the fault judgment result, so as to improve the efficiency and accuracy of power equipment fault diagnosis, handle fault conditions in a timely manner, achieve good adaptability and generalization ability, and have a high level of intelligence, and achieve rapid response to faults while ensuring the diagnosis effect.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A power equipment diagnosis method based on artificial intelligence, characterized in that: include: Selecting a target electric power device and collecting device data of the target electric power device, wherein the device data includes operation data and image data; Acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively; Constructing an artificial intelligence-based power equipment fault diagnosis model, and acquiring historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model, thereby obtaining a target power equipment fault diagnosis model; Acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model to perform fault diagnosis, and obtain a fault diagnosis result; Output the fault judgment result and provide corresponding fault handling suggestions according to the fault judgment result; The constructing of the artificial intelligence-based power equipment fault diagnosis model, and acquiring the historical fault data and normal operation data of the target power equipment to train the power equipment fault diagnosis model, to obtain the target power equipment fault diagnosis model, includes the following operations: The historical fault data and normal operation data of the target power equipment are subjected to feature extraction by the following feature extraction algorithm to obtain historical fault features and normal operation features: The extraction process of the historical fault features is as follows: Feature extraction is performed on the historical fault data to obtain a historical fault matrix, and the expression of the historical fault matrix is: In the formula, represents the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of Calculate the extreme value of the historical fault matrix, the expression of the extreme value of the historical fault matrix is: In the formula, represents the historical fault matrix; Represents the extreme value of the historical fault matrix; , Indicates the pixel position in the historical fault data; Represents the neighborhood standard deviation of pixels in historical fault data; Indicates the historical fault data at the pixel location The second derivative of Indicates the historical fault data at the pixel location , The second derivative of Indicates the historical fault data at the pixel location The second derivative of Select the extreme value of historical fault matrix The corresponding pixel point is used as a historical fault feature point, and a stable historical fault feature point is determined. The historical fault feature direction is determined according to the wavelet feature of the circular neighborhood of the stable historical fault feature point, and a 4*4 rectangular area around the stable historical fault feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable historical fault feature point are counted to obtain the historical fault feature; The extraction process of the normal operation characteristics is as follows: Feature extraction is performed on the normal operation data to obtain a normal operation matrix, and the expression of the normal operation matrix is: In the formula, represents the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of Calculate the extreme value of the normal operation matrix, the expression of the extreme value of the normal operation matrix is: In the formula, represents the normal operation matrix; Indicates the extreme value of the normal operation matrix; , Indicates the pixel position in normal operation data; Represents the neighborhood standard deviation of pixels in normal operating data; Indicates normal operation data at pixel location The second derivative of Indicates normal operation data at pixel location , The second derivative of Indicates normal operation data at pixel location The second derivative of Select the extreme value of normal operation matrix The corresponding pixel point is used as a normal operation feature point, and a stable normal operation feature point is determined. The normal operation feature direction is determined according to the wavelet feature of the circular neighborhood of the stable normal operation feature point, and a 4*4 rectangular area around the stable normal operation feature point is obtained. The sum of the horizontal direction value, the sum of the vertical direction value, the sum of the horizontal direction absolute value, and the sum of the vertical direction absolute value of the stable normal operation feature point are counted to obtain the normal operation feature; Extracting features from the historical fault data and normal operation data of the target power equipment by using a feature extraction algorithm to generate the historical fault features and normal operation features; The difference between the historical fault characteristics and the normal operating characteristics is compared by the following difference verification algorithm to obtain a difference comparison result, that is, the power equipment fault diagnosis model is trained according to the historical fault data and the normal operating data of the target power equipment to obtain the target power equipment fault diagnosis model: Set the historical fault characteristics to The normal operating characteristics are ; The historical fault characteristics and normal operating characteristics After L2 regularization, the expression of the dot product is: In the formula, Indicates historical fault characteristics; Indicates normal operating characteristics; Indicates historical fault characteristics and normal operating characteristics The dot product after L2 regularization; Indicates historical fault characteristics The result of the transpose operation; Indicates historical fault characteristics The L2 norm of ; Indicates normal operating characteristics The L2 norm of ; Indicates the division sign; By performing L2 regularization on the historical fault features and the normal operating features, the difference comparison results of the historical fault features and the normal operating features are generated according to the dot product, that is, according to the historical fault data and normal operating data of the target power equipment, the training process of the power equipment fault diagnosis model is completed based on the image recognition algorithm and the deep learning model framework to obtain the target power equipment fault diagnosis model.

2. The power equipment diagnosis method based on artificial intelligence according to claim 1 is characterized in that: The target sensor on the target power equipment monitors the equipment parameters of the target power equipment in real time to obtain corresponding operation data.

3. The power equipment diagnosis method based on artificial intelligence according to claim 1 is characterized in that: The external device image data and the operating environment image data of the target power equipment are collected by a target camera device to obtain corresponding image data, wherein the image data includes the external device image data and the operating environment image data.

4. The power equipment diagnosis method based on artificial intelligence according to claim 1 is characterized in that: The preprocessing operations include data cleaning, normalization, dimensionality reduction and enhancement.

5. The power equipment diagnosis method based on artificial intelligence according to claim 1 is characterized in that: When the target operation data and the target image data are input into the target power equipment fault diagnosis model for fault judgment, the operation state judgment of the target operation data is performed through a deep learning algorithm to obtain an operation state judgment result; The target image data is subjected to image information recognition through an image recognition algorithm to obtain an image information recognition result, the fault judgment result is obtained according to the operating state judgment result and the image information recognition result, and the fault type and fault location corresponding to the target power equipment are determined according to the fault judgment result.

6. An artificial intelligence-based power equipment diagnostic system, applied to an artificial intelligence-based power equipment diagnostic method as claimed in any one of claims 1 to 5, the diagnostic system comprising: A collection module, used for selecting a target electric power device and collecting device data of the target electric power device, wherein the device data includes operation data and image data; A preprocessing module, used for acquiring the operation data and the image data, and performing preprocessing operations on the operation data and the image data to generate target operation data and target image data respectively; A construction module is used to construct an artificial intelligence-based power equipment fault diagnosis model, and obtain historical fault data and normal operating data of the target power equipment to train the power equipment fault diagnosis model to obtain a target power equipment fault diagnosis model; A judgment module, used for acquiring the target operation data and the target image data, and inputting the target operation data and the target image data into the target power equipment fault diagnosis model for fault judgment to obtain a fault judgment result; The output module is used to output the fault judgment result and provide corresponding fault handling suggestions according to the fault judgment result.

Images