Transformer abnormal loss positioning method and device, storage medium and program product

Abstract

The invention discloses a transformer abnormal loss positioning method and device, a storage medium and a program product, and relates to the technical field of transformers, and the method comprises the steps: obtaining transformer data of a to-be-analyzed transformer; the transformer data comprises actually measured data and corresponding standard data; performing root cause analysis on the to-be-analyzed transformer based on the transformer data and the load loss mechanism model of the transformer to obtain a root cause analysis result; the load loss mechanism model comprises a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values of additional loss influence parameters and additional loss measured values corresponding to each group of measured values; and based on the root cause analysis result, positioning the load loss abnormity of the to-be-analyzed transformer. According to the invention, the root cause causing the abnormal load loss of the transformer can be accurately positioned.

Description

Technical Field

[0001] The present application relates to the field of transformer technology, and in particular to a method, device, storage medium and program product for locating abnormal loss of a transformer. Background Art

[0002] In the related technology, when faced with abnormal transformer load loss, the cause of the abnormality is usually guessed based on the operator's experience, and it is difficult to locate the specific cause in a targeted manner. For production companies, it is difficult to locate the root cause of abnormal transformer load loss, which makes it difficult to control the quality consistency of transformer products during the production process. Therefore, it is necessary to find a method that can accurately locate the root cause of abnormal transformer load loss. Summary of the invention

[0003] The main purpose of this application is to provide a method, device, storage medium and program product for locating abnormal transformer loss, aiming to solve the technical problem in related technologies that it is difficult to accurately locate the root cause of abnormal transformer load loss.

[0004] To achieve the above objectives, the present application proposes a method for locating abnormal transformer losses, comprising: Obtain transformer data of the transformer to be analyzed; the transformer data includes measured data and corresponding standard data; Based on the transformer data and the load loss mechanism model of the transformer, a root cause analysis is performed on the transformer to be analyzed to obtain the root cause analysis results; the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values; Based on the root cause analysis results, the load loss anomaly of the transformer to be analyzed is located.

[0005] In one embodiment, the root cause analysis of the transformer to be analyzed is performed based on the transformer data and the load loss mechanism model of the transformer, and before the step of obtaining the root cause analysis result, the step further includes: Obtain multiple groups of measured values ​​of additional loss influencing parameters of the transformer and additional loss measured values ​​corresponding to each group of measured values; The preset fitting model is trained based on multiple groups of measured values ​​and additional loss measured values ​​to obtain an additional loss fitting model of the transformer; wherein the preset fitting model includes an extremely random tree model.

[0006] In one embodiment, the additional loss influencing parameters include high voltage resistance, low voltage resistance, short circuit impedance, low voltage coil outer diameter, low voltage coil inner diameter, high voltage coil outer diameter and high voltage coil inner diameter.

[0007] In one embodiment, the resistance loss calculation model is: ; in, is the high voltage rated phase current, is the high voltage resistor, is the low voltage rated phase current, It is a low voltage resistor.

[0008] In one embodiment, the root cause analysis is performed on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer, and the step of obtaining the root cause analysis result includes: Based on transformer data and load loss mechanism model, the Adtributor algorithm is used to perform root cause analysis on the transformer to be analyzed, and the main root causes of the load loss of the transformer to be analyzed are determined; wherein the main root causes include resistance loss root causes or additional loss root causes; Conduct root cause analysis based on transformer data and the loss model corresponding to the main root causes to determine the potential root causes of load loss under the main root causes; Obtain root cause analysis results based on the main root causes and potential root causes.

[0009] In one embodiment, the measured data includes measured values ​​of resistance loss influencing parameters; the standard data includes designed values ​​of resistance loss influencing parameters; the resistance loss influencing parameters include high voltage resistance and low voltage resistance; The root cause analysis is based on the transformer data and the loss model corresponding to the main root cause. The steps to determine the potential root causes of load loss under the main root cause include: When the main root cause is the resistance loss root cause, for each resistance loss influencing parameter, the measured value of the corresponding resistance loss influencing parameter is replaced with the design value; Based on the resistance loss calculation model, the resistance loss after parameter replacement is determined; Based on the measured values ​​of the resistance loss influencing parameters, the designed values ​​of the resistance loss influencing parameters, the resistance loss after parameter replacement, and the resistance loss before parameter replacement, the surprisingness and the degree of explanation of the resistance loss influencing parameters are determined; Based on the surprise of all resistance loss influencing parameters and the ranking of the degree of explanation of all resistance loss influencing parameters, the potential root causes of load loss under the resistance loss root cause are determined.

[0010] In one embodiment, the measured data includes measured values ​​of the additional loss influencing parameters; the standard data includes designed values ​​of the additional loss influencing parameters; The root cause analysis is based on the transformer data and the loss model corresponding to the main root cause. The steps to determine the potential root causes of load loss under the main root cause include: When the main root cause is the additional loss root cause, for each additional loss influencing parameter, the measured value of the corresponding additional loss influencing parameter is replaced by the design value; Based on the additional loss fitting model, the additional loss after parameter replacement is determined; Based on the measured values ​​of the additional loss influencing parameters, the designed values ​​of the additional loss influencing parameters, the additional losses after parameter replacement, and the additional losses before parameter replacement, the surprisingness and the degree of explanation of the additional loss influencing parameters are determined; Based on the surprise of all the additional loss influencing parameters and the ranking of the degree of explanation of all the additional loss influencing parameters, the potential root causes of the load loss under the additional loss root causes are determined.

[0011] In addition, to achieve the above-mentioned purpose, the present application also proposes a transformer abnormal loss locating device, the device comprising: a memory, a processor and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the transformer abnormal loss locating method as described above.

[0012] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the transformer abnormal loss locating method as described above are implemented.

[0013] In addition, to achieve the above-mentioned purpose, the present application also proposes a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the transformer abnormal loss locating method as described above are implemented.

[0014] One or more technical solutions proposed in this application have at least the following technical effects: In the transformer abnormal loss positioning method provided in the present application, transformer data of the transformer to be analyzed can be obtained; the transformer data includes measured data and corresponding standard data; thus, root cause analysis can be performed on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer to obtain the root cause analysis result; wherein, the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values; based on the root cause analysis result, the load loss anomaly of the transformer to be analyzed can be positioned.

[0015] Compared with the related art that directly relies on experience judgment to locate the cause of abnormal load loss, the present application can use transformer data and transformer load loss mechanism model to perform root cause analysis of abnormal load loss. The load loss mechanism model comprehensively considers various loss influencing factors and can provide theoretical support for transformer abnormality tracing. It no longer relies on manual experience judgment and can locate the cause of the abnormality more specifically, reducing the analysis deviation caused by subjective factors such as personal experience, thereby improving the accuracy of abnormal loss positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A flowchart diagram of the first embodiment of the method for locating abnormal transformer loss in this application; Figure 2 A comparison chart of training results for multiple preset fitting models; Figure 3 A schematic diagram of a simplified process of a transformer abnormal loss location method according to an example provided in Embodiment 1 of the present application; Figure 4 This is a schematic diagram of removing outliers using the interquartile range method; Figure 5 is a schematic diagram of consistency analysis results of a transformer of an example; Figure 6 This is an overall process flow chart of root cause analysis for an example; Figure 7 This is a schematic diagram of the device structure of the hardware operating environment involved in the transformer abnormal loss locating method in the embodiment of the present application.

[0019] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0020] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0021] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0022] The main solution of the embodiment of the present application is: obtaining transformer data of the transformer to be analyzed; the transformer data includes measured data and corresponding standard data; performing root cause analysis on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer to obtain the root cause analysis result; the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values; based on the root cause analysis result, the load loss anomaly of the transformer to be analyzed is located.

[0023] In related technologies, when encountering the problem of abnormal transformer load loss, the cause of the abnormality is generally guessed based on experience, and it is difficult to locate the specific cause in a targeted and theoretically supported manner; for manufacturers, it is difficult to locate the root cause of abnormal transformer load loss, which makes it difficult to control the quality consistency of transformer products during the production process. Therefore, it is necessary to find a method that can accurately locate the root cause of abnormal transformer load loss.

[0024] The present application provides a solution that can use transformer data and a transformer load loss mechanism model to perform root cause analysis of load loss anomalies. The load loss mechanism model comprehensively considers a variety of loss influencing factors and can provide theoretical support for transformer abnormality tracing. It is no longer solely dependent on manual experience and judgment, and can more specifically locate the cause of the anomaly, reducing analysis deviations caused by subjective factors such as personal experience, thereby improving the accuracy of abnormal loss positioning.

[0025] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, etc., or an electronic device capable of realizing the above functions, etc. The following takes the transformer abnormal loss location device as an example to illustrate this embodiment and the following embodiments.

[0026] Based on this, the embodiment of the present application provides a method for locating abnormal loss of a transformer, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the method for locating abnormal transformer loss in the present application.

[0027] In this embodiment, the transformer abnormal loss location method includes steps S100 to S300: Step S100, obtaining transformer data of a transformer to be analyzed.

[0028] Among them, the transformer data includes measured data and corresponding standard data.

[0029] Step S200: performing root cause analysis on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer to obtain a root cause analysis result.

[0030] Among them, the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model. The additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values.

[0031] Step S300: Based on the root cause analysis result, the load loss anomaly of the transformer to be analyzed is located.

[0032] Specifically, transformer data may include measured data of transformer products and corresponding standard data. The measured data may be obtained by measuring the transformer using sensors or measuring equipment, and the corresponding standard data (such as design parameters or reference data under ideal conditions) may be obtained by consulting transformer design documents or industry specifications and standards. After obtaining the transformer data, the transformer data and the load loss mechanism model of the transformer may be used to perform root cause analysis on the transformer to be analyzed, and obtain the root cause analysis results. The root cause analysis results may include multiple factors that may cause abnormal load loss of the transformer and the proportion of the causes. Therefore, according to the root cause analysis results, the abnormal cause of the load loss of the transformer to be analyzed can be determined, and the abnormal load loss can be accurately located, so as to improve the transformer products in a targeted manner; for example, the root cause analysis results show that the proportion of high load loss caused by resistance loss is high. At this time, it can be determined that the main abnormal cause of load loss is resistance loss according to the root cause analysis results, and the transformer products can be analyzed and improved accordingly.

[0033] It should be noted that the above-mentioned load loss mechanism model is a model used to describe the load loss formation principle of the transformer, which can include a resistance loss calculation model and an additional loss fitting model; among them, the resistance loss calculation model is suitable for describing the energy loss caused by the resistance part of the transformer, and the additional loss fitting model is used to describe the energy loss generated by the remaining parts (such as eddy current loss and stray loss, etc.). In a feasible implementation, the load loss mechanism model of the transformer can be determined under the guidance of business experts and combined with a large number of existing transformer types and data. , .in, is the resistance loss calculation model, is the lead loss model of the transformer, The resistance loss calculation model can be determined based on Ohm's law and Joule's law, and is composed of high-voltage resistance loss and low-voltage resistance loss. Can be: ;in, is the high voltage rated phase current, is the high voltage resistor, is the low voltage rated phase current, is a low-voltage resistor; the lead loss model of the transformer can be regarded as a fixed constant, which is determined according to the design standard of the transformer; since there is no standard calculation model to describe the additional loss of the transformer, the additional loss fitting model of the transformer can be obtained by fitting multiple groups of measured values ​​of the additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values. Therefore, before step S200, steps A100 to A200 may be included to obtain the additional loss fitting model of the transformer: Step A100, obtaining multiple groups of measured values ​​of additional loss influencing parameters of the transformer and additional loss measured values ​​corresponding to each group of measured values.

[0034] Step A200: training a preset fitting model based on multiple groups of measured values ​​and additional loss measured values ​​to obtain an additional loss fitting model of the transformer.

[0035] Among them, the preset fitting model includes an extremely random tree model.

[0036] Specifically, the additional loss influencing parameters may include but are not limited to high-voltage resistance, low-voltage resistance, short-circuit impedance, low-voltage coil outer diameter, low-voltage coil inner diameter, high-voltage coil outer diameter and high-voltage coil inner diameter, etc. For multiple transformers of the same model, the corresponding multiple sets of measured values ​​of additional loss influencing parameters can be obtained; since the additional loss is not convenient to measure directly, the resistance loss calculation model described above can be used for the measured values ​​of each set of additional loss influencing parameters. Calculate the resistance loss value corresponding to the measured value of the group of additional loss influencing parameters; and determine the fixed lead loss value corresponding to the group of measured values ​​according to the transformer model (design data), and the load loss of the transformer can be obtained through actual measurement; substitute the resistance loss value, lead loss value and measured load loss value corresponding to each group of measured values ​​into the above load loss mechanism model , an additional loss value can be calculated, and the additional loss value can be regarded as the additional loss measured value corresponding to the measured value of the additional loss influencing parameter of this group.

[0037] The multiple groups of measured values ​​and the corresponding additional loss measured values ​​obtained can be used as training data to train the preset fitting model to obtain the additional loss fitting model of the transformer; preferably, the preset fitting model can be an extremely random tree model. In practical applications, a variety of representative machine learning models can be selected as preset fitting models for model training, and then the one with the best training result is selected from the trained models as the additional loss fitting model. For example, the least squares support vector machine model LS_SVM, the gradient boosting model Gradient Boost and the extremely random tree model Extra Tree can be used as preset fitting models for model training. The preset fitting model can continuously learn based on the training data, fit the measured value of the input additional loss influencing parameter into an output value, and the output value is the additional loss fitting value. By comparing the additional loss fitting value with the additional loss measured value, the following can be obtained. Figure 2 A comparison chart of training results of multiple preset fitting models is shown. Figure 2 In the figure, the dots represent the measured additional loss values ​​of each transformer. Figure 2 The horizontal axis is the transformer number, and the vertical axis is the additional loss; Figure 2 It can be seen that the curve formed by the additional loss fitting value obtained by the extreme random tree model fitting perfectly connects the points of the additional loss actual value, that is, the additional loss fitting model obtained by the extreme random tree model training can fit the additional loss of the transformer more accurately. In addition, in order to ensure that the optimal model is selected, a comprehensive judgment can also be made based on the quantitative indicators of each fitting model obtained after completing the model training. As shown in Table 1, Table 1 shows the quantitative indicators of the three models.

[0038] Table 1

[0039] In the above Table 1, is the root mean square error. From the order of magnitude, it can be considered that the regression effect of the model is different from the true value on average. , when the fitted value and the true value (measured value) are completely consistent, it is equal to 0, that is, a perfect model; the larger the model error, the larger the value; ; is the measured value of the i-th group of training data, is the fitted value of the i-th group of training data.

[0040] is the mean absolute percentage error. When the fitted value is exactly consistent with the true value, 0% indicates a perfect model. A value greater than 100% indicates a poor quality model; .

[0041] It is a custom accuracy calculation method. k=1 means the accuracy when the difference between the fitted value and the true value is less than 1, k=10 means the accuracy when the difference between the fitted value and the true value is less than 10, and similarly, 50 and 100 can be obtained. ; is the counting function, Count the number of fitted values, The difference between the statistical fitted value and the true value is less than The number of

[0042] According to the quantitative indicators in Table 1 and Figure 2 The comparison chart of the model training results shown in the figure shows that the model obtained by training the extremely random tree model is the optimal additional loss fitting model.

[0043] The additional loss fitting model can better describe the intrinsic relationship between additional loss influencing parameters such as high-voltage resistance, low-voltage resistance, short-circuit impedance, low-voltage coil outer diameter, low-voltage coil inner diameter, high-voltage coil outer diameter and high-voltage coil inner diameter and additional loss, so that the root cause analysis under the additional loss dimension can be carried out based on the additional loss fitting model in the future, and the root cause of the abnormal transformer load loss can be accurately located.

[0044] In a feasible implementation manner, step S200 may specifically include steps S210 to S230: Step S210 , based on the transformer data and the load loss mechanism model, an Adtributor algorithm is used to perform a root cause analysis on the transformer to be analyzed, and a main root cause of the load loss of the transformer to be analyzed is determined.

[0045] Among them, the main root causes include resistance loss root causes or additional loss root causes.

[0046] Step S220: performing root cause analysis based on the transformer data and the loss model corresponding to the main root cause, and determining potential root causes of the load loss under the main root cause.

[0047] Step S230: Obtain root cause analysis results based on the main root causes and potential root causes.

[0048] Specifically, the Adtributor algorithm can be used as a basis for the obtained transformer data and the load loss mechanism model constructed above to perform root cause analysis on the transformer to be analyzed, so as to determine the main root cause of the load loss of the transformer to be analyzed. It can be seen from the above content that the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, so according to the model composition, it can be determined that the main root cause may include the resistance loss root cause and / or the additional loss root cause. It should be noted that the above content mentioned that the load loss mechanism model can also include a lead loss model, but since the lead loss of the transformer is generally a fixed constant determined according to the design standard of the transformer, its impact on the load loss is not considered.

[0049] The above-mentioned Adtributor algorithm is a multi-dimensional anomaly root cause analysis method proposed by Microsoft Research in 2014. It can quantify the root cause through explanatory power (EP) and surprise (S). The difference between the expected value and the measured value distribution is defined as surprise. The dimensions are sorted by calculating the surprise of the dimension (the sum of the surprises of all elements in the dimension) to determine the dimension where the root cause is located. The explanatory power of each element is calculated within the dimension. When the sum of the explanatory power of the elements exceeds the threshold, these elements are considered to be the root cause; the root cause analysis target is called KPI, and the similarity between the prior probability and the posterior probability of KPI can be measured by relative entropy to determine whether the dimension has surprise. The greater the surprise, the greater the possibility that the dimension is the root cause. Relative entropy is also called KL divergence. Compared with KL divergence, JS divergence, as a deformation of KL divergence, has good symmetry, adaptability to zero values, and more accurate similarity judgment. Therefore, JS divergence is used to calculate surprise. The formula for calculating the surprise of the jth element of dimension i is as follows: ;in, p is the prior probability, indicating the expected value distribution , is the expected value of the element, is the KPI expected value; q is the posterior probability, which represents the true value distribution , is the measured value of the element, It is the actual value of KPI. S The value is all under this dimension S Sum of values: ; For each dimension, if the fluctuation of the element accounts for a larger proportion of the fluctuation of the abnormal KPI, it is considered that the element can better explain the occurrence of the KPI anomaly. This definition is the explanatory degree, and the calculation formula of the explanatory degree is as follows: ,in, is the real value of KPI, is the expected value, and m is the abnormality indicator. The EP value can be positive, negative, or greater than 100%, but the sum of all EP elements in each dimension must be 100%. A positive EP indicates that it may be the root cause of the abnormality, a negative EP indicates that it is not the root cause of the abnormality, and a value greater than 100% indicates that it has a very obvious positive correlation with the KPI abnormality. When analyzing quantitative KPIs, according to the formula, the fluctuation ratio of a certain element may exceed 100%, and the explanation threshold TEP may be set to be greater than 1.

[0050] When combining transformer data and load loss mechanism model and using Adtributor algorithm to analyze the main root causes of the transformer to be analyzed, the root cause analysis problem can be positioned as a one-dimensional multi-level root cause location problem. The query dimension of the Adtributor algorithm originally used for multi-dimensional abnormal root cause analysis is defined in one dimension, and the number of search levels is increased to adapt to the actual load loss root cause analysis. For example, the resistance loss of the transformer data can be , Lead loss and additional losses Multiple groups of measured values, standard values ​​(design values) of the same type of transformer, and the load loss mechanism model are input into the load loss root cause analysis model built based on the Adtributor algorithm to analyze the main root causes under the load loss dimension. The load loss root cause analysis model can The measured value is replaced with the corresponding standard value of resistance loss, and the load loss mechanism model is used Determine the corresponding load loss, and then substitute it into the above-mentioned surprise and explanation calculation formula for calculation. At this time, is the measured value of load loss, For the general The load loss obtained after replacing it with the standard value is, is the measured value of resistance loss, is the standard value of resistance loss. Similarly, the additional loss The measured value is replaced with the corresponding additional loss standard value, and the corresponding surprise and explanation degree are calculated. Through the comprehensive ranking of surprise and explanation, the main root causes of transformer load loss can be determined. The main root cause only determines the main level (resistance loss level or additional loss level) of the cause of transformer load loss. There are more potential influencing factors under resistance loss and additional loss. Therefore, in order to more accurately locate the cause of abnormal transformer load loss, after determining the main root cause, it is necessary to conduct further drill-down root cause analysis based on the transformer data and the specific loss model corresponding to the main root cause to determine the potential root cause under the main root cause level. Therefore, a comprehensive root cause analysis report can be generated by combining the main root cause and the potential root cause, so as to accurately locate the root cause of abnormal load loss.

[0051] In a feasible implementation, the measured data of the transformer data may include the measured values ​​of the resistance loss influencing parameters; the standard data may include the design values ​​of the resistance loss influencing parameters; the resistance loss influencing parameters include but are not limited to high-voltage resistance and low-voltage resistance. In this implementation, step S220 may specifically include steps S221 to S224 to implement the drill-down root cause analysis under the resistance loss dimension: Step S221 : when the main root cause is the resistance loss root cause, for each resistance loss influencing parameter, the measured value of the corresponding resistance loss influencing parameter is replaced with the designed value.

[0052] Step S222: determining the resistance loss after parameter replacement based on the resistance loss calculation model.

[0053] Step S223, determining the surprisingness and explanatory power of the resistance loss influencing parameter based on the measured value of the resistance loss influencing parameter, the designed value of the resistance loss influencing parameter, the resistance loss after parameter replacement, and the resistance loss before parameter replacement.

[0054] Step S224: determining the potential root cause of the load loss under the resistance loss root cause based on the surprise of all resistance loss influencing parameters and the ranking order of the degree of explanation of all resistance loss influencing parameters.

[0055] Specifically, when it is determined that the main root cause affecting load loss is resistance loss, the resistance loss calculation model and transformer data can be combined to perform root cause analysis at the resistance loss level. Similarly, the resistance loss root cause analysis model constructed by the Adtributor algorithm can be used to analyze the potential root causes under the resistance loss dimension. For example, the measured values ​​of resistance loss influencing parameters such as the high-voltage resistance and low-voltage resistance of the transformer, the corresponding design values, and the above-mentioned resistance loss calculation model can be input into the resistance loss root cause analysis model. For each resistance loss influencing parameter, the resistance loss root cause analysis model can replace the measured value with the corresponding design value, and use the resistance loss calculation model to determine the corresponding resistance loss, and then substitute it into the above-mentioned surprise and explanation calculation formulas for calculation.

[0056] Taking the cabin transformer as an example, generally, a core with a low-voltage coil and a high-voltage coil can be regarded as a set of transformer units. Three sets of transformer units assembled together can constitute a cabin transformer body. Therefore, for the cabin transformer, there are 6 parameters affecting its resistance loss, namely, the first high-voltage resistance , the second high voltage resistor , the third high voltage resistor , the first low voltage resistor , Second low voltage resistor and the third low voltage resistor ; The resistance loss root cause analysis model can be the first high voltage resistor The measured value is replaced with the corresponding design value, and the corresponding resistance loss is calculated using the resistance loss calculation model, and then substituted into the above-mentioned surprise and explanation calculation formula for calculation. At this time, is the measured value of resistance loss, For the general The resistance loss obtained after replacing it with the standard value is, The first high voltage resistor The measured value of for Similarly, the explanation and surprise calculations for the other resistance loss influencing parameters are also consistent and will not be repeated here.

[0057] After obtaining the surprise and explanation of all resistance loss influencing parameters, all the surprises and explanations can be sorted respectively, and a comprehensive analysis can be conducted based on the sorting results of the two. For example, resistance loss influencing parameters with high surprise and high explanation can be given priority and determined as potential root causes of load loss under the resistance loss root cause.

[0058] In another feasible implementation, the measured data of the transformer data may include the measured values ​​of the additional loss influencing parameters; the standard data includes the design values ​​of the additional loss influencing parameters. In this implementation, step S220 may specifically include steps S225 to S228 to implement the drill-down root cause analysis under the additional loss dimension: Step S225: When the main cause is the additional loss cause, for each additional loss influencing parameter, the actual measured value of the corresponding additional loss influencing parameter is replaced with the design value.

[0059] Step S226, determining the additional loss after parameter replacement based on the additional loss fitting model.

[0060] Step S227, determining the surprisingness and explanatory power of the additional loss influencing parameter based on the measured value of the additional loss influencing parameter, the designed value of the additional loss influencing parameter, the additional loss after parameter replacement, and the additional loss before parameter replacement.

[0061] Step S228: determining the potential root cause of the load loss under the additional loss root cause based on the surprise of all additional loss influencing parameters and the ranking order of the degree of explanation of all additional loss influencing parameters.

[0062] Specifically, similar to the drill-down root cause analysis under the resistance loss dimension, when it is determined that the main root cause affecting the load loss is the additional loss root cause, the additional loss fitting model and transformer data can be combined to perform root cause analysis at the additional loss level. Similarly, the additional loss root cause analysis model constructed by the Adtributor algorithm can be used to analyze the potential root causes under the additional loss dimension. The measured values ​​of the additional loss influencing parameters such as the high-voltage resistance, low-voltage resistance, short-circuit impedance, low-voltage coil outer diameter, low-voltage coil inner diameter, high-voltage coil outer diameter and high-voltage coil inner diameter of the transformer, the corresponding design values ​​and the additional loss fitting model described above that can be used to describe the inherent correspondence between the additional loss influencing parameters and the additional loss can be input into the additional loss root cause analysis model. For each additional loss influencing parameter, the additional loss root cause analysis model can replace the measured value of the parameter with the corresponding design value, and use the additional loss fitting model to calculate the corresponding additional loss, and then substitute it into the above-mentioned surprise and explanation calculation formula for calculation. Taking the cabin transformer as an example, the cabin transformer is composed of three transformer units, so its additional loss influencing parameters can be 19, namely the first high-voltage resistance , the second high voltage resistor , the third high voltage resistor , the first low voltage resistor , Second low voltage resistor , the third low voltage resistor , short-circuit impedance UK, first low-voltage coil outer diameter r_WD1, second low-voltage coil outer diameter r_WD2, third low-voltage coil outer diameter r_WD3, first low-voltage coil inner diameter r_ND1, second low-voltage coil inner diameter r_ND2, third low-voltage coil inner diameter r_ND3, first high-voltage coil outer diameter r_WG1, second high-voltage coil outer diameter r_WG2, third high-voltage coil outer diameter r_WG3, first high-voltage coil inner diameter r_NG1, second high-voltage coil inner diameter r_NG2 and third high-voltage coil inner diameter r_NG3; the additional loss root cause analysis model can convert the first high-voltage resistor The measured value of is replaced with the corresponding design value, and the corresponding additional loss is fitted using the additional loss fitting model, and then substituted into the above-mentioned surprise and explanation calculation formula for calculation. At this time, is the measured value of additional loss, For the general The additional loss obtained after replacing it with the standard value is, The first high voltage resistor The measured value of for Similarly, the explanation and surprise calculations for the other resistance loss influencing parameters are also consistent and will not be repeated here.

[0063] After obtaining the surprise and explanation of all additional loss influencing parameters, all the surprises and explanations can be sorted respectively, and the potential root cause of load loss under the resistance loss root cause can be determined based on a comprehensive analysis of the sorting results of the two.

[0064] It can be understood that the transformer abnormal loss locating method provided in this embodiment can use transformer data and the transformer load loss mechanism model to perform root cause analysis of load loss abnormality. The load loss mechanism model comprehensively considers various loss influencing factors and can provide theoretical support for transformer abnormality tracing. It is no longer solely dependent on manual experience and judgment, and can locate the cause of the abnormality more specifically, reducing the analysis deviation caused by subjective factors such as personal experience. This can improve the accuracy of abnormal loss locating, assist the production end in reverse control, and has guiding significance for industrial units to improve quality.

[0065] For example, to help understand the implementation process of the transformer abnormal loss location method in Example 1, please refer to Figure 3 , Figure 3 Taking the cabin transformer as an example, a brief flow chart of a transformer abnormal loss location method is provided. Specifically: Transformer data of seven types of cabin transformers (models 1 to 7) can be collected, and the transformer data can include all data in the four aspects of design, manufacturing, materials, and testing of each type of cabin transformer, totaling 564 cabin transformer bodies and corresponding coil data; among them, the design data includes the inner and outer diameters, number of layers, number of turns, wire specifications, resistance, and rated current of high and low voltage coils designed for different models; the manufacturing data includes the height, weight, measured number of turns, actual measured inner and outer diameters, resistance, temperature, and humidity of the high and low voltage coils; the material data includes the measured thickness of the wire, the material of the conductor, the size, and the resistivity; the test data includes the operating temperature, short-circuit impedance, load loss, ambient temperature, test current, and measured values ​​of the tapping resistance of the transformer body.

[0066] In order to ensure the accuracy of the subsequent root cause analysis, the transformer data can be subjected to quality analysis and evaluation screening. For example, the specific evaluation objects of the data can include the operating temperature T, the low-voltage resistance value when the low-voltage gear resistance is converted to 75 degrees. (Including the first low voltage resistor , Second low voltage resistor and the third low voltage resistor ), high voltage resistance conversion of high voltage resistance value at 75 degrees (Including the first high voltage resistor , the second high voltage resistor and the third high voltage resistor ), low-voltage coil inner diameter r_ND (including the first low-voltage coil inner diameter r_ND1, the second low-voltage coil inner diameter r_ND2 and the third low-voltage coil inner diameter r_ND3), low-voltage coil outer diameter r_WD (including the first low-voltage coil outer diameter r_WD1, the second low-voltage coil outer diameter r_WD2 and the third low-voltage coil outer diameter r_WD3), high-voltage coil inner diameter r_NG (including the first high-voltage coil inner diameter r_NG1, the second high-voltage coil inner diameter r_NG2 and the third high-voltage coil inner diameter r_NG3), high-voltage coil outer diameter r_WG (including the first high-voltage coil outer diameter r_WG1, the second high-voltage coil outer diameter r_WG2, the third high-voltage coil outer diameter r_WG3), wire foil actual weight weight, coil resistivity ρ, short-circuit impedance UK, load loss Y, etc. After data quality analysis, it can be found that the data has missing items, omitted items, items that do not conform to the normal distribution, and too little data.

[0067] The transformer data is then preprocessed, including outlier processing (converting resistance and temperature according to the formula), missing value filling (average method, simultaneous method), outlier removal (interquartile range method), and normalization (maximum and minimum standardization). The resistance and temperature conversion formula is: ; For resistance to measure temperature, For temperature The measured resistance value at For the general The resistance value is converted to the temperature of 75 degrees. The average value method is to fill in the missing values ​​of phase / line resistance with the average of the five previous and next data. The simultaneous method is to fill in the missing values ​​with the ambient temperature measured at the same time as the standard. The interquartile range method is to treat the data beyond 1.5 times the distance between the number at the 1 / 4 position (denoted as Q1) and the number at the 3 / 4 position (denoted as Q3) in all the data as outliers and remove them. Figure 4 As shown, Figure 4 This is a schematic diagram of the interquartile range method to remove outliers; the interval between Q1 and Q3 is expressed as IQR=Q3-Q1. The maximum and minimum normalization method is to change the data to the range of [0, 1]. The specific normalization formula is as follows: , X min is the minimum value in the data, X max is the maximum value in the data, X is the processed data.

[0068] After the above processing, there are 271 data in total, including 83 for model 1, 19 for model 2, 111 for model 3, 1 for model 4, 45 for model 5, and 12 for model 6. Under the guidance of business experts, the load loss mechanism model of the transformer was determined based on the existing data quality and type. , .in, is the resistance loss calculation model, is the lead loss model of the transformer, Then it is the additional loss fitting model of the transformer. About the load loss mechanism model And its components, such as the resistance loss calculation model And additional loss fitting model For the specific content of the selection, please refer to the relevant description in the aforementioned embodiment 1, which will not be repeated here.

[0069] After collecting the transformer data, the load loss can be analyzed for consistency. The 271 cabin transformers without missing data obtained above are used as the data range for consistency analysis. The design value is used as the standard value. The upper limit UCL and lower limit DCL of the standard value are calculated with a floating range of ±1%, and the cabin transformer (hereinafter referred to as the entity) is defined according to the classification rules shown in the following formula:

[0070] Among them, Y is the measured load loss of each entity, and Y* is the corresponding design load loss; Class A is the entity that exceeds the upper limit of the design load loss (1.01Y), and Class E is the entity that is lower than the lower limit of the design load loss (0.99Y). Class A and Class E are both considered abnormal entities. Consistency analysis can visualize the distribution of abnormal entities through UCL and DCL control lines. Figure 5 As shown, Figure 5 Taking a certain model as an example, the schematic diagram of the consistency analysis results is introduced. There are 111 available transformers of this model, including 8 Class A transformers, 44 Class B transformers, 0 Class C transformers, 46 Class D transformers, and 13 Class E transformers; abnormal transformers (Class A and Class E) with load losses exceeding the design upper and lower limits of 1% account for 18.92%, and 81.08% of the transformers of this model have load losses within the 1% control line, with good load loss consistency; among the abnormal transformers, 61.9% have load losses lower than DCL, that is, the main abnormal performance of this transformer model is low load loss.

[0071] Then, the first-level root cause analysis model (i.e., load loss root cause analysis model) is constructed using the improved Adtributor algorithm to conduct the first-level root cause analysis of load loss and determine the main root cause of the abnormal load loss. The aforementioned 271 non-missing data entities are used as the data scope for this main root cause analysis. Taking the above example as an example, the root cause proportions of various entities are statistically shown in Table 2 below: Table 2

[0072] From the data in Table 2, we can know that the resistance loss It is the main reason for the high load loss and the additional loss This is the main reason for the low load loss.

[0073] Based on the analysis results of the above main root causes, we can further drill down to the root cause analysis in the resistance loss dimension or the additional loss dimension. Figure 3 As shown, if the main root cause is located at the resistance loss , then the resistance loss root cause analysis model can be constructed by improving the Adtributor algorithm as the second-level root cause analysis model, and the root cause analysis of the loss can be continued, and the resistance loss can be drilled down to the coil parameter level. Taking a certain body of model 3 as an example, the surprise of each high-voltage resistor and low-voltage resistor of the body can be calculated based on the JS divergence. The higher the surprise, the higher the possibility that the resistance loss influencing parameter is the root cause. The calculation method of the surprise here can be found in the detailed description of the embodiment part, and will not be repeated here. The resistance loss influencing parameters of the body of this example (including the first low-voltage resistor , Second low voltage resistor , the third low voltage resistor , the first high voltage resistor , the second high voltage resistor and the third high voltage resistor The surprise calculation results of the resistance loss of the main body are shown in Table 3. It can be seen from Table 3 that the surprise of the resistance loss influencing parameters is ranked from large to small as follows: > > > > > .

[0074] Table 3

[0075] Then calculate the explanatory degree of each high-voltage resistor and low-voltage resistor for the resistance loss fluctuation. The higher the explanatory degree, the stronger the explanatory power for the loss anomaly. The calculation method of the explanatory degree refers to the relevant content of the aforementioned embodiment 1, which will not be repeated here. The calculation results of the explanatory degree of the resistance loss influencing parameters are shown in Table 4 below: Table 4

[0076] Combined with the surprise ranking and explanation degree, the root causes that can explain 100%, 80%, and 60% of the load loss anomalies can be located (in order to reduce the difficulty of adjusting model parameters during root cause analysis, the explanation degree threshold TEP can be uniformly designed to be 100%, 80%, and 60%). The final root cause analysis results under the resistance loss dimension are shown in Table 5: Table 5

[0077] As shown in Table 5, It can explain 60% of the anomalies. and It can explain 80% of the anomalies. , Can explain 100% of the anomalies.

[0078] If the main cause is located at the additional loss , we can improve the Adtributor algorithm to build an additional loss root cause analysis model as the second-level root cause analysis model, and continue to drill down the root cause analysis of the loss, and drill down the additional loss to the coil parameter level. The following takes two bodies of a certain model as an example, and calculates the additional loss influencing parameters of the two bodies (including the first high-voltage resistor) according to the JS divergence , the second high voltage resistor , the third high voltage resistor , the first low voltage resistor , Second low voltage resistor , the third low voltage resistor , short-circuit impedance UK, first low-voltage coil outer diameter r_WD1, second low-voltage coil outer diameter r_WD2, third low-voltage coil outer diameter r_WD3, first low-voltage coil inner diameter r_ND1, second low-voltage coil inner diameter r_ND2, third low-voltage coil inner diameter r_ND3, first high-voltage coil outer diameter r_WG1, second high-voltage coil outer diameter r_WG2, third high-voltage coil outer diameter r_WG3, first high-voltage coil inner diameter r_NG1, second high-voltage coil inner diameter r_NG2 and third high-voltage coil inner diameter r_NG3), the ranking results of the surprise calculation of the two entities are consistent, both of which are:

[0079] Then calculate each additional loss influence parameter. However, for the explanation degree of additional loss fluctuation, the calculation method refers to the relevant content of the above-mentioned embodiment 1, which will not be repeated here. The calculation results of the explanation degree of additional loss influence parameters are shown in the following Table 6: Table 6

[0080] Combined with the surprise ranking and explanation degree, the root causes that can explain 100%, 80%, and 60% of the additional loss anomalies are located respectively. The final root cause analysis results under the additional loss dimension are shown in Table 7: Table 7

[0081] From Table 7, we can see that for entity 1, 60% of the anomalies can be explained, and more root causes of anomalies are located in other parameters that have not been collected so far; for entity 2, as the anomaly level increases, the root causes also gradually increase, and the overall anomaly is caused by the combined effect of all factors.

[0082] Figure 6 An example of the overall process flow diagram of root cause analysis is shown as follows: Figure 6 As shown in the figure, in practical applications, after data quality assessment and processing of transformer data, when performing root cause analysis, the load loss mechanism model, resistance loss calculation model and additional loss fitting model can be combined to obtain the loss model of the cabin transformer. When performing root cause analysis of load loss according to the loss model of the cabin transformer, the load loss mechanism model can be used as the basis. The first-level root cause analysis of load loss can be performed by improving the Adtributor algorithm to determine whether the dimension to which the root cause belongs is resistance loss, additional loss or other comprehensive dimensions. Then, according to the determined dimension, the resistance loss calculation model and the additional loss fitting model are combined to perform the second-level root cause analysis of resistance loss or additional loss by improving the Adtributor algorithm. Finally, the elements of the root cause are determined, and the root causes under the three levels of threshold TEP of 100%, 80% and 60% are obtained respectively, so as to locate the root cause of the abnormal load loss of the transformer.

[0083] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the transformer abnormal loss locating method of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0084] The present application provides a transformer abnormal loss locating device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the transformer abnormal loss locating method in the above-mentioned embodiment 1.

[0085] Reference below Figure 7 , which shows a schematic diagram of the structure of a transformer abnormal loss location device suitable for implementing the embodiment of the present application. The transformer abnormal loss location device in the embodiment of the present application may include but is not limited to mobile terminals such as laptop computers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions: tablet computers), etc., and fixed terminals such as desktop computers, etc. Figure 7 The transformer abnormal loss locating device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0086] like Figure 7 As shown, the transformer abnormal loss location device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the transformer abnormal loss location device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the transformer abnormal loss locating device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a transformer abnormal loss locating device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

[0087] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0088] The transformer abnormal loss positioning device provided by the present application adopts the transformer abnormal loss positioning method in the above embodiment, which can solve the technical problem that it is difficult to accurately locate the root cause of the abnormal transformer load loss in the related art. Compared with the related art, the beneficial effects of the transformer abnormal loss positioning device provided by the present application are the same as the beneficial effects of the transformer abnormal loss positioning method provided by the above embodiment, and the other technical features of the transformer abnormal loss positioning device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0089] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0090] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0091] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the transformer abnormal loss locating method in the above-mentioned embodiment.

[0092] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0093] The computer-readable storage medium may be included in the transformer abnormal loss locating device; or may exist independently without being assembled into the transformer abnormal loss locating device.

[0094] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the transformer abnormal loss locating device, the transformer abnormal loss locating device: obtains transformer data of the transformer to be analyzed; the transformer data includes measured data and corresponding standard data; based on the transformer data and the load loss mechanism model of the transformer, a root cause analysis is performed on the transformer to be analyzed to obtain the root cause analysis result; the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and the additional loss measured values ​​corresponding to each group of measured values; based on the root cause analysis result, the load loss anomaly of the transformer to be analyzed is located.

[0095] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0096] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0097] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0098] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned transformer abnormal loss location method, and can solve the technical problem in the related art that it is difficult to accurately locate the root cause of the abnormal transformer load loss. Compared with the related art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the transformer abnormal loss location method provided by the above-mentioned embodiment, and will not be repeated here.

[0099] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned transformer abnormal loss locating method when executed by a processor.

[0100] The computer program product provided by the present application can solve the technical problem in the related art that it is difficult to accurately locate the root cause of the abnormal load loss of the transformer. Compared with the related art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the transformer abnormal loss location method provided by the above embodiment, and will not be repeated here.

[0101] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for locating abnormal transformer loss, characterized in that: The transformer abnormal loss positioning method comprises: Obtaining transformer data of the transformer to be analyzed; the transformer data includes measured data and corresponding standard data; Based on the transformer data and the load loss mechanism model of the transformer, a root cause analysis is performed on the transformer to be analyzed to obtain a root cause analysis result; the load loss mechanism model includes a resistance loss calculation model and an additional loss fitting model, and the additional loss fitting model is obtained by fitting multiple groups of measured values ​​of additional loss influencing parameters and additional loss measured values ​​corresponding to each group of the measured values; Based on the root cause analysis result, the load loss anomaly of the transformer to be analyzed is located.

2. The method for locating abnormal transformer loss according to claim 1, characterized in that: The step of performing root cause analysis on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer and obtaining the root cause analysis result also includes: Obtain multiple groups of measured values ​​of additional loss influencing parameters of the transformer and additional loss measured values ​​corresponding to each group of the measured values; The preset fitting model is trained based on the multiple groups of measured values ​​and the additional loss measured values ​​to obtain an additional loss fitting model of the transformer; wherein the preset fitting model includes an extremely random tree model.

3. The method for locating abnormal transformer loss according to claim 1, characterized in that: The additional loss influencing parameters include high voltage resistance, low voltage resistance, short circuit impedance, low voltage coil outer diameter, low voltage coil inner diameter, high voltage coil outer diameter and high voltage coil inner diameter.

4. The method for locating abnormal transformer loss according to claim 1, characterized in that: The resistance loss calculation model is: ; in, is the high voltage rated phase current, is the high voltage resistor, is the low voltage rated phase current, It is a low voltage resistor.

5. The method for locating abnormal transformer loss according to claim 1, characterized in that: The step of performing root cause analysis on the transformer to be analyzed based on the transformer data and the load loss mechanism model of the transformer to obtain the root cause analysis result comprises: Based on the transformer data and the load loss mechanism model, an Adtributor algorithm is used to perform a root cause analysis on the transformer to be analyzed, and a main root cause of the load loss of the transformer to be analyzed is determined; wherein the main root cause includes a resistance loss root cause or an additional loss root cause; Performing root cause analysis based on the transformer data and the loss model corresponding to the main root cause, and determining the potential root cause of the load loss under the main root cause; The root cause analysis result is obtained based on the main root cause and the potential root cause.

6. The method for locating abnormal transformer loss according to claim 5, characterized in that: The measured data includes measured values ​​of resistance loss influencing parameters; the standard data includes design values ​​of resistance loss influencing parameters; the resistance loss influencing parameters include high voltage resistance and low voltage resistance; The step of performing root cause analysis based on the transformer data and the loss model corresponding to the main root cause to determine the potential root cause of the load loss under the main root cause includes: In the case where the main root cause is the resistance loss root cause, for each resistance loss influencing parameter, replacing the measured value of the corresponding resistance loss influencing parameter with the designed value; Based on the resistance loss calculation model, determining the resistance loss after parameter replacement; Determine the surprisingness and the degree of explanation of the resistance loss influencing parameter based on the measured value of the resistance loss influencing parameter, the designed value of the resistance loss influencing parameter, the resistance loss after the parameter replacement, and the resistance loss before the parameter replacement; Based on the surprise of all the resistance loss influencing parameters and the ranking order of the degree of explanation of all the resistance loss influencing parameters, the potential root cause of the load loss under the resistance loss root cause is determined.

7. The method for locating abnormal transformer loss according to claim 5, characterized in that: The measured data includes the measured value of the additional loss influencing parameter; the standard data includes the design value of the additional loss influencing parameter; The step of performing root cause analysis based on the transformer data and the loss model corresponding to the main root cause to determine the potential root cause of the load loss under the main root cause includes: In the case where the main root cause is an additional loss root cause, for each additional loss influencing parameter, replacing the corresponding measured value of the additional loss influencing parameter with the designed value; Based on the additional loss fitting model, determining the additional loss after parameter replacement; Determine the surprise and explanatory power of the additional loss influencing parameter based on the measured value of the additional loss influencing parameter, the designed value of the additional loss influencing parameter, the additional loss after the parameter replacement, and the additional loss before the parameter replacement; Based on the surprise of all the additional loss influencing parameters and the ranking order of the degree of explanation of all the additional loss influencing parameters, the potential root cause of the load loss under the additional loss root cause is determined.

8. A transformer abnormal loss positioning device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for locating abnormal loss of a transformer according to any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the transformer abnormal loss locating method according to any one of claims 1 to 7 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method for locating abnormal loss of a transformer according to any one of claims 1 to 7 are implemented.

Images