Transformer substation fault determination method and device and computer readable storage medium

Abstract

The invention relates to the technical field of electric power, and provides a transformer substation fault determination method and device and a computer readable storage medium, and the method comprises the steps: obtaining an original data set of a transformer substation; preprocessing data in the original data set to obtain a target data set; inputting data in the target data set into the target graph model to obtain fault information of the transformer substation in the first time period; the target graph model is obtained by training based on a historical data set of the transformer substation, the target graph model comprises a target constraint function, and the target constraint function is used for indicating physical constraint information of operation of electrical equipment in the transformer substation; the fault information is used for indicating that fault electrical equipment in the multiple pieces of electrical equipment breaks down, and the fault information can also indicate state prediction results such as equipment health indexes and residual life. The accuracy and comprehensiveness of determining the fault of the transformer substation can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power, and particularly relates to a substation fault determination method and device and a computer readable storage medium. BACKGROUND

[0002] A substation is a core link for realizing energy conversion and control in an electric power system, and whether electrical equipment in the substation is faulty directly relates to the safety and reliability of the power grid. To determine whether the substation is faulty, an existing scheme usually installs sensors on each electrical equipment of the substation to determine working parameters of the electrical equipment, and in the case that the working parameters are abnormal, it is determined that the substation is faulty.

[0003] The existing scheme is mostly limited to a single electrical equipment or a single physical process, so that the accuracy and comprehensiveness of determining the fault of the substation are low. SUMMARY

[0004] The present application provides a substation fault determination method and device and a computer readable storage medium, which can improve the accuracy and comprehensiveness of determining the fault of the substation.

[0005] To achieve the above object, the present application adopts the following technical scheme: In a first aspect, a substation fault determination method is provided. The method comprises: obtaining an original data set of a substation; the substation comprising a plurality of electrical devices, the original data set comprising a plurality of sub-original data sets, the plurality of sub-original data sets corresponding one-to-one to the plurality of electrical devices, each sub-original data set comprising a plurality of working parameter original value sets of an electrical device, each working parameter original value set corresponding one-to-one to a working parameter of the electrical device, and each working parameter original value set comprising a plurality of parameter values of the working parameter of the electrical device within a first time period; preprocessing data in the original data set to obtain a target data set; inputting data in the target data set into a target graph model to obtain fault information of the substation within the first time period; the target graph model comprising a plurality of nodes and connection relationships between the plurality of nodes, the plurality of nodes corresponding one-to-one to the plurality of electrical devices, the connection relationships between the plurality of nodes being connection relationships between the plurality of electrical devices, the target graph model being trained based on a historical data set of the substation, the historical data set comprising a plurality of sub-historical data sets, the plurality of sub-historical data sets corresponding one-to-one to the plurality of electrical devices, each sub-historical data set comprising a plurality of working parameter historical value sets of an electrical device, each working parameter historical value set corresponding one-to-one to a working parameter of the electrical device, and each working parameter historical value set comprising a plurality of parameter values of the working parameter of the electrical device within a historical time period, the target graph model comprising a target constraint function, and the target constraint function being used to indicate physical constraint information of operation of the electrical devices in the substation; and the fault information being used to indicate that a fault electrical device in the plurality of electrical devices is faulty, and the fault information further being used to indicate state prediction results such as a device health index and a remaining life.

[0006] In combination with the first aspect, in some embodiments of the first aspect, before obtaining the original data set of the substation, the method further comprises: obtaining a historical data set of the substation; determining feature information of the plurality of electrical devices within a historical time period according to the historical data set; and determining the target graph model according to the feature information.

[0007] In combination with the first aspect, in some embodiments of the first aspect, the feature information comprises data statistical feature information and first feature vector information, the first feature vector information being determined based on a preset deep learning model, and the feature information of the plurality of electrical devices within the historical time period being determined according to the historical data set comprises: for each electrical device in the plurality of electrical devices, determining a mathematical statistical feature of a working parameter corresponding to each working parameter historical value set in a sub-historical data set corresponding to the electrical device; and inputting parameter values in each working parameter historical value set in the sub-historical data set corresponding to the electrical device into the preset deep learning model to obtain first feature vector information of each working parameter input by the preset deep learning model.

[0008] In some embodiments of the first aspect, the target graph model is determined according to the feature information, including: for each of the plurality of electrical devices, inputting the mathematical statistical feature of each of the plurality of working parameters of the electrical device and the first feature vector information into a preset self-attention model to obtain second feature vector information of the electrical device output by the preset self-attention model; and determining the target graph model according to the second feature vector information of each of the plurality of electrical devices.

[0009] In some embodiments of the first aspect, the target graph model is determined according to the second feature vector information of each of the plurality of electrical devices, including: obtaining the electrical connection relationship between the plurality of electrical devices; taking the second feature vector of each electrical device as the feature of a node and taking the electrical connection relationship as the connection relationship between nodes to obtain an original graph model; and training the original graph model to obtain the target graph model.

[0010] In some embodiments of the first aspect, the target graph model is determined by training the original graph model, including: obtaining a target constraint function; embedding the target constraint function into a loss function of the original graph model; and training the original graph model based on the loss function of the original graph model to obtain the target graph model.

[0011] In some embodiments of the first aspect, the data in the original data set is preprocessed to obtain a target data set, including: performing data cleaning on the data in the original data set to obtain an intermediate data set; and performing normalization and dimension unification processing on the data in the intermediate data set to obtain the target data set.

[0012] In a second aspect, a substation fault determination apparatus is provided for implementing the substation fault determination method of the first aspect. The substation fault determination apparatus includes modules, units, or means corresponding to the above-mentioned method, which can be implemented by hardware, software, or by hardware executing corresponding software. The hardware or software includes one or more modules or units corresponding to the above-mentioned functions.

[0013] With reference to the second aspect, in some embodiments of the second aspect, the apparatus comprises: an obtaining module and a processing module; the obtaining module is configured to obtain an original data set of the power substation; the power substation comprises a plurality of electrical devices, the original data set comprises a plurality of sub-original data sets, the plurality of sub-original data sets correspond to the plurality of electrical devices one by one, each sub-original data set comprises a plurality of working parameter original value sets of an electrical device, the plurality of working parameter original value sets correspond to a plurality of working parameters of the electrical device one by one, and each working parameter original value set comprises a plurality of parameter values of the working parameter of the electrical device within a first time period; the processing module is configured to pre-process data in the original data set to obtain a target data set; and the processing module is further configured to input data in the target data set into a target graph model to obtain fault information of the power substation within the first time period; the target graph model comprises a plurality of nodes and connection relationships between the plurality of nodes, the plurality of nodes correspond to the plurality of electrical devices one by one, the connection relationships between the plurality of nodes are connection relationships between the plurality of electrical devices, the target graph model is trained based on a historical data set of the power substation, the historical data set comprises a plurality of sub-historical data sets, the plurality of sub-historical data sets correspond to the plurality of electrical devices one by one, each sub-historical data set comprises a plurality of working parameter historical value sets of an electrical device, the plurality of working parameter historical value sets correspond to a plurality of working parameters of the electrical device one by one, each working parameter historical value set comprises a plurality of parameter values of the working parameter of the electrical device within a historical time period, and the target graph model comprises a target constraint function, which is used to indicate physical constraint information of the electrical devices in the power substation; and the fault information is used to indicate a fault electrical device in the plurality of electrical devices.

[0014] With reference to the second aspect, in some embodiments of the second aspect, before the obtaining module is configured to obtain the original data set of the power substation, the processing apparatus is further configured to: obtain a historical data set of the power substation; determine feature information of the plurality of electrical devices within a historical time period based on the historical data set; and determine the target graph model based on the feature information.

[0015] With reference to the second aspect, in some embodiments of the second aspect, the feature information comprises data statistical feature information and first feature vector information, the first feature vector information is determined based on a preset deep learning model, and the processing module, configured to determine the feature information of the plurality of electrical devices within the historical time period based on the historical data set, comprises: for each electrical device in the plurality of electrical devices, determining a mathematical statistical feature of a working parameter corresponding to each working parameter historical value set in the sub-historical data set corresponding to the electrical device; and inputting parameter values in each working parameter historical value set in the sub-historical data set corresponding to the electrical device into the preset deep learning model to obtain first feature vector information of each working parameter input by the preset deep learning model.

[0016] With reference to the second aspect, in some embodiments of the second aspect, the processing module, configured to determine the target graph model according to the feature information, comprises: inputting, for each of the plurality of electrical devices, the mathematical statistical feature of each of the plurality of working parameters of the electrical device and the first feature vector information into a preset self-attention model to obtain second feature vector information of the electrical device output by the preset self-attention model; and determining the target graph model according to the second feature vector information of each of the plurality of electrical devices.

[0017] With reference to the second aspect, in some embodiments of the second aspect, the processing module, configured to train the original graph model according to the second feature vector information of each of the plurality of electrical devices to obtain the target graph model, comprises: obtaining the electrical connection relationship between the plurality of electrical devices; taking the second feature vector of each of the electrical devices as the feature of the node and taking the electrical connection relationship as the connection relationship between the nodes to obtain the original graph model; and training the original graph model to obtain the target graph model.

[0018] With reference to the second aspect, in some embodiments of the second aspect, the processing module, configured to train the original graph model to obtain the target graph model, comprises: obtaining a target constraint function; embedding the target constraint function into a loss function of the original graph model; and training the original graph model based on the loss function of the original graph model to obtain the target graph model.

[0019] With reference to the second aspect, in some embodiments of the second aspect, the processing module, configured to preprocess the data in the original data set to obtain the target data set, comprises: performing data cleaning on the data in the original data set to obtain an intermediate data set; and performing normalization and dimension unification processing on the data in the intermediate data set to obtain the target data set.

[0020] A substation fault determination apparatus is provided in a third aspect. The substation fault determination apparatus comprises at least one processor and a memory storing processor-executable instructions. The processor is configured to execute the instructions to implement the method provided in the first aspect and any possible implementation thereof.

[0021] A computer-readable storage medium is provided in a fourth aspect. When instructions in the computer-readable storage medium are executed by a processor of a substation fault determination apparatus, the substation fault determination apparatus is enabled to perform the method provided in the first aspect and any possible implementation thereof.

[0022] A computer program product containing instructions is provided in a fifth aspect. When the computer program product is run on a computer, the computer is enabled to perform the method provided in the first aspect and any possible implementation thereof.

[0023] Compared with the prior scheme of installing sensors on each electrical device of a substation to determine the working parameters of the electrical device and determining that the substation has a fault in the case where the working parameters are abnormal, the scheme provided in the application obtains an original data set including parameter values of working parameters of a plurality of electrical devices of a substation in a first time period, obtains a target data set by preprocessing the original data set, inputs data in the target data set into a target graph model, and obtains fault information of the substation in the first time period output by the target graph model. Since the target graph model includes a plurality of nodes and connection relationships between the plurality of nodes, the plurality of nodes correspond to the plurality of electrical devices one by one, the connection relationships between the plurality of nodes are connection relationships between the plurality of electrical devices, the target graph model is trained based on a historical data set of the substation, and the target graph model further includes a target constraint function for indicating physical constraint information of the electrical devices in the substation, the fault information of the substation in the first time period output by the target graph model can be obtained. By determining the fault information of the substation according to the connection relationships between the plurality of electrical devices and the physical constraint information, the accuracy and comprehensiveness of determining the fault of the substation can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 An architecture schematic diagram of a substation fault determination system provided in the application; Figure 2 A flowchart schematic diagram of a substation fault determination method provided in the application; Figure 3 A flowchart schematic diagram of another substation fault determination method provided in the application; Figure 4 A flowchart schematic diagram of another substation fault determination method provided in the application; Figure 5 A flowchart schematic diagram of another substation fault determination method provided in the application; Figure 6 A flowchart schematic diagram of another substation fault determination method provided in the application; Figure 7 A flowchart schematic diagram of another substation fault determination method provided in the application; Figure 8 A structure schematic diagram of a substation fault determination apparatus provided in the application; Figure 9 A structure schematic diagram of another substation fault determination apparatus provided in the application. DETAILED DESCRIPTION

[0025] In the description of the present application, "a plurality of" means two or more than two, unless otherwise specified. "At least one of the following" or similar expressions means any combination of the items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0026] In addition, in order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, "first", "second" and the like are used to distinguish the same items or similar items with basically the same function and role. Those skilled in the art can understand that "first", "second" and the like do not limit the quantity and execution order, and "first", "second" and the like do not necessarily mean different.

[0027] Meanwhile, in the embodiments of the present application, the words such as "exemplary" or "for example" are used to mean as an example, illustration or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words such as "exemplary" or "for example" are intended to present the relevant concept in a specific manner, for easy understanding.

[0028] It can be understood that the "embodiments" mentioned throughout the specification mean that the specific features, structures or characteristics related to the embodiments are included in at least one embodiment of the present application. Therefore, the various embodiments throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It can be understood that in various embodiments of the present application, the size of the sequence number of each process does not mean the execution order, and the execution order of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0029] It can be understood that in the present application, "when", "if" and "if" all refer to the corresponding processing under certain objective circumstances, not limited to time, and do not require a judgment action when implemented, nor does it mean that there are other limitations.

[0030] It can be understood that some optional features in the embodiments of the present application can be implemented independently in some scenarios without relying on other features, such as the scheme currently based on, to solve the corresponding technical problems and achieve the corresponding effects. In some scenarios, it can also be combined with other features according to demand. Correspondingly, the devices given in the embodiments of the present application can also realize these features or functions, which will not be described here.

[0031] In the present application, the same or similar parts among various embodiments can be mutually referred to, unless otherwise specified. In the various embodiments in the present application, and the various implementation methods in the various embodiments, the terms and / or descriptions among different embodiments, and the various implementation methods in the various embodiments are consistent, and can be mutually referred to, unless otherwise specified and in conflict with logic, and the technical features in different embodiments, and the various implementation methods in the various embodiments can be combined to form new embodiments, implementation manners, implementation methods, or implementation methods according to their inherent logical relationship. The following implementation manners of the present application do not constitute a limitation on the protection scope of the present application.

[0032] The substation is a core link for realizing energy conversion and control in the power system, and whether the electrical equipment in the substation fails directly relates to the safety and reliability of the power grid. In order to determine whether the substation fails, the existing scheme usually installs sensors on each electrical equipment in the substation to determine the working parameters of the electrical equipment, and determines that the substation fails in the case that the working parameters are abnormal.

[0033] Most of the existing schemes are limited to a single electrical equipment or a single physical process, so that the accuracy and comprehensiveness of determining the failure of the substation are low.

[0034] To solve the problem, the present application provides a substation failure determination method, which comprises: obtaining an original data set of a substation; the substation comprises a plurality of electrical equipment, the original data set comprises a plurality of sub-original data sets, the plurality of sub-original data sets correspond one by one to the plurality of electrical equipment, the sub-original data set comprises a plurality of working parameter original value sets of the electrical equipment, the plurality of working parameter original value sets correspond one by one to a plurality of working parameters of the electrical equipment, and the working parameter original value set comprises a plurality of parameter values of the working parameter of the electrical equipment in a first time period; obtaining a target data set by preprocessing the data in the original data set; inputting the data in the target data set into a target graph model to obtain failure information of the substation in the first time period; the target graph model comprises a plurality of nodes and a connection relationship between the plurality of nodes, the plurality of nodes correspond one by one to the plurality of electrical equipment, the connection relationship between the plurality of nodes is a connection relationship between the plurality of electrical equipment, the target graph model is obtained based on a historical data set of the substation, the historical data set comprises a plurality of sub-historical data sets, the plurality of sub-historical data sets correspond one by one to the plurality of electrical equipment, the sub-historical data set comprises a plurality of working parameter historical value sets of the electrical equipment, the plurality of working parameter historical value sets correspond one by one to the plurality of working parameters of the electrical equipment, the working parameter historical value set comprises a plurality of parameter values of the working parameter of the electrical equipment in a historical time period, and the target graph model comprises a target constraint function, the target constraint function being used to indicate physical constraint information of the electrical equipment running in the substation; the failure information is used to indicate that a fault electrical equipment in the plurality of electrical equipment fails.

[0035] Based on the scheme, compared with the existing scheme of installing sensors on each electrical device of the substation to determine the working parameters of the electrical device, in the case that the working parameters appear abnormity, the scheme of the present application determines the failure of the substation. The scheme of the present application obtains the original data set including the parameter values of the working parameters of the plurality of electrical devices of the substation in the first time period, obtains the target data set after pre-processing the original data set, and inputs the data in the target data set into the target graph model. Since the target graph model includes a plurality of nodes and a connection relationship between the plurality of nodes, the plurality of nodes correspond to the plurality of electrical devices one by one, the connection relationship between the plurality of nodes is the connection relationship between the plurality of electrical devices, the target graph model is obtained based on the historical data set of the substation, and the target graph model further includes a target constraint function for indicating the physical constraint information of the electrical devices in the substation. Therefore, the failure information of the substation in the first time period output by the target graph model can be obtained, and the accuracy and comprehensiveness of determining the failure of the substation can be improved by determining the failure information of the substation according to the connection relationship between the plurality of electrical devices and the physical constraint information.

[0036] Figure 1 The technical scheme of the embodiment of the present application can be applied to the architecture of the substation failure determination system provided by the present application. As shown in the architecture of the substation failure determination system provided by the present application, the substation failure determination system 10 includes a substation failure determination device 11 and an electronic electrical device 12. Figure 1 As shown in the architecture of the substation failure determination system provided by the present application, the substation failure determination system 10 includes a substation failure determination device 11 and an electronic electrical device 12. Figure 1 As shown in the architecture of the substation failure determination system provided by the present application, the substation failure determination system 10 includes a substation failure determination device 11 and an electronic electrical device 12.

[0037] The substation failure determination device 11 and the electronic electrical device 12 are directly or indirectly connected. In the connection relationship, a wired connection or a wireless connection can be used, and the embodiment of the present application does not limit this.

[0038] The substation failure determination device 11 and the electronic electrical device 12 can exchange data.

[0039] It should be noted that the substation failure determination device 11 and the electronic electrical device 12 can be independent electrical devices or integrated into the same electrical device, and the present application does not make specific limitations.

[0040] When the substation failure determination device 11 and the electronic electrical device 12 are integrated into the same electrical device, the communication mode between the substation failure determination device 11 and the electronic electrical device 12 is the communication between the internal modules of the electrical device. In this case, the communication process between the two is the same as the communication process between the substation failure determination device 11 and the electronic electrical device 12 when they are independent of each other.

[0041] In the following embodiments provided by the present application, the substation fault determination device 11 and the electronic and electrical equipment 12 are taken as examples to be described independently of each other.

[0042] In actual applications, the substation fault determination method provided by the embodiments of the present application can be applied to the substation fault determination device 11, and can also be applied to the device included in the substation fault determination device 11.

[0043] The substation fault determination method provided by the embodiments of the present application will be described below with reference to the drawings, taking the substation fault determination method applied to the substation fault determination device 11 as an example.

[0044] Figure 2 A flowchart of a substation fault determination method provided by the present application is shown in FIG. 1, which includes the following steps. Figure 2 S201, the substation fault determination device acquires a raw data set of the substation.

[0045] The substation includes a plurality of electrical equipment, and the raw data set includes a plurality of sub-raw data sets, the plurality of sub-raw data sets correspond to the plurality of electrical equipment one by one, the sub-raw data set includes a plurality of working parameter raw value sets of the electrical equipment, the plurality of working parameter raw value sets correspond to a plurality of working parameters of the electrical equipment one by one, and the working parameter raw value set includes a plurality of parameter values of the working parameter of the electrical equipment within a first time period.

[0046] It should be noted that the electrical equipment in the substation can include a transformer, a circuit breaker, a disconnector, a bus, a current transformer, and a voltage transformer. Of course, the electrical equipment can also include other devices, which are not limited by the present application.

[0047] The first time period can have any length, and the first time period can have any starting time and any ending time, which are not limited by the present application. The working parameter can include electrical data, heat and equipment temperature data, vibration and acoustic data, partial discharge monitoring data, oil chromatographic analysis (DGA) data, or environmental and working condition data.

[0048] The electrical data can reflect the electrical operating characteristics of the electrical equipment, and can include voltage, current, active / reactive power, frequency, harmonic component, etc., which can be collected by a current transformer (CT), a voltage transformer (VT), and a power quality monitoring device.

[0049] ​Thermal and temperature data can reflect the thermal state and heat dissipation capacity of electrical equipment, and can include transformer top oil temperature, winding hot spot temperature estimation, switch contact temperature rise, bus and joint temperature, etc., which can be determined by infrared temperature measurement, optical fiber temperature sensor and thermocouple measurement.

[0050] Vibration and acoustic data can reflect the mechanical wear and insulation defects of electrical equipment, and can include circuit breaker operation vibration, transformer core and winding vibration, partial discharge ultrasonic signal, etc., which are mainly collected by accelerometers, piezoelectric vibration sensors and ultrasonic / acoustic emission sensors.

[0051] Partial discharge monitoring data can identify insulation defects and discharge development trends of electrical equipment, and can be collected by ultra-high frequency (UHF) sensors, acoustic-electric joint probes or pulse current methods to obtain PRPD spectrum, discharge amplitude and phase distribution, etc.

[0052] Oil chromatographic analysis (DGA) data can reveal the decomposition law and potential fault types of the insulation oil of electrical equipment, and can collect characteristic gases such as hydrogen, acetylene, methane, carbon monoxide and their change rates by online DGA devices or gas chromatographs.

[0053] Environmental and working condition data can reflect the operating environment conditions and working condition background of electrical equipment, and can include environmental temperature, humidity, wind speed, cooling method and load level around the electrical equipment, etc., which are collected by temperature and humidity sensors, anemometers and SCADA / EMS systems.

[0054] The sampling rate of electrical data can be 1-10 kHz to capture power frequency fundamental wave, harmonic and transient characteristics.

[0055] The sampling period of thermal and temperature data can be 1-5 minutes to reflect slow variable characteristics such as hot spot temperature and contact temperature rise.

[0056] The sampling rate of vibration and acoustic data can be 10-100 kHz to obtain transient characteristics such as circuit breaker operation shock waveform and partial discharge ultrasonic signal.

[0057] The sampling rate of partial discharge monitoring data can be 10-200 MHz (UHF signal) or 100 kHz-10 MHz (pulse current method) to ensure the capture of high-frequency discharge pulses.

[0058] The sampling period of oil chromatographic (DGA) data can be 30 minutes-1 day, with online monitoring updated at the hourly level and offline detection generally at the weekly / monthly level.

[0059] The sampling period of environmental and working condition data can be 1-10 minutes to reflect the environmental changes of electrical equipment.

[0060] As a possible implementation manner, in combination with Figure 1 The substation fault determination apparatus sends a request message for requesting the original data set of the substation to the electronic device, and correspondingly, the electronic device receives the request message of the substation fault determination apparatus.

[0061] The electronic device generates a response message including the original data set of the substation, and sends the response message to the substation fault determination apparatus. Correspondingly, the substation fault determination apparatus receives the response message, and obtains the original data set of the substation from the response message.

[0062] S202, the substation fault determination apparatus pre-processes the data in the original data set to obtain a target data set.

[0063] As a possible implementation manner, the substation fault determination apparatus performs data cleaning on the data in the original data set to obtain an intermediate data set, and performs normalization and dimension unification processing on the data in the intermediate data set to obtain the target data set.

[0064] As an example, the substation fault determination apparatus removes the repetition, abnormal jump and saturation value in the original data set, detects and eliminates the long-time zero value or constant value data, then performs wavelet denoising and band-pass filtering on the electrical data, and performs empirical mode decomposition (EMD) and adaptive filtering on the vibration and acoustic data, then fills in the missing values in the original data set by using time series interpolation, Kalman filtering or a method based on Bayesian inference to obtain the intermediate data set.

[0065] The substation fault determination apparatus unifies different physical quantities in the intermediate data set to dimensionless form or standard physical unit, and adopts a robust normalization method, for example, quantile scaling, to obtain the target data set.

[0066] S203, the substation fault determination apparatus inputs the data in the target data set into a target graph model to obtain the fault information of the substation in the first time period.

[0067] The target graph model includes a plurality of nodes and a connection relationship between the plurality of nodes, the plurality of nodes correspond to the plurality of electrical devices one by one, the connection relationship between the plurality of nodes is a connection relationship between the plurality of electrical devices, the target graph model is obtained based on a historical data set of the substation, the historical data set includes a plurality of sub historical data sets, the plurality of sub historical data sets correspond to the plurality of electrical devices one by one, the sub historical data set includes a plurality of working parameter historical value sets of the electrical device, the plurality of working parameter historical value sets correspond to a plurality of working parameters of the electrical device one by one, the working parameter historical value set includes a plurality of parameter values of the working parameter of the electrical device in a historical time period, the target graph model includes a target constraint function, and the target constraint function is used to indicate physical constraint information of the electrical device in the substation; and the fault information is used to indicate that a fault electrical device in the plurality of electrical devices is faulty, and the fault information can also indicate state prediction results such as a device health index and a remaining life.

[0068] After the substation fault determination apparatus inputs data in the target data set into the target graph model, the target graph model can determine fault information of the substation in the first time period based on the target data set and output the fault information, so that the substation fault determination apparatus obtains the fault information of the substation in the first time period.

[0069] Based on the scheme, compared with the existing scheme of installing sensors on each electrical device in the substation to determine working parameters of the electrical device and determining that the substation is faulty in the case of abnormal working parameters, in the scheme provided in the present application, the original data set including parameter values of working parameters of the plurality of electrical devices in the substation in the first time period is obtained, the target data set is obtained by preprocessing the original data set, data in the target data set is input into the target graph model, because the target graph model includes a plurality of nodes and a connection relationship between the plurality of nodes, the plurality of nodes correspond to the plurality of electrical devices one by one, the connection relationship between the plurality of nodes is a connection relationship between the plurality of electrical devices, the target graph model is obtained based on a historical data set of the substation, and the target graph model further includes a target constraint function used to indicate physical constraint information of the electrical device in the substation, so that the fault information of the substation in the first time period output by the target graph model can be obtained, and by determining the fault information of the substation according to the connection relationship between the plurality of electrical devices and the physical constraint information, the accuracy and comprehensiveness of determining the fault of the substation can be improved.

[0070] The above is a general description of the substation fault determination method provided in the present application, and the substation fault determination method provided in the present application will be further described below with reference to the accompanying drawings.

[0071] In one design, Figure 3 A flowchart of still another substation fault determination method provided in the present application is shown in FIG. 6.Figure 3 Before S201, the method further includes: S301, the substation fault determination apparatus acquires a historical data set of the substation.

[0072] It should be noted that the length of the historical time period can be the same as that of the first time period, or it can be different, and the present application does not make a specific limitation thereon.

[0073] The end moment of the historical time period is before the start moment of the first time period.

[0074] The working parameters corresponding to the historical data set and the collection frequency of the working parameters are described with reference to the related description in S201, and the present application will not be described here.

[0075] As a possible implementation manner, in combination with Figure 1 The substation fault determination apparatus sends a request message for requesting the historical data set of the substation to the electronic device, and correspondingly, the electronic device receives the request message of the substation fault determination apparatus.

[0076] The electronic device generates a response message including the historical data set of the substation, and sends the response message to the substation fault determination apparatus. Correspondingly, the substation fault determination apparatus receives the response message and acquires the historical data set of the substation from the response message.

[0077] S302, the substation fault determination apparatus determines the feature information of the plurality of electrical equipment in the historical time period according to the historical data set.

[0078] It should be noted that the feature information can include data statistical feature information and first feature vector information, and the first feature vector information is determined based on a preset deep learning model.

[0079] As a possible implementation manner, for each electrical equipment in the plurality of electrical equipment, the substation fault determination apparatus determines the mathematical statistical feature of the working parameter corresponding to each working parameter historical value set in the sub-history data set corresponding to the electrical equipment.

[0080] The substation fault determination apparatus inputs the parameter value in each working parameter historical value set in the sub-history data set corresponding to the electrical equipment into the preset deep learning model to obtain the first feature vector information of each working parameter output by the preset deep learning model.

[0081] It should be noted that the specific description of this possible implementation manner can refer to the related description in the subsequent part of the specific embodiment of the present application, and the present application will not be described here.

[0082] S303, the substation fault determination apparatus determines a target graph model according to the feature information.

[0083] As a possible implementation manner, the substation fault determination apparatus inputs, for each electrical device in the plurality of electrical devices, the mathematical statistical feature of each working parameter of the plurality of working parameters of the electrical device and the first feature vector information into a preset self-attention model to obtain second feature vector information of the electrical device output by the preset self-attention model.

[0084] The substation fault determination apparatus determines a target graph model according to the second feature vector information of each electrical device in the plurality of electrical devices.

[0085] It should be noted that the specific description of this possible implementation manner can refer to the related description in the subsequent part of the specific embodiments of the present application, which is not described herein.

[0086] Based on S301-S303, by obtaining the historical data set of the substation, the feature information of the plurality of electrical devices in the historical time period can be determined according to the historical data set, so as to determine the target graph model according to the feature information, so as to determine the fault information of the substation based on the target graph model subsequently.

[0087] In one design, the feature information includes data statistical feature information and first feature vector information, the first feature vector information being determined based on a preset deep learning model, Figure 4 Another flowchart of a substation fault determination method provided by the present application is shown as follows, Figure 4 S302 can specifically include the following steps: S401, the substation fault determination apparatus, for each electrical device in the plurality of electrical devices, determines the mathematical statistical feature of each working parameter in the working parameter historical value set corresponding to the working parameter historical value set according to each working parameter historical value set in the sub-historical data set corresponding to the electrical device.

[0088] It should be noted that the mathematical statistical feature can include time domain features (for example, mean, variance, skewness, kurtosis, pulse factor, margin factor), frequency domain features (for example, FFT extraction of fundamental wave, harmonic and frequency component), time-frequency domain features (for example, STFT, wavelet packet decomposition), and professional features (for example, PRPD spectrum texture parameters, oil spectrum ratio method indicators, temperature change rate and load correlation), of course, the mathematical statistical feature can also include other features, which are not limited by the present application.

[0089] As a possible implementation manner, the substation fault determination apparatus acquires a preset correspondence relationship, the preset correspondence relationship including a mathematical statistical feature corresponding to each working parameter, and the substation fault determination apparatus, for each working parameter, searches for the mathematical statistical feature corresponding to the working parameter in the preset correspondence relationship, and then determines the mathematical statistical feature corresponding to the working parameter based on the historical value set of each working parameter.

[0090] S402, the substation fault determination apparatus inputs the parameter value in each working parameter historical value set in the substation fault determination apparatus corresponding to the electrical equipment into the preset deep learning model, and obtains the first feature vector information of each working parameter input by the preset deep learning model.

[0091] It should be noted that the preset deep learning model can be a CNN feature model, a TCN feature model, or a recurrent network model. Of course, the preset deep learning model can also be other models, which are not limited in the present application.

[0092] It can be understood that after the substation fault determination apparatus inputs the parameter value in the working parameter historical value set into the preset deep learning model, the preset deep learning model can analyze the parameter value to obtain the first feature vector information of the working parameter.

[0093] Based on S401-S402, by determining the mathematical statistical feature of the working parameter of the electrical equipment, the overall distribution of the working parameter of the electrical equipment can be reflected from the data statistical level, and by determining the first feature vector information of the working parameter of the electrical equipment based on the preset deep model, the running state of the electrical equipment can be determined based on the working parameter of the electrical equipment.

[0094] In one design, Figure 5 Another flowchart of a substation fault determination method provided by the present application is shown in FIG. 6. Figure 5 As shown in FIG. 6, S303 can specifically include the following steps: S501, the substation fault determination apparatus inputs the mathematical statistical feature and the first feature vector information of each working parameter in the multiple working parameters of the electrical equipment into the preset self-attention model, and obtains the second feature vector information of the electrical equipment output by the preset self-attention model.

[0095] It should be noted that the preset self-attention model can be a BERT model, a GPT model, or a BART model. Of course, the preset self-attention model can also be other models, which are not limited in the present application.

[0096] It can be understood that the substation fault determination apparatus inputs the mathematical statistical features of each of the plurality of working parameters of the electrical equipment and the first feature vector information into the preset self-attention model, and the preset self-attention model can analyze the data to obtain the second feature vector information of the electrical equipment.

[0097] S502, the substation fault determination apparatus determines a target graph model according to the second feature vector information of each of the plurality of electrical equipment.

[0098] As a possible implementation manner, the substation fault determination apparatus obtains an electrical connection relationship between the plurality of electrical equipment; takes the second feature vector of each electrical equipment as a feature of a node, and takes the electrical connection relationship as a connection relationship between nodes to obtain an original graph model; and trains the original graph model to obtain the target graph model.

[0099] It should be noted that the specific description of this possible implementation manner can be referred to the related description in the subsequent part of the specific embodiment of the present application, which is not described herein.

[0100] Based on S501-S502, by inputting the mathematical statistical features of each of the plurality of working parameters of the electrical equipment and the first feature vector information into the preset self-attention model, the second feature vector information of the electrical equipment output by the preset self-attention model is obtained, and then the target graph model is determined according to the second feature vector information of each of the plurality of electrical equipment. The advantages of traditional statistical features and deep learning features are fused through the self-attention mechanism, a highly condensed and context-aware second feature vector is generated for each electrical equipment, which not only more accurately represents the comprehensive state of the device itself, but more importantly, these high-quality feature vectors lay a reliable foundation for subsequent graph model construction, so that the potential complex association, dependency relationship or group failure mode between devices can be accurately mined and presented, thereby realizing a leap from individual deep perception to group intelligence insight, and greatly improving the intelligent level of device cluster management and risk prediction.

[0101] In one design, Figure 6 Another flowchart of a substation fault determination method provided by the present application is shown in FIG. 6. Figure 6 As shown in FIG. 6, S502 can include the following steps: S601, the substation fault determination apparatus obtains an electrical connection relationship between the plurality of electrical equipment.

[0102] As a possible implementation manner, in combination with Figure 1 , the substation fault determination apparatus sends a request message for requesting the electrical connection relationship between the plurality of electrical equipment to the electronic device, and correspondingly, the electronic device receives the request message of the substation fault determination apparatus.

[0103] The electronic device generates a response message including the electrical connection relationship between the plurality of electrical devices, and sends the response message to the substation fault determination apparatus. Correspondingly, the substation fault determination apparatus receives the response message and obtains the electrical connection relationship between the plurality of electrical devices from the response message.

[0104] S602, the substation fault determination apparatus takes the second feature vector of each electrical device as the feature of a node, and takes the electrical connection relationship as the connection relationship between nodes, to obtain an original graph model.

[0105] It can be understood that the substation fault determination apparatus can construct the original graph model after determining the features of each node and the connection relationship between nodes. The specific construction method can refer to the existing scheme, and will not be described herein.

[0106] S603, the substation fault determination apparatus trains the original graph model to obtain a target graph model.

[0107] As a possible implementation manner, the substation fault determination apparatus obtains a target constraint function; embeds the target constraint function into a loss function of the original graph model; and trains the original graph model based on the loss function of the original graph model to obtain the target graph model.

[0108] It should be noted that the specific description of this possible implementation manner can refer to the related description in the subsequent part of the specific embodiment of the present application, which will not be described herein.

[0109] Based on S601-S603, by taking the second feature vector of each electrical device as the feature of a node and taking the electrical connection relationship between the plurality of electrical devices as the connection relationship between nodes, the inherent electrical connection relationship between the devices in the substation can be used to construct the graph model, which makes the model not only analyze the working state of each device itself (node feature), but also consider the propagation and correlation influence of the fault on the physical connection. Through the information transmission of the graph structure, the model can more accurately locate the fault source point and effectively distinguish the abnormality of the device itself and the abnormality caused by the adjacent fault wave, thereby improving the accuracy and robustness of fault diagnosis.

[0110] In one design, Figure 7 Another flowchart of a substation fault determination method provided by the present application is shown in FIG. 6. Figure 7 As shown in FIG. 6, S603 can include the following steps: S701, the substation fault determination apparatus obtains a target constraint function.

[0111] It should be noted that the target constraint function can include an electrical constraint residual term: in the electrical constraint residual term, denotes a set of nodes in the target graph model, is the number of nodes, denotes a set of branches or current channels connected to the node denotes the current through the th branch or channel, denotes a set of loops or lines surrounding the node denotes the voltage or potential difference contribution on the th loop / line. The target constraint function can also include a thermal conduction constraint residual term:

[0112] in the thermal conduction constraint residual term, denotes the average squared error of the residual of the heat conduction equation at all sampling points, denotes the temperature field distribution inside (or on the surface) of the substation equipment or its components, is time, is the equivalent thermal diffusivity (comprehensively reflecting the thermal conductivity, density and specific heat capacity of the material), is the Laplacian operator with respect to spatial coordinates, i.e. the second-order spatial derivative, denotes the internal heat source term per unit volume (such as copper loss, iron loss, contact resistance heating, etc.); after discretization, and denote the temperature and heat source values at the th sampling space-time point, and is the number of sampling points involved in the thermal mechanism constraint. The target constraint function can also include a mechanical vibration dynamics constraint residual term:

[0113] in the mechanical vibration dynamics constraint residual term, is the damping coefficient, is the equivalent mass, is the equivalent stiffness coefficient, denotes the mechanical displacement or vibration displacement, and denote the first-order time derivative (velocity) and the second-order time derivative (acceleration) of the displacement, respectively; denotes the external excitation or equivalent load (such as electromagnetic force, wind force or impact caused by faults) acting on the structure at time ​​​is the value of the external force at the jth sampling time or state, is the number of time / samples involved in the constraint.

[0114] The target constraint function can also include an insulation degradation constraint residual term: In the insulation degradation constraint residual term, represents an insulation life or insulation health index (such as residual life, insulation strength, or equivalent aging factor), is the rate of change of insulation state over time, is the pre-exponential factor (frequency factor) in the Arrhenius model, determined in combination with device material properties and stress levels, is the apparent activation energy of insulation aging, is the universal gas constant, is the absolute temperature of the insulation medium or oil-paper system, and represent the insulation state and temperature values at the i th time point or working condition, is the number of time / samples involved in the constraint.

[0115] The target constraint function can include one or more of an electrical constraint residual term, a thermal conduction constraint residual term, a mechanical vibration dynamics constraint residual term, and an insulation degradation constraint residual term, which are not specifically limited in the present application.

[0116] As a possible implementation manner, in combination with Figure 1 , the substation fault determination apparatus sends a request message for requesting the target constraint function to the electronic device, and correspondingly, the electronic device receives the request message of the substation fault determination apparatus.

[0117] The electronic device generates a response message including the target constraint function, and sends the response message to the substation fault determination apparatus. Correspondingly, the substation fault determination apparatus receives the response message and obtains the target constraint function from the response message.

[0118] S702, the substation fault determination apparatus embeds the target constraint function into the loss function of the original graph model.

[0119] As a possible implementation manner, the substation fault determination apparatus embeds the target constraint function in the form of a partial differential equation residual into the loss function, and the specific embedding manner can refer to the existing scheme, which will not be described herein.

[0120] S703, the substation fault determination apparatus trains the original graph model based on the loss function of the original graph model to obtain a target graph model.

[0121] It should be noted that the substation fault determination apparatus can train the original graph model in an unsupervised manner, or the substation fault determination apparatus can also be in a supervised manner, in which case the substation fault determination apparatus can also obtain data labels, and train the original graph model based on the data labels.

[0122] For example, the data label can be a fault work order, which sets the identification of the electrical equipment, the fault type, and the start and end time of the fault time period. The fault work order can be used as a data label for the parameter value of the working parameter of the electrical equipment within the start and end time.

[0123] The training process can be implemented in the physical information neural network (PINN) paradigm. The specific training process can refer to the existing scheme, and the present application will not be described.

[0124] Based on S701-S703, by obtaining the target constraint function, embedding the target constraint function into the loss function of the original graph model, training the original graph model based on the loss function of the original graph model, and obtaining the target graph model, the physical limit of the substation equipment can be integrated into the model training process as an explicit optimization target, thereby ensuring that the output result of the model not only meets the accuracy requirements of data-driven, but also strictly meets the actual physical law. By embedding the target constraint function into the loss function of the original graph model, the model can learn data features and physical rules simultaneously during training, effectively avoiding invalid solutions that may be caused by relying solely on data that violate physical feasibility. This method improves the reliability and applicability of the model in real scenarios, and the generated target graph model can more stably serve the fault analysis of the substation equipment.

[0125] The above mainly introduces the scheme provided by the embodiments of the present application from the perspective of the substation fault determination apparatus executing the substation fault determination method. In order to realize the above functions, the substation fault determination apparatus includes the hardware structure and / or software module corresponding to each function. Those skilled in the art should easily realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in hardware or computer software driven hardware depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0126] The substation fault determination apparatus can be divided into functional modules according to the above method examples. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The integrated module can be implemented in the form of hardware or in the form of a software functional module. Alternatively, the division of the modules in the embodiments of the present application is illustrative, and is only a logical functional division. In actual implementation, another division manner can be used. In addition, the "module" can refer to an application-specific integrated circuit (ASIC), a circuit, a processor and a memory that execute one or more software or firmware programs, an integrated logic circuit, and / or other devices that can provide the above functions.

[0127] In the case of using functional module division, Figure 8 A structure diagram of a substation fault determination apparatus is shown. As Figure 8 shown, the substation fault determination apparatus 80 includes an acquisition module 801 and a processing module 802.

[0128] In some embodiments, the substation fault determination apparatus 80 can further include a storage module (not shown) for storing program instructions and data. Figure 8

[0129] ​The obtaining module 801 is configured to obtain an original data set of a transformer substation; the transformer substation includes a plurality of electrical devices; the original data set includes a plurality of sub-original data sets; the plurality of sub-original data sets correspond to the plurality of electrical devices in a one-to-one manner; each sub-original data set includes a plurality of working parameter original value sets of an electrical device; the plurality of working parameter original value sets correspond to a plurality of working parameters of the electrical device in a one-to-one manner; each working parameter original value set includes a plurality of parameter values of the working parameter of the electrical device in a first time period; the processing module 802 is configured to pre-process data in the original data set to obtain a target data set; the processing module 802 is further configured to input data in the target data set into a target graph model to obtain fault information of the transformer substation in the first time period; the target graph model includes a plurality of nodes and connection relationships between the plurality of nodes; the plurality of nodes correspond to the plurality of electrical devices in a one-to-one manner; the connection relationships between the plurality of nodes are connection relationships between the plurality of electrical devices; the target graph model is trained based on a historical data set of the transformer substation; the historical data set includes a plurality of sub-historical data sets; the plurality of sub-historical data sets correspond to the plurality of electrical devices in a one-to-one manner; each sub-historical data set includes a plurality of working parameter historical value sets of an electrical device; the plurality of working parameter historical value sets correspond to a plurality of working parameters of the electrical device in a one-to-one manner; each working parameter historical value set includes a plurality of parameter values of the working parameter of the electrical device in a historical time period; the target graph model includes a target constraint function; the target constraint function is used to indicate physical constraint information of the electrical device in operation in the transformer substation; and the fault information is used to indicate a fault electrical device in the plurality of electrical devices.

[0130] Optionally, before the obtaining module 801 is configured to obtain the original data set of the transformer substation, the processing apparatus is further configured to: obtain a historical data set of the transformer substation; determine feature information of the plurality of electrical devices in the historical time period according to the historical data set; and determine the target graph model according to the feature information.

[0131] Optionally, the feature information includes data statistical feature information and first feature vector information; the first feature vector information is determined based on a preset deep learning model; the processing module 802 is configured to determine the feature information of the plurality of electrical devices in the historical time period according to the historical data set, including: for each electrical device in the plurality of electrical devices, determining a mathematical statistical feature of a working parameter corresponding to each working parameter historical value set in the sub-historical data set corresponding to the electrical device according to the working parameter historical value set; and inputting parameter values in each working parameter historical value set in the sub-historical data set corresponding to the electrical device into the preset deep learning model to obtain first feature vector information of each working parameter input by the preset deep learning model.

[0132] Optionally, the processing module 802 is configured to determine the target graph model according to the feature information, including: for each of the plurality of electrical devices, inputting the mathematical statistical feature of each of the plurality of working parameters of the electrical device and the first feature vector information into a preset self-attention model to obtain second feature vector information of the electrical device output by the preset self-attention model; and determining the target graph model according to the second feature vector information of each of the plurality of electrical devices.

[0133] Optionally, the processing module 802 is configured to train the original graph model according to the second feature vector information of each of the plurality of electrical devices to obtain the target graph model, including: obtaining the electrical connection relationship between the plurality of electrical devices; taking the second feature vector of each of the electrical devices as the feature of the node and taking the electrical connection relationship as the connection relationship between the nodes to obtain the original graph model; and training the original graph model to obtain the target graph model.

[0134] Optionally, the processing module 802 is configured to train the original graph model to obtain the target graph model, including: obtaining a target constraint function; embedding the target constraint function into a loss function of the original graph model; and training the original graph model based on the loss function of the original graph model to obtain the target graph model.

[0135] Optionally, the processing module 802 is configured to pre-process the data in the original data set to obtain the target data set, including: performing data cleaning on the data in the original data set to obtain an intermediate data set; and performing normalization and dimension unification processing on the data in the intermediate data set to obtain the target data set.

[0136] All related contents of each step involved in the above method embodiments can be cited to the function description of the corresponding function module, which will not be repeated here.

[0137] In the case of realizing the functions of the above function modules in the form of hardware, Figure 9 Another structure diagram of a substation fault determination device is shown. As Figure 9 shown, the substation fault determination device 90 includes a processor 901, a memory 902 and a bus 903. The processor 901 and the memory 902 can be connected through the bus 903.

[0138] The processor 901 is the control center of the substation fault determination device 90, which can be one processor or a general term of multiple processing elements. For example, the processor 901 can be a general central processing unit (CPU), or other general-purpose processors, etc. Among them, the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0139] As one embodiment, processor 901 may include one or more CPUs, for example Figure 9 CPU 0 and CPU 1 are shown in the diagram.

[0140] The memory 902 may be a read-only memory (ROM) or other type of static storage electrical device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage electrical device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage medium or other magnetic storage electrical device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0141] As one possible implementation, the memory 902 can exist independently of the processor 901. The memory 902 can be connected to the processor 901 via a bus 903 and is used to store instructions or program code. When the processor 901 calls and executes the instructions or program code stored in the memory 902, it can implement the substation fault determination method provided in this application embodiment.

[0142] In another possible implementation, the memory 902 can also be integrated with the processor 901.

[0143] Bus 903 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0144] It should be pointed out that, Figure 9 The structure shown does not constitute a limitation on the substation fault determination device 90. Except... Figure 9 In addition to the components shown, the substation fault determination device 90 may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0145] As an example, combinedFigure 8 The functions implemented by the obtaining module 801 and the processing module 802 in the substation fault determination apparatus 80 are the same as the functions of the processor 901 in the substation fault determination apparatus 80. Figure 9

[0146] Optionally, as shown in Figure 9 , the substation fault determination apparatus 90 provided by the embodiment of the present application can further include a communication interface 904.

[0147] The communication interface 904 is configured to connect with other electrical equipment through a communication network. The communication network can be an Ethernet, a wireless access network, a wireless local area network (WLAN), etc. The communication interface 904 can include a receiving unit for receiving data, and a sending unit for sending data.

[0148] In a possible implementation, in the substation fault determination apparatus 90 provided by the embodiment of the present application, the communication interface 904 can be integrated in the processor 901, which is not limited in the embodiment of the present application.

[0149] As a possible product form, the substation fault determination apparatus provided by the embodiment of the present application can also be implemented by using one or more field programmable gate arrays (FPGA), programmable logic devices (PLD), controllers, state machines, gate logic, discrete hardware components, any other suitable circuits, or any combination of circuits capable of performing the various functions described throughout this application.

[0150] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of functional units is exemplified. In actual application, the above-mentioned functions can be completed by different functional units according to needs, that is, the internal structure of the apparatus is divided into different functional units to complete all or part of the above-described functions. The specific working process of the above-described system, apparatus and unit can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0151] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program or instructions, and the computer program or instructions are executed to make the computer execute each step in the method flow shown in the foregoing method embodiments.

[0152] ​The embodiment of the present application provides a computer program product comprising instructions which, when executed on a computer, cause the computer to carry out each of the steps in the method procedure shown in the above method embodiment.

[0153] The embodiment of the present application provides a chip system, comprising: a processor and an interface circuit; the interface circuit is used for receiving a computer program or instructions and transmitting to the processor; the processor is used for executing the computer program or instructions, so that the chip system executes each of the steps in the method procedure shown in the above method embodiment.

[0154] In some embodiments, a computer readable storage medium can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any combination of the above. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a register, a hard disk, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above, or any other form of computer readable storage medium known in the art. An exemplary storage medium is coupled to the processor, so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application-specific integrated circuit (ASIC) in the present embodiment. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus or device.

[0155] Since the substation fault determination device, the computer readable storage medium and the computer program product provided by the embodiment can be applied to the substation fault determination method provided by the embodiment, the technical effects that can be obtained by the substation fault determination device, the computer readable storage medium and the computer program product are also referable to the above method embodiment, and the present embodiment will not be described here again.

[0156] Although the present application is described herein in conjunction with various embodiments, other variations of the disclosed embodiments can be understood and effected by those skilled in the art in the practice of the claimed application, from the appended drawings, the disclosure and the specific description of the features.

[0157] Although the present application has been described in connection with certain specific features and embodiments thereof, it is to be understood that it is provided as an example to assist those skilled in the art and is not intended to limit the scope of the application, which is outlined by the appended claims. Accordingly, no reference signs in the claims should be considered as limiting the scope of the claims. The disclosure of all patents and publications cited above are hereby incorporated by reference.

Claims

1. A method for determining faults in a substation, characterized in that, The method includes: Obtain the original data set of the substation; the substation includes multiple electrical devices, the original data set includes multiple sub-original data sets, each sub-original data set corresponds one-to-one with the multiple electrical devices, each sub-original data set includes multiple sets of original values ​​of operating parameters of the electrical devices, each set of original values ​​of operating parameters corresponds one-to-one with multiple operating parameters of the electrical devices, and each set of original values ​​of operating parameters includes multiple parameter values ​​of the operating parameters of the electrical devices within a first time period; The target data set is obtained by preprocessing the data in the original data set; The data from the target dataset is input into the target graph model to obtain the fault information of the substation within the first time period. The target graph model includes multiple nodes and the connection relationships between the multiple nodes. Each node corresponds one-to-one with a multiple electrical device, and the connection relationships between the multiple nodes are the connection relationships between the multiple electrical devices. The target graph model is trained based on the historical dataset of the substation. The historical dataset includes multiple sub-historical datasets, each corresponding one-to-one with a multiple electrical device. Each sub-historical dataset includes a set of historical values ​​of multiple operating parameters of the electrical device, each corresponding one-to-one with multiple operating parameters of the electrical device. The set of historical operating parameter values ​​includes multiple parameter values ​​of the operating parameters of the electrical device within a historical time period. The target graph model includes a target constraint function, which is used to indicate the physical constraint information for the operation of the electrical devices in the substation. The fault information is used to indicate that a faulty electrical device among the multiple electrical devices has occurred.

2. The method according to claim 1, characterized in that, Before acquiring the original data set of the substation, the method further includes: Obtain the historical data set of the substation; Based on the historical data set, the characteristic information of the multiple electrical devices within the historical time period is determined; The target graph model is determined based on the feature information.

3. The method according to claim 2, characterized in that, The feature information includes data statistical feature information and first feature vector information. The first feature vector information is determined based on a preset deep learning model. The feature information of the multiple electrical devices within the historical time period is determined based on the historical data set, including: For each of the plurality of electrical devices, based on the historical value set of each working parameter in the sub-historical data set corresponding to the electrical device, the mathematical statistical characteristics of the working parameters corresponding to the historical value set of the working parameters are determined; The parameter values ​​in the historical value set of each working parameter in the sub-historical data set corresponding to the electrical equipment are input into the preset deep learning model to obtain the first feature vector information of each working parameter input into the preset deep learning model.

4. The method according to claim 3, characterized in that, Determining the target graph model based on the feature information includes: For each of the plurality of electrical devices, the mathematical statistical features and first feature vector information of each of the plurality of operating parameters of the electrical device are input into a preset self-attention model to obtain the second feature vector information of the electrical device output by the preset self-attention model; The target graph model is determined based on the second feature vector information of each of the plurality of electrical devices.

5. The method according to claim 4, characterized in that, The original graph model is trained based on the second feature vector information of each of the plurality of electrical devices to obtain the target graph model, including: Obtain the electrical connection relationships between the plurality of electrical devices; The second feature vector of each electrical device is used as the feature of the node, and the electrical connection relationship is used as the connection relationship between the nodes to obtain the original graph model; The original graph model is trained to obtain the target graph model.

6. The method according to claim 5, characterized in that, The original graph model is trained to obtain the target graph model, including: Obtain the target constraint function; The objective constraint function is embedded into the loss function of the original graphical model; The original graph model is trained based on the loss function of the original graph model to obtain the target graph model.

7. The method according to any one of claims 1-6, characterized in that, The target dataset is obtained by preprocessing the data in the original dataset, including: The data in the original dataset is cleaned to obtain an intermediate dataset; The data in the intermediate dataset is normalized and dimensionally unified to obtain the target dataset.

8. A substation fault determination device, characterized in that, The device includes: an acquisition module and a processing module; The acquisition module is used to acquire the original data set of the substation; the substation includes multiple electrical devices, the original data set includes multiple sub-original data sets, each sub-original data set corresponds one-to-one with the multiple electrical devices, each sub-original data set includes multiple sets of original values ​​of operating parameters of the electrical devices, each set of original values ​​of operating parameters corresponds one-to-one with multiple operating parameters of the electrical devices, and each set of original values ​​of operating parameters includes multiple parameter values ​​of the operating parameters of the electrical devices within a first time period; The processing module is used to preprocess the data in the original data set to obtain the target data set; The processing module is further configured to input the data from the target data set into the target graph model to obtain the fault information of the substation within the first time period; the target graph model includes multiple nodes and the connection relationships between the multiple nodes, each node corresponding to one of the multiple electrical devices, and the connection relationships between the multiple nodes being the connection relationships between the multiple electrical devices; the target graph model is trained based on the historical data set of the substation, the historical data set including multiple sub-historical data sets, each sub-historical data set corresponding to one of the multiple electrical devices, each sub-historical data set including multiple sets of historical values ​​of operating parameters of the electrical devices, each set of historical values ​​of operating parameters corresponding to multiple operating parameters of the electrical devices, each set of historical values ​​of operating parameters including multiple parameter values ​​of the operating parameters of the electrical devices within a historical time period; the target graph model includes a target constraint function, the target constraint function being used to indicate the physical constraint information of the operation of the electrical devices in the substation; the fault information is used to indicate that a faulty electrical device among the multiple electrical devices has failed.

9. A substation fault determination device, characterized in that, The substation fault determination device includes: a processor coupled to a memory for storing programs or instructions, which, when executed by the processor, cause the device to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed, they cause the computer to perform the method as described in any one of claims 1 to 7.

Images