Method and device for positioning turn-to-turn short circuit fault position of transformer, medium and terminal

Abstract

The invention discloses a method and a device for positioning a turn-to-turn short circuit fault position of a transformer, a medium and a terminal, relates to the technical field of transformer fault diagnosis, and mainly aims to solve the problem of internal faults of the transformer caused by untimely response, low action sensitivity and incapability of accurately positioning the turn-to-turn short circuit fault position in the prior art. Comprising the steps that when it is detected that a turn-to-turn short circuit fault exists in a target transformer, micro-temperature field simulation data of the target transformer are obtained, and the micro-temperature field simulation data are obtained based on a micro-temperature field digital twinborn model of the target transformer; based on a convolutional neural network, extracting turn-to-turn short circuit fault characteristic parameters from the micro-temperature field simulation data; and determining the position of the turn-to-turn short circuit fault of the target transformer based on a Bayesian reasoning method according to the turn-to-turn short circuit fault characteristic parameter.

Description

Technical Field

[0001] The present application relates to the technical field of transformer fault diagnosis, and in particular to a method and device, medium, and terminal for locating the position of a transformer turn-to-turn short-circuit fault. Background Art

[0002] Power transformers are core equipment in power system equipment. If a fault occurs, it will seriously affect the reliability of power supply and the safety of operation of the power system. Among transformer faults, short circuit between winding turns is a common fault type of power transformers. Short circuit between winding turns can easily lead to problems such as aging of solid insulation and diffusion of local hot spots.

[0003] In order to avoid the above problems, at present, most projects use gas protection, differential protection, and dissolved gas in oil as the main protection for transformers.

[0004] However, since the action time of gas protection is affected by factors such as insulating oil and the location of turn-to-turn short-circuit fault, it is very easy to cause the problem of untimely response, resulting in internal faults of the transformer; and the differential protection can basically only respond to more than 3% of the turn-to-turn short-circuit in terms of the number of turns, and the action sensitivity is low, which in turn causes internal faults of the transformer; the dissolved gas in the oil determines the abnormality of the transformer through the change of the gas and the composition of the gas when the fault occurs, which makes it impossible to accurately locate the turn-to-turn short-circuit fault location. Therefore, a method for locating the position of the transformer turn-to-turn short-circuit fault is urgently needed to solve the above problems. Summary of the invention

[0005] In view of this, the present application provides a method and device, medium, and terminal for locating the position of a transformer inter-turn short-circuit fault. The main purpose is to solve the problem of internal faults in the transformer caused by untimely response, low action sensitivity, and inability to accurately locate the position of the inter-turn short-circuit fault.

[0006] According to one aspect of the present application, a method for locating a transformer turn-to-turn short circuit fault position is provided, comprising:

[0007] When a turn-to-turn short circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer;

[0008] Based on a convolutional neural network, extracting turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data;

[0009] According to the turn-to-turn short-circuit fault characteristic parameters, a location where the turn-to-turn short-circuit fault occurs in the target transformer is determined based on a Bayesian reasoning method.

[0010] Preferably, when a turn-to-turn short circuit fault is detected in the target transformer, before obtaining the micro-temperature field simulation data of the target transformer, the method further includes:

[0011] Based on the electromagnetic field differential equation and the natural convection heat transfer mathematical model, a thermal field and flow field coupling model is constructed. The electromagnetic field differential equation is expressed as the following formula:

[0012]

[0013] in, represents the vector differential operator, H represents the magnetic field intensity, J represents the current density, D represents the electric displacement vector, t represents the time, B represents the magnetic induction intensity, E represents the electric field intensity, p represents the charge density,

[0014] The natural convection heat transfer mathematical model is expressed as the following formula

[0015] q=h(T b -T o )

[0016] Where q represents the heat flux density, h represents the convective heat transfer coefficient, T b Indicates the transformer tank temperature, T o Indicates the air temperature;

[0017] The loss value is calculated based on the core and winding loss formula, which is expressed as the following formula

[0018]

[0019] Among them, P 总 Represents the loss value, P fe0 It represents the core loss per unit volume before the winding is short-circuited, B represents the maximum magnetic flux density in the core before the winding is short-circuited, B0 represents the maximum magnetic flux density in the core after the winding is short-circuited, P wN Indicates the loss per unit volume of the winding at rated current, I b Indicates the winding current with short circuit between turns, I N Indicates the rated current of the winding with short circuit between turns, I sh Represents the equivalent fault resistance current of turn-to-turn short circuit, R sh It represents the equivalent fault resistance value of inter-turn short circuit, n represents the proportion of inter-turn short circuit, and V represents the volume of fault winding;

[0020] The loss value and the preset inlet flow velocity value in the flow field are input into the thermal field and flow field coupling model to obtain a multi-field coupling simulation model for the target transformer.

[0021] Preferably, after obtaining the multi-field coupling simulation model for the target transformer, the method further includes:

[0022] Acquire real signals collected by various sensors arranged in the micro-temperature field of the target transformer in real time;

[0023] The parameters of the multi-field coupling simulation model are corrected using the real signal until a preset accuracy threshold is met, so as to obtain a micro-temperature field digital twin model of the target transformer.

[0024] Preferably, the method of determining the location where the turn-to-turn short-circuit fault of the target transformer occurs based on the Bayesian reasoning method according to the turn-to-turn short-circuit fault characteristic parameter includes:

[0025] Based on the total probability calculation formula, the total probability value of observing the turn-to-turn short circuit fault characteristic is calculated according to the turn-to-turn short circuit fault characteristic parameter. The total probability calculation formula is expressed as the following formula

[0026]

[0027] Among them, Y n It represents the characteristic parameter of turn-to-turn short-circuit fault, P(Y n ) represents the total probability of observing the characteristics of turn-to-turn short-circuit faults, Indicates the frequency of occurrence of the characteristic corresponding to the inter-turn short-circuit fault, X m Indicates that a turn-to-turn open circuit fault occurs in area m, where m represents areas A, B, and C. Indicates the total frequency of turn-to-turn short-circuit faults in the three areas A, B, and C;

[0028] Based on the following formula, the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition of turn-to-turn short-circuit fault occurring in each area is calculated:

[0029]

[0030] Among them, P(Y n |X m ) represents the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in region m, It indicates the probability value of turn-to-turn short circuit fault occurring in the three regions A, B and C respectively;

[0031] Based on the Bayesian fault location formula, according to the total probability value of the observed turn-to-turn short-circuit fault characteristics and the probability value of the observed turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in each area, the probability value of the turn-to-turn short-circuit fault occurring in each area under the condition that the turn-to-turn short-circuit fault characteristics are observed is calculated respectively. The Bayesian fault location formula is expressed as the following formula

[0032]

[0033] Among them, P(X m |Y n ) represents the probability value of turn-to-turn short-circuit fault occurring in region m under the condition that the turn-to-turn short-circuit fault characteristics are observed, P(X m ) represents the preset prior probability value of the turn-to-turn short circuit fault occurring in region m;

[0034] An area corresponding to a maximum probability value is selected from the probability values ​​of the turn-to-turn short-circuit fault, and the area is determined as a location where the turn-to-turn short-circuit fault of the target transformer occurs.

[0035] Preferably, the micro-temperature field simulation data includes time series parameter data and image parameter data, and the method based on the convolutional neural network to extract the turn-to-turn short circuit fault characteristic parameters from the micro-temperature field simulation data includes:

[0036] Based on a preset matrix rule, the timing parameter data is arranged and processed to obtain a timing parameter matrix that conforms to the preset matrix rule;

[0037] Converting the time series parameter matrix into a form so that the time series parameter matrix and the data image have the same representation form;

[0038] Extracting a first turn-to-turn short-circuit fault characteristic parameter corresponding to the timing parameter from the timing parameter after the form conversion based on a first convolutional neural network;

[0039] Extracting a second turn-to-turn short-circuit fault characteristic parameter corresponding to the image parameter from the image parameter data based on a second convolutional neural network;

[0040] Based on a preset weight ratio, the first turn-to-turn short-circuit fault characteristic parameter and the second turn-to-turn short-circuit fault characteristic parameter are fused to obtain the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

[0041] Preferably, when a turn-to-turn short circuit fault is detected in the target transformer, obtaining micro-temperature field simulation data of the target transformer includes:

[0042] When a turn-to-turn short circuit fault is detected in the target transformer, a target sensor group is selected according to the comprehensive score coefficients of the sensors arranged in the micro-temperature field of the target transformer, wherein the comprehensive score coefficient is obtained by fusing the basic weight vectors of the correlation between the sensor temperature and the hotspot temperature of each sensor, the basic weight vectors of the reliability of the sensor data, and the basic weight vectors of the symmetry between the sensor signals to obtain a comprehensive score vector, and performing weight distribution based on a game weighting algorithm;

[0043] The micro-temperature field simulation data corresponding to each sensor in the target sensor group is obtained as the micro-temperature field simulation data of the target transformer to locate the turn-to-turn short-circuit fault position.

[0044] Preferably, when a turn-to-turn short circuit fault is detected in the target transformer, a target sensor group is screened out according to comprehensive score coefficients of various sensors arranged in the micro-temperature field of the target transformer, including:

[0045] When a turn-to-turn short circuit fault is detected in the target transformer, the basic weight vector of the correlation between the sensor temperature and the hotspot temperature is calculated based on the Pearson correlation coefficient formula, and the Pearson correlation coefficient formula is expressed as follows:

[0046]

[0047] Where r represents the correlation vector between sensor temperature and hotspot temperature, M represents the number of sensors, and x j represents the hot spot temperature in the jth sensor, represents the average temperature of the hot spot in the sensor, y j represents the temperature of the jth sensor, Indicates the average temperature of the sensor;

[0048] Determining a sensor data reliability basic weight vector based on the signal variance of each of the sensors;

[0049] Based on the symmetry index calculation formula, the symmetry index between each pair of sensor signals is calculated to generate the symmetry basic weight vector between the sensor signals. The symmetry index calculation formula is expressed as the following formula:

[0050] DS=R*ρ

[0051] Where DS represents the symmetry index between each pair of sensor signals, R represents the distance coefficient, R = -ln(|r1-r2|), r1 and r2 represent the root mean square values ​​of the current monitoring points of the two sensors, and ρ represents the Spearman correlation coefficient;

[0052] The basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are fused to obtain a comprehensive score vector

[0053]

[0054] Among them, W represents the comprehensive score vector, λ α represents the weight coefficient, α=1 represents the basic weight vector of the correlation between sensor temperature and hot spot temperature, α=2 represents the basic weight vector of sensor data reliability, α=3 represents the basic weight vector of symmetry between sensor signals, and T represents the combination coefficient;

[0055] The weight coefficient is optimized based on a game weighting algorithm so that the deviations between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are all minimum values, thereby obtaining the optimal weight coefficient;

[0056] Normalizing the optimal weight coefficient to obtain a comprehensive score coefficient of each sensor;

[0057] Based on preset rules, a target sensor group is screened out according to the comprehensive score coefficients of each sensor, so as to locate the turn-to-turn short circuit fault position according to the micro-temperature field simulation data corresponding to each sensor in the target sensor group.

[0058] According to another aspect of the present application, a device for locating a transformer turn-to-turn short-circuit fault position is provided, comprising:

[0059] A micro-temperature field simulation data acquisition module is used to acquire micro-temperature field simulation data of the target transformer when a turn-to-turn short circuit fault is detected in the target transformer, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer;

[0060] A turn-to-turn short-circuit fault characteristic parameter extraction module, used to extract turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data based on a convolutional neural network;

[0061] The turn-to-turn short-circuit fault location positioning module is used to determine the location of the turn-to-turn short-circuit fault of the target transformer based on the Bayesian reasoning method according to the turn-to-turn short-circuit fault characteristic parameters.

[0062] Preferably, before the micro-temperature field simulation data acquisition module, the device further includes a micro-temperature field digital twin model construction module, which is used to:

[0063] Based on the electromagnetic field differential equation and the natural convection heat transfer mathematical model, a thermal field and flow field coupling model is constructed. The electromagnetic field differential equation is expressed as the following formula:

[0064]

[0065] in, represents the vector differential operator, H represents the magnetic field intensity, J represents the current density, D represents the electric displacement vector, t represents the time, B represents the magnetic induction intensity, E represents the electric field intensity, p represents the charge density,

[0066] The natural convection heat transfer mathematical model is expressed as the following formula

[0067] q=h(T b -T o )

[0068] Where q represents the heat flux density, h represents the convective heat transfer coefficient, T b Indicates the transformer tank temperature, T o Indicates the air temperature;

[0069] The loss value is calculated based on the core and winding loss formula, which is expressed as the following formula

[0070]

[0071] Among them, P 总 Represents the loss value, P fe0 It represents the core loss per unit volume before the winding is short-circuited, B represents the maximum magnetic flux density in the core before the winding is short-circuited, B0 represents the maximum magnetic flux density in the core after the winding is short-circuited, P wN Indicates the loss per unit volume of the winding at rated current, I b Indicates the winding current with short circuit between turns, I N Indicates the rated current of the winding with short circuit between turns, I sh Represents the equivalent fault resistance current of turn-to-turn short circuit, R sh It represents the equivalent fault resistance value of inter-turn short circuit, n represents the proportion of inter-turn short circuit, and V represents the volume of fault winding;

[0072] The loss value and the preset inlet flow velocity value in the flow field are input into the thermal field and flow field coupling model to obtain a multi-field coupling simulation model for the target transformer.

[0073] Preferably, the micro-temperature field digital twin model construction module is also used for:

[0074] Acquire real signals collected by various sensors arranged in the micro-temperature field of the target transformer in real time;

[0075] The parameters of the multi-field coupling simulation model are corrected using the real signal until a preset accuracy threshold is met, so as to obtain a micro-temperature field digital twin model of the target transformer.

[0076] Preferably, the turn-to-turn short-circuit fault location positioning module is used to:

[0077] Based on the total probability calculation formula, the total probability value of observing the turn-to-turn short circuit fault characteristic is calculated according to the turn-to-turn short circuit fault characteristic parameter. The total probability calculation formula is expressed as the following formula

[0078]

[0079] Among them, Y n It represents the characteristic parameter of turn-to-turn short-circuit fault, P(Y n ) represents the total probability of observing the characteristics of turn-to-turn short-circuit faults, Indicates the frequency of occurrence of the characteristic corresponding to the inter-turn short-circuit fault, X m Indicates that a turn-to-turn open circuit fault occurs in area m, where m represents areas A, B, and C. Indicates the total frequency of turn-to-turn short-circuit faults in the three areas A, B, and C;

[0080] Based on the following formula, the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition of turn-to-turn short-circuit fault occurring in each area is calculated:

[0081]

[0082] Among them, P(Y n |X m ) represents the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in region m, It indicates the probability value of turn-to-turn short circuit fault occurring in the three regions A, B and C respectively;

[0083] Based on the Bayesian fault location formula, according to the total probability value of the observed turn-to-turn short-circuit fault characteristics and the probability value of the observed turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in each area, the probability value of the turn-to-turn short-circuit fault occurring in each area under the condition that the turn-to-turn short-circuit fault characteristics are observed is calculated respectively. The Bayesian fault location formula is expressed as the following formula

[0084]

[0085] Among them, P(X m |Y n) represents the probability value of turn-to-turn short-circuit fault occurring in region m under the condition that the turn-to-turn short-circuit fault characteristics are observed, P(X m ) represents the preset prior probability value of the turn-to-turn short circuit fault occurring in region m;

[0086] An area corresponding to a maximum probability value is selected from the probability values ​​of the turn-to-turn short-circuit fault, and the area is determined as a location where the turn-to-turn short-circuit fault of the target transformer occurs.

[0087] Preferably, the micro-temperature field simulation data includes time series parameter data and image parameter data, and the turn-to-turn short circuit fault characteristic parameter extraction module is used to:

[0088] Based on a preset matrix rule, the timing parameter data is arranged and processed to obtain a timing parameter matrix that conforms to the preset matrix rule;

[0089] Converting the time series parameter matrix into a form so that the time series parameter matrix and the data image have the same representation form;

[0090] Extracting a first turn-to-turn short-circuit fault characteristic parameter corresponding to the timing parameter from the timing parameter after the form conversion based on a first convolutional neural network;

[0091] Extracting a second turn-to-turn short-circuit fault characteristic parameter corresponding to the image parameter from the image parameter data based on a second convolutional neural network;

[0092] Based on a preset weight ratio, the first turn-to-turn short-circuit fault characteristic parameter and the second turn-to-turn short-circuit fault characteristic parameter are fused to obtain the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

[0093] Preferably, the micro-temperature field simulation data acquisition module includes:

[0094] A target sensor group screening unit is used to screen out a target sensor group according to the comprehensive score coefficients of each sensor arranged in the micro-temperature field of the target transformer when a turn-to-turn short circuit fault is detected in the target transformer, wherein the comprehensive score coefficient is obtained by fusing the basic weight vectors of the correlation between the sensor temperature and the hotspot temperature of each sensor, the basic weight vectors of the reliability of the sensor data, and the basic weight vectors of the symmetry between the sensor signals to obtain a comprehensive score vector, and performing weight distribution based on a game weighting algorithm;

[0095] The micro-temperature field simulation data acquisition unit is used to acquire the micro-temperature field simulation data corresponding to each sensor in the target sensor group as the micro-temperature field simulation data of the target transformer to locate the turn-to-turn short-circuit fault position.

[0096] Preferably, the target sensor group screening unit is used to:

[0097] When a turn-to-turn short circuit fault is detected in the target transformer, the basic weight vector of the correlation between the sensor temperature and the hotspot temperature is calculated based on the Pearson correlation coefficient formula, and the Pearson correlation coefficient formula is expressed as follows:

[0098]

[0099] Where r represents the correlation vector between sensor temperature and hotspot temperature, M represents the number of sensors, and x j represents the hot spot temperature in the jth sensor, x represents the average hot spot temperature in the sensor, y j represents the temperature of the jth sensor, and y represents the average temperature of the sensor;

[0100] Determining a sensor data reliability basic weight vector based on the signal variance of each of the sensors;

[0101] Based on the symmetry index calculation formula, the symmetry index between each pair of sensor signals is calculated to generate the symmetry basic weight vector between the sensor signals. The symmetry index calculation formula is expressed as the following formula:

[0102] DS=R*ρ

[0103] Where DS represents the symmetry index between each pair of sensor signals, R represents the distance coefficient, R = -ln(|r1-r2|), r1 and r2 represent the root mean square values ​​of the current monitoring points of the two sensors, and ρ represents the Spearman correlation coefficient;

[0104] The basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are fused to obtain a comprehensive score vector

[0105]

[0106] Among them, W represents the comprehensive score vector, λ α represents the weight coefficient, α=1 represents the basic weight vector of the correlation between sensor temperature and hot spot temperature, α=2 represents the basic weight vector of sensor data reliability, α=3 represents the basic weight vector of symmetry between sensor signals, and T represents the combination coefficient;

[0107] The weight coefficient is optimized based on a game weighting algorithm so that the deviations between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are all minimum values, thereby obtaining the optimal weight coefficient;

[0108] Normalizing the optimal weight coefficient to obtain a comprehensive score coefficient of each sensor;

[0109] Based on preset rules, a target sensor group is screened out according to the comprehensive score coefficients of each sensor, so as to locate the turn-to-turn short circuit fault position according to the micro-temperature field simulation data corresponding to each sensor in the target sensor group.

[0110] According to another aspect of the present application, a storage medium is provided, in which at least one executable instruction is stored, and the executable instruction enables a processor to execute operations corresponding to the method for locating the position of a transformer inter-turn short-circuit fault as described above.

[0111] According to another aspect of the present application, a terminal is provided, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0112] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for locating the position of the transformer inter-turn short-circuit fault.

[0113] By means of the above technical solution, the technical solution provided by the embodiment of the present application has at least the following advantages:

[0114] The present application provides a method and device, medium, and terminal for locating the position of a transformer turn-to-turn short-circuit fault. First, when a turn-to-turn short-circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained. The micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer. The micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer; secondly, based on a convolutional neural network, characteristic parameters of the turn-to-turn short-circuit fault are extracted from the micro-temperature field simulation data; finally, based on the characteristic parameters of the turn-to-turn short-circuit fault, the position of the turn-to-turn short-circuit fault of the target transformer is determined based on a Bayesian reasoning method. Compared with the prior art, the embodiment of the present application, based on the multi-field coupling simulation model of the target transformer, uses the real signal collected by the sensor arranged in the micro-temperature field of the target transformer to correct the micro-temperature field digital twin model, so that the simulation data obtained based on the micro-temperature field digital twin model is highly similar to the actual operation of the target transformer. On this basis, the simulation data is used to locate the fault in advance before the fault occurs, thereby avoiding the problem of internal fault of the transformer caused by untimely response; further, based on the convolutional neural network, the turn-to-turn short-circuit fault characteristic parameters are extracted from the micro-temperature field simulation data, and by utilizing the high sensitivity characteristics of the convolutional neural network, the problem of not being able to detect the fault characteristics in time due to the slow temperature change is overcome, the recognition sensitivity of the fault characteristics is improved, and then the action sensitivity is improved; finally, by using the Bayesian reasoning method to determine the exact location of the turn-to-turn short-circuit fault according to the turn-to-turn short-circuit fault characteristic parameters, the defect of the existing method that the turn-to-turn short-circuit fault position cannot be accurately located is overcome.

[0115] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0116] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0117] Figure 1 A flow chart of a method for locating a transformer turn-to-turn short-circuit fault position provided in an embodiment of the present application is shown;

[0118] Figure 2A flow chart of another method for locating a transformer turn-to-turn short-circuit fault position provided in an embodiment of the present application is shown;

[0119] Figure 3 A schematic diagram of the heat dissipation process of the transformer micro-temperature field provided in an embodiment of the present application is shown;

[0120] Figure 4 An equivalent circuit diagram of a transformer provided in an embodiment of the present application is shown;

[0121] Figure 5 A schematic diagram of grid division provided in an embodiment of the present application is shown;

[0122] Figure 6 A comparison curve diagram of the experimental measurement results and the simulation results of the winding temperature under normal conditions provided by the embodiment of the present application is shown;

[0123] Figure 7 A comparison curve diagram of the experimental measurement results and the simulation results of the winding temperature under the fault state provided by the embodiment of the present application is shown;

[0124] Figure 8 A block diagram of a device for locating a transformer turn-to-turn short-circuit fault position provided by an embodiment of the present application is shown;

[0125] Fig. 9 A schematic diagram of the structure of a terminal provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0126] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0127] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0128] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0129] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0130] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0131] The embodiments of the present application can be applied to computer systems / servers, which can operate with many other general or special computing system environments or configurations. Examples of well-known computing systems, environments and / or configurations suitable for use with computer systems / servers include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0132] Computer systems / servers may be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. In general, program modules may include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers may be implemented in a distributed cloud computing environment where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules may be located on local or remote computing system storage media including storage devices.

[0133] The present application embodiment provides a method for locating a transformer turn-to-turn short circuit fault position, such as Figure 1 As shown, the method includes:

[0134] 101. When a turn-to-turn short circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained.

[0135] Among them, the micro-temperature field simulation data is obtained based on the micro-temperature field digital twin model of the target transformer; the micro-temperature field digital twin model is obtained by correcting the real signal collected by the sensor arranged in the micro-temperature field of the target transformer on the basis of the multi-field coupling simulation model for the target transformer. It should be noted that since the micro-temperature field simulation data is not limited by factors such as the sensor position, data at any position can be obtained, so that the accuracy of fault location positioning based on the simulation data is higher; further, in the embodiment of the present application, on the basis of the multi-field coupling simulation model for the target transformer, the real signal collected by the sensor arranged in the micro-temperature field of the target transformer is used to correct the micro-temperature field digital twin model, so that the simulation data obtained based on the micro-temperature field digital twin model is highly similar to the actual operation of the target transformer. On this basis, the simulation data is used to locate the fault in advance before the fault occurs, avoiding the problem of internal fault of the transformer caused by untimely response. In the embodiment of the present application, the current execution end can be a turn-to-turn short circuit fault detection module in the transformer fault detection system, which can be used to locate the occurrence position of the turn-to-turn short circuit fault.

[0136] 102. Based on convolutional neural network, the characteristic parameters of turn-to-turn short-circuit fault are extracted from micro-temperature field simulation data.

[0137] Since temperature change is a slow process, it is very easy to be ignored, making it difficult to find fault features in time. In order to overcome this problem, in the embodiment of the present application, by utilizing the high sensitivity of the convolutional neural network, the recognition sensitivity of the fault features is improved, thereby improving the sensitivity of the action.

[0138] 103. According to the characteristic parameters of the inter-turn short-circuit fault, the location of the inter-turn short-circuit fault of the target transformer is determined based on the Bayesian reasoning method.

[0139] Among them, the Bayesian reasoning method is a statistical method based on probability density, which is used to describe the mapping relationship between characteristic parameters and faults, and is continuously adjusted according to new information to finally approach the facts. In the embodiment of the present application, the characteristic parameters of the turn-to-turn short-circuit fault extracted from the micro-temperature field simulation data are used as the new information required by the Bayesian reasoning method to participate in the iteration, so as to determine the exact location of the turn-to-turn short-circuit fault, thereby overcoming the defect of the existing method that the turn-to-turn short-circuit fault location cannot be accurately located.

[0140] Compared with the prior art, the embodiment of the present application, based on the multi-field coupling simulation model of the target transformer, uses the real signal collected by the sensor arranged in the micro-temperature field of the target transformer to correct the micro-temperature field digital twin model, so that the simulation data obtained based on the micro-temperature field digital twin model is highly similar to the actual operation of the target transformer. On this basis, the simulation data is used to locate the fault in advance before the fault occurs, thereby avoiding the problem of internal fault of the transformer caused by untimely response; further, based on the convolutional neural network, the turn-to-turn short-circuit fault characteristic parameters are extracted from the micro-temperature field simulation data, and by utilizing the high sensitivity characteristics of the convolutional neural network, the problem of not being able to detect the fault characteristics in time due to the slow temperature change is overcome, the recognition sensitivity of the fault characteristics is improved, and then the action sensitivity is improved; finally, by using the Bayesian reasoning method to determine the exact location of the turn-to-turn short-circuit fault according to the turn-to-turn short-circuit fault characteristic parameters, the defect of the existing method that the turn-to-turn short-circuit fault position cannot be accurately located is overcome.

[0141] The present application embodiment provides another method for locating the position of a transformer turn-to-turn short circuit fault, such as Figure 2 As shown, the method includes:

[0142] 201. Construct a micro-temperature field digital twin model of the target transformer.

[0143] In the embodiment of the present application, first, a multi-field coupling simulation model of the target transformer is constructed. Since the temperature distribution inside the transformer involves multi-physical field coupling, in which the electromagnetic field is used as the basis of the micro-temperature field, after the transformer is operated, the loss of the core and winding will generate heat to increase the oil temperature, resulting in a heat transfer process, and the heat dissipation process of the transformer micro-temperature field, such as Figure 3 As shown, since the heat of the iron core and winding in the transformer is mainly dissipated to the outer surface through heat conduction, when the temperature rises, the oil around the iron core and winding flows upward. At this time, the temperature of the upper surface is lower and flows downward, forming natural convection of the oil. When the heated oil flows to the heat sink, the temperature will be transferred to the heat sink, and the heat sink relies on natural convection and heat radiation to transfer to the external air, and the circulation realizes the heat dissipation of the transformer. Therefore, the heat exchange method between the transformer oil tank and the air mainly considers the natural convection heat transfer method in thermal convection. Based on this, specifically, based on the electromagnetic field differential equation and the natural convection heat transfer mathematical model, a thermal field and flow field coupling model is constructed, and the electromagnetic field differential equation is expressed as the following formula

[0144]

[0145] in, represents the vector differential operator, H represents the magnetic field intensity, J represents the current density, D represents the electric displacement vector, t represents the time, B represents the magnetic induction intensity, E represents the electric field intensity, p represents the charge density,

[0146] The mathematical model of natural convection heat transfer is expressed as the following formula

[0147] q=h(T b -T o )

[0148] Where q represents the heat flux density, h represents the convective heat transfer coefficient, T b Indicates the transformer tank temperature, T o Indicates the air temperature;

[0149] The loss value is calculated based on the core and winding loss formula. The core and winding loss formula is expressed as the following formula

[0150]

[0151] Among them, P 总 Represents the loss value, P fe0 It represents the core loss per unit volume before the winding is short-circuited, B represents the maximum magnetic flux density in the core before the winding is short-circuited, B0 represents the maximum magnetic flux density in the core after the winding is short-circuited, P wN Indicates the loss per unit volume of the winding at rated current, I b Indicates the winding current with short circuit between turns, I N Indicates the rated current of the winding with short circuit between turns, I sh Represents the equivalent fault resistance current of turn-to-turn short circuit, R sh It represents the equivalent fault resistance value of inter-turn short circuit, n represents the proportion of inter-turn short circuit, and V represents the volume of fault winding;

[0152] The loss value and the preset inlet flow velocity value in the flow field (such as 0.02 m / s) are input into the thermal field and flow field coupling model to obtain a multi-field coupling simulation model for the target transformer.

[0153] Optionally, when constructing a multi-field coupling simulation model, the coupling of the thermal field and the flow field can be achieved through the solid heat transfer module and the laminar flow module in the numerical simulation software COMSOL.

[0154] Furthermore, the multi-field coupling simulation model is corrected through actual data to achieve digital twins, which can truly and consistently reflect the changes in the micro-temperature field. Specifically, the real signals collected by each sensor arranged in the micro-temperature field of the target transformer are obtained in real time; the parameters of the multi-field coupling simulation model are corrected using the real signals until the preset accuracy threshold is met, so as to obtain the micro-temperature field digital twin model of the target transformer.

[0155] Optionally, when the multi-field coupling simulation model is corrected by actual data, the coupling relationship between the multi-physical field inside the transformer and the transformer structure can be used to effectively analyze and calculate the unknown parameters in the micro-temperature field digital twin model, thereby realizing dynamic tracking of the real equipment, which can be achieved based on the following formula,

[0156]

[0157] Among them, m(t) represents the current system state, u(t) represents the input data, n(t) represents the output data, F represents the state function, and H represents the output function. Furthermore, the above model dual can be simplified into an equivalent system

[0158]

[0159] However, the number of poles of the dual simplified system is much smaller than that of the original system. Therefore, after evaluating the dual simplified system, it is necessary to implement the meshing of the two models through the microservice unit to achieve the dynamic update of the micro-thermal field.

[0160] In a specific application scenario, when there is an internal short circuit between turns, according to the circuit principle, the equivalent circuit can be regarded as two windings connected in series, such as Figure 4 shown.

[0161] 202. Based on the micro-temperature field digital twin model, the micro-temperature field simulation temperature distribution data of the target transformer is obtained.

[0162] It is understandable that simulation data of any time period and any position can be selected from the micro-temperature field simulation temperature distribution data.

[0163] 203. When a turn-to-turn short circuit fault is detected in the target transformer, a target sensor group is selected according to comprehensive score coefficients of various sensors arranged in the micro-temperature field of the target transformer.

[0164] Among them, the comprehensive score coefficient is obtained by fusing the basic weight vector of the correlation between the sensor temperature and the hotspot temperature of each sensor, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between sensor signals to obtain a comprehensive score vector, and then allocating the weights based on the game weighted algorithm.

[0165] Due to the huge number of sensors, if fault location is performed based on the collected data of all sensors, it is very easy to cause the problem of slowing down the positioning efficiency due to the need to process massive data. Based on this, in the embodiment of the present application, a valuable target sensor group is pre-screened therefrom. It should be noted that since the location of the inter-turn short circuit is random, each sensor must reflect the fault characteristics, but due to the different fault locations and degrees, the importance of each sensor to reflecting the fault characteristics is not the same. According to the contribution of the micro-temperature field analysis, valuable sensors can be selected to capture the hotspot temperature and locate the fault position. In the embodiment of the present application, the comprehensive score coefficient is used as an evaluation index for sensor screening. Specifically, when an inter-turn short circuit fault is detected in the target transformer, the basic weight vector of the correlation between the sensor temperature and the hotspot temperature is calculated based on the Pearson correlation coefficient formula. The Pearson correlation coefficient formula is expressed as the following formula

[0166]

[0167] Where r represents the correlation vector between sensor temperature and hotspot temperature, M represents the number of sensors, and x j represents the hot spot temperature in the jth sensor, represents the average temperature of the hot spot in the sensor, y j represents the temperature of the jth sensor, Indicates the average temperature of the sensor;

[0168] Based on the signal variance of each sensor, the basic weight vector of sensor data reliability is determined, which can be expressed as V r =[V r1 ,V r2 ,...,V rj ], where V r Represents the basic weight vector of sensor data reliability, V rj represents the signal variance of the jth sensor;

[0169] Based on the symmetry index calculation formula, the symmetry index between each pair of sensor signals is calculated to generate the symmetry basic weight vector between the sensor signals. The symmetry index calculation formula is expressed as follows:

[0170] DS=R*ρ

[0171] Where DS represents the symmetry index between each pair of sensor signals, R represents the distance coefficient, R = -ln(r1-r2), r1 and r2 represent the root mean square values ​​of the current monitoring points of the two sensors, and ρ represents the Spearman correlation coefficient;

[0172] The basic weight vector of symmetry between sensor signals can be expressed as follows:

[0173]

[0174] Among them, d q represents the basic weight vector of symmetry between sensor signals, d ij represents the symmetry index between the i-th sensor and the j-th sensor;

[0175] The basic weight vector of the correlation between sensor temperature and hotspot temperature, the basic weight vector of sensor data reliability, and the basic weight vector of symmetry between sensor signals are integrated to obtain a comprehensive score vector

[0176]

[0177] Among them, W represents the comprehensive score vector, λ α represents the weight coefficient, α=1 represents the basic weight vector of the correlation between sensor temperature and hot spot temperature, α=2 represents the basic weight vector of sensor data reliability, α=3 represents the basic weight vector of symmetry between sensor signals, and T represents the combination coefficient;

[0178] The weight coefficient is optimized based on the game weighted algorithm so that the deviations between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hot spot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are all minimum values, and the optimal weight coefficient is obtained. Preferably, the weight coefficient can be optimized by solving the minimum objective function, and the minimum objective function can be expressed as the following formula

[0179]

[0180] Among them, y1 represents the minimum objective function between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hot spot temperature, y2 represents the minimum objective function between the comprehensive score vector and the basic weight vector of the sensor data reliability, and y3 represents the minimum objective function between the comprehensive score vector and the basic weight vector of the symmetry between the sensor signals;

[0181] The optimal weight coefficient is normalized to obtain the comprehensive score coefficient of each sensor;

[0182] Based on preset rules, a target sensor group is screened out according to the comprehensive score coefficients of each sensor, so as to locate the turn-to-turn short circuit fault according to the micro-temperature field simulation data corresponding to each sensor in the target sensor group. The preset rules can be to sort the comprehensive score coefficients of each sensor in sequence, select a preset number of sensors from the first one to form a target sensor group, or set a comprehensive score coefficient threshold to form a target sensor group with sensors exceeding the threshold. No specific limitation is made in the embodiments of the present application.

[0183] 204. Acquire micro-temperature field simulation data corresponding to each sensor in the target sensor group as micro-temperature field simulation data of the target transformer to locate the turn-to-turn short-circuit fault position.

[0184] 205. Based on convolutional neural network, the characteristic parameters of turn-to-turn short-circuit fault are extracted from micro-temperature field simulation data.

[0185] Among them, the micro-temperature field simulation data includes time series parameter data and image parameter data.

[0186] For the time series parameter data, firstly, the time series parameter data is arranged and processed based on the preset matrix rules to obtain a time series parameter matrix that meets the preset matrix rules; further, the time series parameter matrix is ​​converted in form so that the time series parameter matrix has the same representation form as the data image; further, the first turn-to-turn short circuit fault characteristic parameter corresponding to the time series parameter is extracted from the time series parameter after the form conversion based on the first convolutional neural network. Among them, the time series parameter data may include oil temperature, oil speed, oil flow, etc.; the preset matrix rules, for example, a matrix of size K×N, K represents the category of the time series, and N represents the length of the time series; the first convolutional neural network can adopt a shallow convolutional neural network, and the structure includes, for example, a convolution layer containing 32 convolution kernels of size 5×5, a nonlinear function Relu, a maximum pooling layer of size 2×2, and a fully connected layer of size 125.

[0187] For the image parameter data, the second turn-to-turn short-circuit fault characteristic parameter corresponding to the image parameter is extracted from the image parameter data based on the second convolutional neural network. The second convolutional neural network can adopt a mature convolutional neural network, such as the Faster R-CNN neural network, which can include two convolutional layers, each of which contains 64 3×3 convolution kernels.

[0188] Finally, based on a preset weight ratio, the first turn-to-turn short-circuit fault characteristic parameter and the second turn-to-turn short-circuit fault characteristic parameter are fused to obtain the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

[0189] Exemplarily, assuming that the weight ratio of the first turn-to-turn short-circuit fault characteristic parameter is W1, the second turn-to-turn short-circuit fault characteristic parameter is W2, the first turn-to-turn short-circuit fault characteristic parameter is a 1×125 characteristic matrix, and the second turn-to-turn short-circuit fault characteristic parameter is a 1×4096 characteristic matrix. By connecting the two according to their respective weight ratios, a 1×4221 fused feature matrix can be obtained. At this time, the fused feature matrix can be input into a fully connected layer with a size of 1×4096, and a fused feature matrix with a size of 1×4096 can be obtained, that is, the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

[0190] Preferably, W1 and W2 can be regarded as trainable parameters and trained using a cross entropy loss or a mean square error minimization objective, thereby reducing the influence of human factors and improving the objectivity of the characteristic parameters of the turn-to-turn short-circuit fault.

[0191] 206. According to the characteristic parameters of the inter-turn short-circuit fault, the location of the inter-turn short-circuit fault of the target transformer is determined based on the Bayesian reasoning method.

[0192] Specifically, based on the total probability calculation formula, the total probability value of observing the turn-to-turn short-circuit fault characteristics is calculated according to the turn-to-turn short-circuit fault characteristic parameters. The total probability calculation formula is expressed as follows:

[0193]

[0194] Among them, Y n It represents the characteristic parameter of turn-to-turn short-circuit fault, P(Y n ) represents the total probability of observing the characteristics of turn-to-turn short-circuit faults, Indicates the frequency of occurrence of the characteristic corresponding to the inter-turn short-circuit fault, X m Indicates that a turn-to-turn open circuit fault occurs in area m, where m represents areas A, B, and C. Indicates the total frequency of turn-to-turn short-circuit faults in the three areas A, B, and C;

[0195] Based on the following formula, the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition of turn-to-turn short-circuit fault occurring in each area is calculated:

[0196]

[0197] Among them, P(Y n |X m ) represents the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in region m, It indicates the probability value of turn-to-turn short circuit fault occurring in the three regions A, B and C respectively;

[0198] Based on the Bayesian fault location formula, according to the total probability value of observing the inter-turn short-circuit fault characteristics and the probability value of observing the inter-turn short-circuit fault characteristics under the condition that the inter-turn short-circuit fault occurs in each area, the probability value of the inter-turn short-circuit fault occurring in each area under the condition that the inter-turn short-circuit fault characteristics are observed is calculated respectively. The Bayesian fault location formula is expressed as the following formula

[0199]

[0200] Among them, P(X m |Y n ) represents the probability value of turn-to-turn short-circuit fault occurring in region m under the condition that the turn-to-turn short-circuit fault characteristics are observed, P(Xm ) represents the preset prior probability value of the turn-to-turn short circuit fault occurring in region m;

[0201] The area corresponding to the maximum probability value is screened out from the probability values ​​of the turn-to-turn short-circuit fault, and the area is determined as the location where the turn-to-turn short-circuit fault occurs in the target transformer.

[0202] In order to verify the positioning accuracy of the method for locating the transformer turn-to-turn short-circuit fault position provided in the embodiment of the present application, a single-phase transformer is taken as an example for simulation and verification, and the basic parameters of the transformer are shown in Table 1. In order to facilitate the simulation of the turn-to-turn short-circuit model, a turn-to-turn short-circuit is set in the primary winding, and the winding is a pancake structure to simulate turn-to-turn short-circuits at different positions.

[0203] Table 1 Transformer parameters

[0204] The winding and core parameters of the transformer are shown in Table 2.

[0205]

[0206] Density, specific heat capacity, dynamic viscosity, etc. will change continuously with temperature.

[0207] Table 2 Physical properties of transformer materials

[0208]

[0209] Since the three-dimensional model is closer to the actual transformer, in the embodiment of the present application, the three-dimensional transformer model is used for simulation. Considering the fault location and severity when the transformer has a turn-to-turn short circuit, in the modeling process, in order to eliminate the large error caused by poor mesh quality and ensure the reliability of the results, the primary winding, the secondary winding and the oil channel position are meshed finely until certain control parameters do not change. The meshing degree in the embodiment of the present application is as follows: Figure 5As shown. In order to timely obtain the heat distribution of each winding in the oil channel when a turn-to-turn short circuit occurs, each winding is monitored by a sensor, and the transformer is evenly divided into three zones. Sensors are evenly installed in the three zones to provide a priori probability for the Bayesian reasoning of the probability of turn-to-turn short circuit failure in each zone. In order to verify the accuracy and effectiveness of the simulation model, in the embodiment of the present application, the transformer simulates the operating conditions of the transformer through a temperature rise experiment. For the entire transformer industry, it is very costly to install a large number of optical fiber probes inside the transformer to detect the internal temperature of the transformer, and it is impractical to develop a scale consistent with the prototype in the research scenario. Therefore, a reduced-size experimental transformer model was designed and manufactured to specifically analyze the heat transfer of the transformer, but this does not affect the experimental results, so as to obtain the conclusion. In terms of temperature measurement, considering the complexity of installing optical fiber sensors along the transformer windings, a group of T22 optical fiber temperature probes are placed in the oil channel between the primary winding and the secondary winding for verification with the simulation model. The measurement range is -80℃-250℃, and the response time is 250ms. The experimental measurement results of the winding temperature under normal conditions are compared with the simulation results. Figure 6 As shown in Figure 2, the experimental measurement results of the winding temperature under fault conditions are compared with the simulation results. Figure 7 As shown in the figure, the winding simulation calculated temperature and the experimental measured temperature have roughly the same trend with the winding height, and the overall temperature error is less than 3K. The error in the experiment mainly comes from the simplification of the simulation model, which makes the simulation results inconsistent with the experimental results.

[0210] The present application provides a method for locating the position of a transformer inter-turn short-circuit fault. First, when a target transformer is detected to have an inter-turn short-circuit fault, micro-temperature field simulation data of the target transformer is obtained. The micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer. The micro-temperature field digital twin model is based on a multi-field coupling simulation model of the target transformer and is corrected using real signals collected by sensors arranged in the micro-temperature field of the target transformer. Secondly, based on a convolutional neural network, characteristic parameters of the inter-turn short-circuit fault are extracted from the micro-temperature field simulation data. Finally, based on the characteristic parameters of the inter-turn short-circuit fault, the position of the inter-turn short-circuit fault of the target transformer is determined based on a Bayesian reasoning method. Compared with the prior art, the embodiment of the present application, based on the multi-field coupling simulation model of the target transformer, uses the real signal collected by the sensor arranged in the micro-temperature field of the target transformer to correct the micro-temperature field digital twin model, so that the simulation data obtained based on the micro-temperature field digital twin model is highly similar to the actual operation of the target transformer. On this basis, the simulation data is used to locate the fault in advance before the fault occurs, thereby avoiding the problem of internal fault of the transformer caused by untimely response; further, based on the convolutional neural network, the turn-to-turn short-circuit fault characteristic parameters are extracted from the micro-temperature field simulation data, and by utilizing the high sensitivity characteristics of the convolutional neural network, the problem of not being able to detect the fault characteristics in time due to the slow temperature change is overcome, the recognition sensitivity of the fault characteristics is improved, and then the action sensitivity is improved; finally, by using the Bayesian reasoning method to determine the exact location of the turn-to-turn short-circuit fault according to the turn-to-turn short-circuit fault characteristic parameters, the defect of the existing method that the turn-to-turn short-circuit fault position cannot be accurately located is overcome.

[0211] Furthermore, as a response to the above Figure 1 The implementation of the method shown in the figure, the embodiment of the present application provides a device for locating the position of a transformer inter-turn short circuit fault, such as Figure 8 As shown, the device comprises:

[0212] Micro-temperature field simulation data acquisition module 31, turn-to-turn short-circuit fault characteristic parameter extraction module 32, turn-to-turn short-circuit fault position positioning module 33;

[0213] A micro-temperature field simulation data acquisition module 31 is used to acquire micro-temperature field simulation data of the target transformer when a turn-to-turn short circuit fault is detected in the target transformer, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer;

[0214] A turn-to-turn short-circuit fault characteristic parameter extraction module 32, used to extract turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data based on a convolutional neural network;

[0215] The turn-to-turn short-circuit fault location positioning module 33 is used to determine the location of the turn-to-turn short-circuit fault of the target transformer based on the Bayesian reasoning method according to the turn-to-turn short-circuit fault characteristic parameters.

[0216] In a specific application scenario, before the micro-temperature field simulation data acquisition module, the device further includes a micro-temperature field digital twin model construction module, which is used to:

[0217] Based on the electromagnetic field differential equation and the natural convection heat transfer mathematical model, a thermal field and flow field coupling model is constructed. The electromagnetic field differential equation is expressed as the following formula:

[0218]

[0219] in, represents the vector differential operator, H represents the magnetic field intensity, J represents the current density, D represents the electric displacement vector, t represents the time, B represents the magnetic induction intensity, E represents the electric field intensity, p represents the charge density,

[0220] The natural convection heat transfer mathematical model is expressed as the following formula

[0221] q=h(T b -T o )

[0222] Where q represents the heat flux density, h represents the convective heat transfer coefficient, T b Indicates the transformer tank temperature, T o Indicates the air temperature;

[0223] The loss value is calculated based on the core and winding loss formula, which is expressed as the following formula

[0224]

[0225] Among them, P 总 Represents the loss value, P fe0 It represents the core loss per unit volume before the winding is short-circuited, B represents the maximum magnetic flux density in the core before the winding is short-circuited, B0 represents the maximum magnetic flux density in the core after the winding is short-circuited, P wN Indicates the loss per unit volume of the winding at rated current, I b Indicates the winding current with short circuit between turns, I N Indicates the rated current of the winding with short circuit between turns, I sh Represents the equivalent fault resistance current of turn-to-turn short circuit, R shIt represents the equivalent fault resistance value of inter-turn short circuit, n represents the proportion of inter-turn short circuit, and V represents the volume of fault winding;

[0226] The loss value and the preset inlet flow velocity value in the flow field are input into the thermal field and flow field coupling model to obtain a multi-field coupling simulation model for the target transformer.

[0227] In specific application scenarios, the micro-temperature field digital twin model construction module is also used to:

[0228] Acquire real signals collected by various sensors arranged in the micro-temperature field of the target transformer in real time;

[0229] The parameters of the multi-field coupling simulation model are corrected using the real signal until a preset accuracy threshold is met, so as to obtain a micro-temperature field digital twin model of the target transformer.

[0230] In a specific application scenario, the turn-to-turn short-circuit fault location positioning module is used to:

[0231] Based on the total probability calculation formula, the total probability value of observing the turn-to-turn short circuit fault characteristic is calculated according to the turn-to-turn short circuit fault characteristic parameter. The total probability calculation formula is expressed as the following formula

[0232]

[0233] Among them, Y n It represents the characteristic parameter of turn-to-turn short-circuit fault, P(Y n ) represents the total probability of observing the characteristics of turn-to-turn short-circuit faults, Indicates the frequency of occurrence of the characteristic corresponding to the inter-turn short-circuit fault, X m Indicates that a turn-to-turn open circuit fault occurs in area m, where m represents areas A, B, and C. Indicates the total frequency of turn-to-turn short-circuit faults in the three areas A, B, and C;

[0234] Based on the following formula, the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition of turn-to-turn short-circuit fault occurring in each area is calculated:

[0235]

[0236] Among them, P(Y n |X m ) represents the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in region m, It indicates the probability value of turn-to-turn short circuit fault occurring in the three regions A, B and C respectively;

[0237] Based on the Bayesian fault location formula, according to the total probability value of the observed turn-to-turn short-circuit fault characteristics and the probability value of the observed turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in each area, the probability value of the turn-to-turn short-circuit fault occurring in each area under the condition that the turn-to-turn short-circuit fault characteristics are observed is calculated respectively. The Bayesian fault location formula is expressed as the following formula

[0238]

[0239] Among them, P(X m |Y n ) represents the probability value of turn-to-turn short-circuit fault occurring in region m under the condition that the turn-to-turn short-circuit fault characteristics are observed, P(X m ) represents the preset prior probability value of the turn-to-turn short circuit fault occurring in region m;

[0240] An area corresponding to a maximum probability value is selected from the probability values ​​of the turn-to-turn short-circuit fault, and the area is determined as a location where the turn-to-turn short-circuit fault of the target transformer occurs.

[0241] In a specific application scenario, the micro-temperature field simulation data includes time series parameter data and image parameter data, and the turn-to-turn short circuit fault characteristic parameter extraction module is used to:

[0242] Based on a preset matrix rule, the timing parameter data is arranged and processed to obtain a timing parameter matrix that conforms to the preset matrix rule;

[0243] Converting the time series parameter matrix into a form so that the time series parameter matrix and the data image have the same representation form;

[0244] Extracting a first turn-to-turn short-circuit fault characteristic parameter corresponding to the timing parameter from the timing parameter after the form conversion based on a first convolutional neural network;

[0245] Extracting a second turn-to-turn short-circuit fault characteristic parameter corresponding to the image parameter from the image parameter data based on a second convolutional neural network;

[0246] Based on a preset weight ratio, the first turn-to-turn short-circuit fault characteristic parameter and the second turn-to-turn short-circuit fault characteristic parameter are fused to obtain the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

[0247] In a specific application scenario, the micro-temperature field simulation data acquisition module includes:

[0248] A target sensor group screening unit is used to screen out a target sensor group according to the comprehensive score coefficients of each sensor arranged in the micro-temperature field of the target transformer when a turn-to-turn short circuit fault is detected in the target transformer, wherein the comprehensive score coefficient is obtained by fusing the basic weight vectors of the correlation between the sensor temperature and the hotspot temperature of each sensor, the basic weight vectors of the reliability of the sensor data, and the basic weight vectors of the symmetry between the sensor signals to obtain a comprehensive score vector, and performing weight distribution based on a game weighting algorithm;

[0249] The micro-temperature field simulation data acquisition unit is used to acquire the micro-temperature field simulation data corresponding to each sensor in the target sensor group as the micro-temperature field simulation data of the target transformer to locate the turn-to-turn short-circuit fault position.

[0250] In a specific application scenario, the target sensor group screening unit is used to:

[0251] When a turn-to-turn short circuit fault is detected in the target transformer, the basic weight vector of the correlation between the sensor temperature and the hotspot temperature is calculated based on the Pearson correlation coefficient formula, and the Pearson correlation coefficient formula is expressed as follows:

[0252]

[0253] Where r represents the correlation vector between sensor temperature and hotspot temperature, M represents the number of sensors, and x j represents the hot spot temperature in the jth sensor, represents the average temperature of the hot spot in the sensor, y j represents the temperature of the jth sensor, Indicates the average temperature of the sensor;

[0254] Determining a sensor data reliability basic weight vector based on the signal variance of each of the sensors;

[0255] Based on the symmetry index calculation formula, the symmetry index between each pair of sensor signals is calculated to generate the symmetry basic weight vector between the sensor signals. The symmetry index calculation formula is expressed as the following formula:

[0256] DS=R*ρ

[0257] Where DS represents the symmetry index between each pair of sensor signals, R represents the distance coefficient, R = -ln(|r1-r2|), r1 and r2 represent the root mean square values ​​of the current monitoring points of the two sensors, and ρ represents the Spearman correlation coefficient;

[0258] The basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are fused to obtain a comprehensive score vector

[0259]

[0260] Among them, W represents the comprehensive score vector, λ α represents the weight coefficient, α=1 represents the basic weight vector of the correlation between sensor temperature and hot spot temperature, α=2 represents the basic weight vector of sensor data reliability, α=3 represents the basic weight vector of symmetry between sensor signals, and T represents the combination coefficient;

[0261] The weight coefficient is optimized based on a game weighting algorithm so that the deviations between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are all minimum values, thereby obtaining the optimal weight coefficient;

[0262] Normalizing the optimal weight coefficient to obtain a comprehensive score coefficient of each sensor;

[0263] Based on preset rules, a target sensor group is screened out according to the comprehensive score coefficients of each sensor, so as to locate the turn-to-turn short circuit fault position according to the micro-temperature field simulation data corresponding to each sensor in the target sensor group.

[0264] The present application provides a device for locating the position of a transformer turn-to-turn short-circuit fault. First, when a turn-to-turn short-circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained, and the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer. The micro-temperature field digital twin model is based on a multi-field coupling simulation model of the target transformer and is corrected using real signals collected by sensors arranged in the micro-temperature field of the target transformer; secondly, based on a convolutional neural network, turn-to-turn short-circuit fault characteristic parameters are extracted from the micro-temperature field simulation data; finally, based on the turn-to-turn short-circuit fault characteristic parameters, the position of the turn-to-turn short-circuit fault of the target transformer is determined based on a Bayesian reasoning method. Compared with the prior art, the embodiment of the present application, based on the multi-field coupling simulation model of the target transformer, uses the real signal collected by the sensor arranged in the micro-temperature field of the target transformer to correct the micro-temperature field digital twin model, so that the simulation data obtained based on the micro-temperature field digital twin model is highly similar to the actual operation of the target transformer. On this basis, the simulation data is used to locate the fault in advance before the fault occurs, thereby avoiding the problem of internal fault of the transformer caused by untimely response; further, based on the convolutional neural network, the turn-to-turn short-circuit fault characteristic parameters are extracted from the micro-temperature field simulation data, and by utilizing the high sensitivity characteristics of the convolutional neural network, the problem of not being able to detect the fault characteristics in time due to the slow temperature change is overcome, the recognition sensitivity of the fault characteristics is improved, and then the action sensitivity is improved; finally, by using the Bayesian reasoning method to determine the exact location of the turn-to-turn short-circuit fault according to the turn-to-turn short-circuit fault characteristic parameters, the defect of the existing method that the turn-to-turn short-circuit fault position cannot be accurately located is overcome.

[0265] According to an embodiment of the present application, a storage medium is provided, wherein the storage medium stores at least one executable instruction, and the computer executable instruction can execute the method for locating the position of a transformer inter-turn short circuit fault in any of the above method embodiments.

[0266] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0267] Fig. 9 A schematic diagram of the structure of a terminal provided according to an embodiment of the present application is shown, and the specific embodiment of the present application does not limit the specific implementation of the terminal.

[0268] like Fig. 9As shown, the terminal may include: a processor (processor) 402 , a communication interface (Communications Interface) 404 , a memory (memory) 406 , and a communication bus 408 .

[0269] The processor 402 , the communication interface 404 , and the memory 406 communicate with each other via a communication bus 408 .

[0270] The communication interface 404 is used to communicate with other devices such as clients or other servers.

[0271] The processor 402 is used to execute the program 410, and specifically can execute the relevant steps in the above-mentioned transformer inter-turn short circuit fault location positioning method embodiment.

[0272] Specifically, the program 410 may include program codes, which include computer operation instructions.

[0273] The processor 402 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in the computer device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0274] The memory 406 is used to store the program 410. The memory 406 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0275] The program 410 may be specifically configured to enable the processor 402 to perform the following operations:

[0276] When a turn-to-turn short circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer;

[0277] Based on a convolutional neural network, extracting turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data;

[0278] According to the turn-to-turn short-circuit fault characteristic parameters, a location where the turn-to-turn short-circuit fault occurs in the target transformer is determined based on a Bayesian reasoning method.

[0279] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the physical device hardware and software resources of the method for locating the position of the transformer inter-turn short-circuit fault, and supports the operation of the information processing program and other software and / or programs. The network communication module is used to realize the communication between the components inside the storage medium, and the communication with other hardware and software in the information processing physical device.

[0280] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0281] The method and system of the present application may be implemented in many ways. For example, the method and system of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present application may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers a recording medium storing a program for executing the method according to the present application.

[0282] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0283] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for locating a transformer turn-to-turn short-circuit fault, characterized in that: include: When a turn-to-turn short circuit fault is detected in a target transformer, micro-temperature field simulation data of the target transformer is obtained, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer; Based on a convolutional neural network, extracting turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data; According to the turn-to-turn short-circuit fault characteristic parameters, a location where the turn-to-turn short-circuit fault occurs in the target transformer is determined based on a Bayesian reasoning method.

2. The method according to claim 1, characterized in that: When a turn-to-turn short circuit fault is detected in the target transformer, before obtaining micro-temperature field simulation data of the target transformer, the method further includes: Based on the electromagnetic field differential equation and the natural convection heat transfer mathematical model, a thermal field and flow field coupling model is constructed. The electromagnetic field differential equation is expressed as the following formula: in, represents the vector differential operator, H represents the magnetic field intensity, J represents the current density, D represents the electric displacement vector, t represents the time, B represents the magnetic induction intensity, E represents the electric field intensity, p represents the charge density, The natural convection heat transfer mathematical model is expressed as the following formula q=h(T b -T o ) Where q represents the heat flux density, h represents the convective heat transfer coefficient, T b Indicates the transformer tank temperature, T o Indicates the air temperature; The loss value is calculated based on the core and winding loss formula, which is expressed as the following formula Among them, P 总 Represents the loss value, P fe0 It represents the core loss per unit volume before the winding is short-circuited, B represents the maximum magnetic flux density in the core before the winding is short-circuited, B0 represents the maximum magnetic flux density in the core after the winding is short-circuited, P wN Indicates the loss per unit volume of the winding at rated current, I b Indicates the winding current with short circuit between turns, I N Indicates the rated current of the winding with short circuit between turns, I sh Represents the equivalent fault resistance current of turn-to-turn short circuit, R sh It represents the equivalent fault resistance value of inter-turn short circuit, n represents the proportion of inter-turn short circuit, and V represents the volume of fault winding; The loss value and the preset inlet flow velocity value in the flow field are input into the thermal field and flow field coupling model to obtain a multi-field coupling simulation model for the target transformer.

3. The method according to claim 2, characterized in that After obtaining the multi-field coupling simulation model for the target transformer, the method further includes: Acquire real signals collected by various sensors arranged in the micro-temperature field of the target transformer in real time; The parameters of the multi-field coupling simulation model are corrected using the real signal until a preset accuracy threshold is met, so as to obtain a micro-temperature field digital twin model of the target transformer.

4. The method according to claim 1, characterized in that The determining, based on the characteristic parameters of the turn-to-turn short-circuit fault and based on the Bayesian reasoning method, the location where the turn-to-turn short-circuit fault occurs in the target transformer comprises: Based on the total probability calculation formula, the total probability value of observing the turn-to-turn short circuit fault characteristic is calculated according to the turn-to-turn short circuit fault characteristic parameter. The total probability calculation formula is expressed as the following formula Among them, Y n It represents the characteristic parameter of turn-to-turn short-circuit fault, P(Y n ) represents the total probability of observing the characteristics of turn-to-turn short-circuit faults, Indicates the frequency of occurrence of the characteristic corresponding to the inter-turn short-circuit fault, X m Indicates that a turn-to-turn open circuit fault occurs in area m, where m represents areas A, B, and C. Indicates the total frequency of turn-to-turn short-circuit faults in the three areas A, B, and C; Based on the following formula, the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition of turn-to-turn short-circuit fault occurring in each area is calculated: Among them, P(Y n |X m ) represents the probability value of observing the turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in region m, It indicates the probability value of turn-to-turn short circuit fault occurring in the three regions A, B and C respectively; Based on the Bayesian fault location formula, according to the total probability value of the observed turn-to-turn short-circuit fault characteristics and the probability value of the observed turn-to-turn short-circuit fault characteristics under the condition that the turn-to-turn short-circuit fault occurs in each area, the probability value of the turn-to-turn short-circuit fault occurring in each area under the condition that the turn-to-turn short-circuit fault characteristics are observed is calculated respectively. The Bayesian fault location formula is expressed as the following formula Among them, P(X m |Y n ) represents the probability value of turn-to-turn short-circuit fault occurring in region m under the condition that the turn-to-turn short-circuit fault characteristics are observed, P(X m ) represents the preset prior probability value of the turn-to-turn short circuit fault occurring in region m; An area corresponding to a maximum probability value is selected from the probability values ​​of the turn-to-turn short-circuit fault, and the area is determined as a location where the turn-to-turn short-circuit fault of the target transformer occurs.

5. The method according to claim 1, characterized in that The micro-temperature field simulation data includes time series parameter data and image parameter data. The method based on the convolutional neural network extracts the turn-to-turn short circuit fault characteristic parameters from the micro-temperature field simulation data, including: Based on a preset matrix rule, the timing parameter data is arranged and processed to obtain a timing parameter matrix that conforms to the preset matrix rule; Converting the time series parameter matrix into a form so that the time series parameter matrix and the data image have the same representation form; Extracting a first turn-to-turn short-circuit fault characteristic parameter corresponding to the timing parameter from the timing parameter after the form conversion based on a first convolutional neural network; Extracting a second turn-to-turn short-circuit fault characteristic parameter corresponding to the image parameter from the image parameter data based on a second convolutional neural network; Based on a preset weight ratio, the first turn-to-turn short-circuit fault characteristic parameter and the second turn-to-turn short-circuit fault characteristic parameter are fused to obtain the turn-to-turn short-circuit fault characteristic parameter of the micro-temperature field simulation data.

6. The method according to claim 1, characterized in that When a turn-to-turn short circuit fault is detected in the target transformer, micro-temperature field simulation data of the target transformer is obtained, including: When a turn-to-turn short circuit fault is detected in the target transformer, a target sensor group is selected according to the comprehensive score coefficients of the sensors arranged in the micro-temperature field of the target transformer, wherein the comprehensive score coefficient is obtained by fusing the basic weight vectors of the correlation between the sensor temperature and the hotspot temperature of each sensor, the basic weight vectors of the reliability of the sensor data, and the basic weight vectors of the symmetry between the sensor signals to obtain a comprehensive score vector, and performing weight distribution based on a game weighting algorithm; The micro-temperature field simulation data corresponding to each sensor in the target sensor group is obtained as the micro-temperature field simulation data of the target transformer to locate the turn-to-turn short-circuit fault position.

7. The method according to claim 6, characterized in that When a turn-to-turn short circuit fault is detected in the target transformer, a target sensor group is selected according to the comprehensive score coefficients of the sensors arranged in the micro-temperature field of the target transformer, including: When a turn-to-turn short circuit fault is detected in the target transformer, the basic weight vector of the correlation between the sensor temperature and the hotspot temperature is calculated based on the Pearson correlation coefficient formula, and the Pearson correlation coefficient formula is expressed as follows: Where r represents the correlation vector between sensor temperature and hotspot temperature, M represents the number of sensors, and x j represents the hot spot temperature in the jth sensor, represents the average temperature of the hot spot in the sensor, y j represents the temperature of the jth sensor, Indicates the average temperature of the sensor; Determining a sensor data reliability basic weight vector based on the signal variance of each of the sensors; Based on the symmetry index calculation formula, the symmetry index between each pair of sensor signals is calculated to generate the symmetry basic weight vector between the sensor signals. The symmetry index calculation formula is expressed as the following formula: DS=R*ρ Where DS represents the symmetry index between each pair of sensor signals, R represents the distance coefficient, R = -ln(|r1-r2|), r1 and r2 represent the root mean square values ​​of the current monitoring points of the two sensors, and ρ represents the Spearman correlation coefficient; The basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are integrated to obtain a comprehensive score vector Among them, W represents the comprehensive score vector, λ α represents the weight coefficient, α=1 represents the basic weight vector of the correlation between sensor temperature and hot spot temperature, α=2 represents the basic weight vector of sensor data reliability, α=3 represents the basic weight vector of symmetry between sensor signals, and T represents the combination coefficient; The weight coefficient is optimized based on a game weighting algorithm so that the deviations between the comprehensive score vector and the basic weight vector of the correlation between the sensor temperature and the hotspot temperature, the basic weight vector of the sensor data reliability, and the basic weight vector of the symmetry between the sensor signals are all minimum values, thereby obtaining the optimal weight coefficient; Normalizing the optimal weight coefficient to obtain a comprehensive score coefficient of each sensor; Based on preset rules, a target sensor group is screened out according to the comprehensive score coefficients of each sensor, so as to locate the turn-to-turn short circuit fault position according to the micro-temperature field simulation data corresponding to each sensor in the target sensor group.

8. A device for locating the position of a transformer turn-to-turn short-circuit fault, characterized in that: include: A micro-temperature field simulation data acquisition module is used to acquire micro-temperature field simulation data of the target transformer when a turn-to-turn short circuit fault is detected in the target transformer, wherein the micro-temperature field simulation data is obtained based on a micro-temperature field digital twin model of the target transformer, and the micro-temperature field digital twin model is obtained by correcting a real signal collected by a sensor arranged in the micro-temperature field of the target transformer on the basis of a multi-field coupling simulation model for the target transformer; A turn-to-turn short-circuit fault characteristic parameter extraction module, used to extract turn-to-turn short-circuit fault characteristic parameters from the micro-temperature field simulation data based on a convolutional neural network; The turn-to-turn short-circuit fault location positioning module is used to determine the location of the turn-to-turn short-circuit fault of the target transformer based on the Bayesian reasoning method according to the turn-to-turn short-circuit fault characteristic parameters.

9. A storage medium, wherein at least one executable instruction is stored in the storage medium, characterized in that: The executable instructions enable the processor to execute operations corresponding to the method for locating the position of a transformer turn-to-turn short-circuit fault according to any one of claims 1 to 7.

10. A terminal, comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, wherein the executable instruction enables the processor to perform operations corresponding to the method for locating the position of a transformer inter-turn short-circuit fault according to any one of claims 1-7.

Images