CN117992860A - Method and device for early warning and identifying defects of oil immersed power transformer - Google Patents
Method and device for early warning and identifying defects of oil immersed power transformer Download PDFInfo
- Publication number
- CN117992860A CN117992860A CN202410402329.4A CN202410402329A CN117992860A CN 117992860 A CN117992860 A CN 117992860A CN 202410402329 A CN202410402329 A CN 202410402329A CN 117992860 A CN117992860 A CN 117992860A
- Authority
- CN
- China
- Prior art keywords
- period
- time sequence
- key
- sequence variable
- index score
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 79
- 230000007547 defect Effects 0.000 title claims abstract description 58
- 239000000523 sample Substances 0.000 claims abstract description 101
- 238000009826 distribution Methods 0.000 claims abstract description 94
- 239000011159 matrix material Substances 0.000 claims abstract description 70
- 238000005259 measurement Methods 0.000 claims abstract description 38
- 239000013074 reference sample Substances 0.000 claims abstract description 33
- 238000005070 sampling Methods 0.000 claims description 80
- 238000012544 monitoring process Methods 0.000 claims description 78
- 239000007789 gas Substances 0.000 claims description 76
- 230000036541 health Effects 0.000 claims description 27
- 238000004587 chromatography analysis Methods 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 10
- 230000015654 memory Effects 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000002131 composite material Substances 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 11
- 238000012545 processing Methods 0.000 abstract description 11
- 238000010586 diagram Methods 0.000 description 9
- 230000002159 abnormal effect Effects 0.000 description 8
- 230000007704 transition Effects 0.000 description 8
- 238000003745 diagnosis Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 239000003550 marker Substances 0.000 description 3
- 239000000835 fiber Substances 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 239000004215 Carbon black (E152) Substances 0.000 description 1
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical class [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910002090 carbon oxide Inorganic materials 0.000 description 1
- 229920002678 cellulose Polymers 0.000 description 1
- 239000001913 cellulose Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005312 nonlinear dynamic Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Landscapes
- Housings And Mounting Of Transformers (AREA)
Abstract
The invention discloses a defect early warning and identifying method and device for an oil immersed power transformer, which relate to the field of data processing and comprise the following steps: determining key time sequence variables, constructing a reference sample matrix and an actual measurement sample matrix, fitting each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain reference distribution and disturbance distribution, calculating inconsistent index scores of each key time sequence variable according to the reference distribution and the disturbance distribution, and calculating global comprehensive inconsistent index scores of each time period according to the inconsistent index scores of each key time sequence variable; and determining whether to trigger an alarm according to the global comprehensive inconsistent index score in the continuous time period, and performing fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm in combination with a characteristic gas method to determine the defect type, thereby solving the problems that the running state of the transformer is difficult to identify, the fault type analysis is difficult to perform and the like.
Description
Technical Field
The invention relates to the field of data processing, in particular to a defect early warning and identifying method and device for an oil-immersed power transformer.
Background
Transformers are a critical component in power systems, and their normal operation is critical to the safety and stability of the entire power network. When the transformer fails internally, hydrocarbon gas, hydrogen gas, carbon oxides and the like are generated and are fused into the transformer oil. Analysis of the Dissolved Gas (DGA) in the oil is therefore effective to reveal the type of fault that may occur inside the transformer and its severity. The traditional transformer fault diagnosis method based on DGA mainly comprises a characteristic gas method and a three-ratio method. The characteristic gas method is used for comprehensively analyzing the content of dissolved gas in the oil and the change trend with time, and dividing the dissolved gas in the oil into main characteristic gas and secondary characteristic gas so as to judge the fault type. The method has the advantages that the method is simple and direct, the main characteristic gas can not be distinguished directly for quantitative operation, and fault diagnosis is easy to be influenced by experience. The three-ratio method can be used only when the content of each component of the gas or the gas growth rate exceeds the alarm threshold value, and the alarm threshold value is difficult to set; second, the three-ratio method is prone to bias in judging some initial defects and very low energy defects. Some students train a diagnosis model to perform fault diagnosis through a large number of sample data of transformers in different typical states based on machine learning, and compared with a traditional diagnosis method, accuracy and robustness are improved. However, in practical engineering application, the sample size of various typical fault states of the transformer is limited, and individual differences exist among different transformers, so that a universal fault sample is difficult to find for identifying the transition of the running state of the transformer.
Critical phase transition is a very important concept in nonlinear dynamics, and this theory has been gradually developed and applied in various fields in recent years. When a complex system is developed to critical conditions, small changes in control parameters or stresses within the system may cause critical phase changes, resulting in accidents. In the process of critical phase transition of a complex system, a "critical point" usually exists, and the system shows features such as frequent fluctuation and critical slowdown near the critical point, and these features can be regarded as "early warning signs" of the critical phase transition.
The concept of dynamic network markers (DYNAMICAL NETWORK MARKER, DNM) proposed by scholars on the basis of critical phase transition theory is used for describing the critical phase transition dynamic characteristics of a multivariable complex system, and the effectiveness of the proposed method is demonstrated in biological, ecological and financial systems. In the field of fault diagnosis, fang Ruiming et al first apply DNM to early defect early warning of transformers, and the method only needs to use monitoring data of the transformers to be diagnosed, does not need to rely on typical sample data of fault states, and achieves good effects in the aspect of transformer fault early warning. However, the method has high calculation cost, and depending on screening network key nodes, how to select a proper method to determine the key network becomes a difficult problem restricting the application of the method. The DNM method can obtain a group of variable subsets with highest contribution to system faults, but the lack of quantization indexes for measuring importance differences among various variables is unfavorable for subsequent fault type analysis.
Disclosure of Invention
The application aims to provide a defect early warning and identifying method and device for an oil-immersed power transformer aiming at the technical problems.
In a first aspect, the invention provides a defect early warning and identifying method for an oil immersed power transformer, which comprises the following steps:
taking the content of dissolved gas in oil with high correlation with the running state of an oil immersed power transformer as a key time sequence variable in the variables monitored by an oil chromatograph on-line monitoring system, acquiring actual monitoring values of the key time sequence variable at each sampling time in a current sampling period, and constructing an actual measurement sample data set based on the actual monitoring values;
Constructing a health state theoretical value prediction model aiming at each key time sequence variable, acquiring a historical monitoring value of each sampling time of each key time sequence variable in a previous sampling period, which is acquired by an oil chromatography online monitoring system when the oil immersed power transformer is in a health state, inputting the historical monitoring value into the health state theoretical value prediction model to obtain a predicted value of each sampling time of each key time sequence variable in a current sampling period, and constructing a predicted sample data set based on the predicted value;
constructing a reference sample matrix and an actual measurement sample matrix under each period based on a predicted sample data set and an actual measurement sample data set, fitting each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain reference distribution and disturbance distribution, calculating the inconsistent index score of each key time sequence variable by using a Jansen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and calculating the global comprehensive inconsistent index score of each period according to the inconsistent index score of each key time sequence variable;
And determining whether to trigger an alarm according to the global comprehensive inconsistent index score in the continuous time period, and performing fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
Preferably, the actual monitoring value and the historical monitoring value are both data after normalization processing.
Preferably, the state of health theoretical value prediction model comprises a trained long-term and short-term memory neural network.
Preferably, the construction of the reference sample matrix and the measured sample matrix at each period based on the predicted sample data set and the measured sample data set specifically includes:
dividing the predicted sample data set and the measured sample data set into N time periods, wherein each time period comprises N sampling points;
based on n sets of predicted data within the t-th period Constructing a reference sample matrix X shown in the following formula, wherein/>Predictive data composed of predictive values of n sampling points in a t-th period for key time sequence variables of m dimensions;
；
Wherein, Representing key timing variables/>Predicted value at j-th sample point,/>，M represents the total number of key timing variables;
Based on n measured data sets in the t-th period Constructing a measured sample matrix X' shown in the following formula, wherein/>Actual monitoring data formed by actual monitoring values of n sampling points of the key time sequence variable of m dimensions in the t-th period;
；
Wherein, Representing key timing variables/>Actual monitored value at the j-th sampling point.
Preferably, the fitting of the reference sample matrix and the actually measured sample matrix to each key time sequence variable is performed to obtain a reference distribution and a disturbance distribution, which specifically comprises:
According to critical timing variables in a matrix X of reference samples Calculates the mean/>, of n predictorsAnd standard deviation/>According to the mean/>And standard deviation/>Calculating key timing variables/>, in a reference sample matrix XPredicted value/>, at j-th sample pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a reference distribution P:
；
According to critical timing variables in the measured sample matrix X Calculated mean value of n actual monitored values of (2)And standard deviation/>According to the mean/>And standard deviation/>Calculating key time sequence variable/>, in measured sample matrix XActual monitoring value/>, at the j-th sampling pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a disturbance distribution Q:
。
Preferably, the inconsistent indicator score of each key time sequence variable is calculated by using a jensen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and the global comprehensive inconsistent indicator score of each period is calculated according to the inconsistent indicator score of each key time sequence variable, which specifically comprises the following steps:
calculating the inconsistency index score of each key time sequence variable in the t period by adopting the following steps:
；
；
Wherein, Representing the i-th dimension row vector in the reference distribution,Representing the i-th dimension row vector in the disturbance distribution,Representing the probabilities of the ith row and jth column in the reference distribution,Representing the probability of the ith row and jth column in the disturbance distribution,Representing critical timing variables at time tIs used for the value of the KL divergence,Is shown inTime period critical timing variableIs a non-uniformity index score for (1);
Calculating a global comprehensive inconsistency index score for a period t using ：
。
Preferably, whether to trigger an alarm is determined according to a global comprehensive inconsistent index score in a continuous period, fault identification is performed by utilizing inconsistent index scores of each key time sequence variable in the period of triggering the alarm in combination with a characteristic gas method, and defect types are determined, which specifically comprises:
constructing the corresponding relation between the defect type and the main characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tIf the difference value of the (2) is larger than the threshold value, generating an alarm signal, sequencing the inconsistent index scores of the key time sequence variables in the t+1 time period from large to small, and enabling the inconsistent index score to be larger than the global comprehensive inconsistent index score/>, of the t+1 time periodRegarding key time sequence variables of (2) as main characteristic gases, and scoring the inconsistency index to be smaller than the global comprehensive inconsistency index score/> of the t+1 periodThe key time sequence variable of the (1) is regarded as secondary characteristic gas, and the defect type corresponding to the oil immersed power transformer in the t+1 period is determined according to the corresponding relation between the defect type and the primary characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tAnd if the difference value of the current voltage is smaller than or equal to the threshold value, determining that the oil-immersed power transformer is in a healthy state, and repeating the steps by taking the t+1 period as the t period.
In a second aspect, the present invention provides an apparatus for early warning and identifying defects of an oil-immersed power transformer, including:
the actual measurement module is configured to take the content of dissolved gas in oil with high correlation with the running state of the oil immersed power transformer as a key time sequence variable in the variables monitored by the oil chromatography on-line monitoring system, obtain actual monitoring values of the key time sequence variable at each sampling time in the current sampling period, and construct an actual measurement sample data set based on the actual monitoring values;
The prediction module is configured to construct a health state theoretical value prediction model for each key time sequence variable, acquire historical monitoring values of sampling times of each key time sequence variable in a previous sampling period, acquired by the oil chromatography on-line monitoring system when the oil immersed power transformer is in a health state, input the historical monitoring values into the health state theoretical value prediction model, acquire predicted values of sampling times of each key time sequence variable in a current sampling period, and construct a predicted sample data set based on the predicted values;
The index score calculation module is configured to construct a reference sample matrix and an actual measurement sample matrix under each time period based on the prediction sample data set and the actual measurement sample data set, obtain reference distribution and disturbance distribution for each key time sequence variable by fitting based on the reference sample matrix and the actual measurement sample matrix, calculate inconsistent index scores of each key time sequence variable by using a Jasen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and calculate global comprehensive inconsistent index scores of each time period according to the inconsistent index scores of each key time sequence variable;
the monitoring and identifying module is configured to determine whether to trigger an alarm according to the overall comprehensive inconsistent index score in the continuous time period, and perform fault identification by utilizing the inconsistent index score of each key time sequence variable of the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
In a third aspect, the present invention provides an electronic device comprising one or more processors; and storage means for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the method as described in any of the implementations of the first aspect.
In a fourth aspect, the present invention provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method as described in any of the implementations of the first aspect.
Compared with the prior art, the invention has the following beneficial effects:
The defect early warning and identifying method for the oil immersed power transformer utilizes the real-time monitoring data of the oil chromatography on-line monitoring system to construct quantitative indexes describing the dynamic change of the transformer, so that the critical state of the transformer in the transition from the healthy state to the abnormal state is identified. The method does not need to rely on typical sample data, and the data is easy to acquire; the modeling process only needs to utilize the monitoring data of the transformer to be diagnosed, so that the generalization problem is avoided; collaborative analysis is performed for multiple variables in the transformer, not just for single variables; the method is based on the critical turning signal of the dynamic change detection system of the global variable, and does not need to rely on critical network screening compared with the traditional dynamic network marker method; the influence degree of different characteristic variables on system faults can be intuitively analyzed by constructing the inconsistency index of each characteristic gas, a quantifiable characteristic gas distinguishing means is provided for fault diagnosis by applying a characteristic gas method, and early defect early warning is facilitated for the transformer.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is an exemplary device frame pattern to which an embodiment of the present application may be applied;
fig. 2 is a flow chart of a fault early warning and identifying method for an oil immersed power transformer according to an embodiment of the application;
Fig. 3 is a diagram showing a score of an inconsistency index of a system of a transformer in a healthy state at each period in an oil immersed power transformer defect pre-warning and identifying method according to an embodiment of the present application A result graph;
Fig. 4 is a diagram showing the inconsistent index scores of the dissolved gas variables in each oil at each time period of the transformer in the healthy state in the defect pre-warning and identifying method of the oil-immersed power transformer according to the embodiment of the application Is a three-dimensional evolution result graph of (2);
fig. 5 is a diagram showing a score of an inconsistency index of a system of a transformer in an abnormal state at each period in an oil-immersed power transformer defect pre-warning and identifying method according to an embodiment of the present application A result graph;
FIG. 6 is a graph showing the inconsistent index scores of the dissolved gas variables in each oil at each time period of the transformer in an abnormal state in the defect pre-warning and identifying method of an oil-immersed power transformer according to an embodiment of the present application Is a three-dimensional evolution result graph of (2);
FIG. 7 is a graph showing the inconsistent index scores of the dissolved gas variables in each oil during abnormal operation period of the transformer in the defect pre-warning and identifying method of the oil-immersed power transformer according to the embodiment of the application A result graph;
FIG. 8 is a schematic diagram of a defect warning and identifying device for an oil immersed power transformer according to an embodiment of the present application;
fig. 9 is a schematic structural view of a computer device suitable for use in an electronic apparatus for implementing an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Fig. 1 illustrates an exemplary device architecture 100 to which the oil-immersed power transformer defect pre-warning and identification method or device of the present application may be applied.
As shown in fig. 1, the apparatus architecture 100 may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used as a medium to provide communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.
The user may interact with the server 105 via the network 104 using the first terminal device 101, the second terminal device 102, the third terminal device 103, to receive or send messages, etc. Various applications, such as a data processing class application, a file processing class application, and the like, may be installed on the terminal device one 101, the terminal device two 102, and the terminal device three 103.
The first terminal device 101, the second terminal device 102 and the third terminal device 103 may be hardware or software. When the first terminal device 101, the second terminal device 102, and the third terminal device 103 are hardware, they may be various electronic devices, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like. When the first terminal apparatus 101, the second terminal apparatus 102, and the third terminal apparatus 103 are software, they can be installed in the above-listed electronic apparatuses. Which may be implemented as multiple software or software modules (e.g., software or software modules for providing distributed services) or as a single software or software module. The present invention is not particularly limited herein.
The server 105 may be a server that provides various services, such as a background data processing server that processes files or data uploaded by the terminal device one 101, the terminal device two 102, and the terminal device three 103. The background data processing server can process the acquired file or data to generate a processing result.
It should be noted that, the defect early warning and identifying method for the oil-immersed power transformer provided by the embodiment of the application may be executed by the server 105, or may be executed by the first terminal device 101, the second terminal device 102, or the third terminal device 103, and accordingly, the defect early warning and identifying device for the oil-immersed power transformer may be set in the server 105, or may be set in the first terminal device 101, the second terminal device 102, or the third terminal device 103.
It should be understood that the number of terminal devices, networks and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation. In the case where the processed data does not need to be acquired from a remote location, the above-described apparatus architecture may not include a network, but only a server or terminal device.
Fig. 2 shows a defect early warning and identifying method for an oil immersed power transformer, which includes the following steps:
S1, taking the content of dissolved gas in oil with high correlation with the running state of an oil immersed power transformer as a key time sequence variable in variables monitored by an oil chromatography on-line monitoring system, acquiring actual monitoring values of the key time sequence variable at each sampling time in a current sampling period, and constructing an actual measurement sample data set based on the actual monitoring values.
Specifically, the variable monitored by the oil chromatography on-line monitoring system is utilized to select the content of dissolved gas in oil with high correlation with the running state of the oil immersed power transformer as a key time sequence variable, and an actual measurement sample data set is obtained based on the actual monitoring value of each key time sequence variable.
Specifically, taking a main transformer of a certain factory as an example, the embodiment of the application selects the content of dissolved gas in oil with high correlation with the running state of the transformer as a key time sequence variable based on real-time monitoring data of an oil chromatography on-line monitoring system, and the monitored dissolved gas in main oil is as follows: . The transformer is a 3X 277 MV.A, 515kV/22kV single-phase transformer group manufactured by Alston corporation, england, YNd1 is connected, the middle part of the high-low voltage side is connected, and the impedance voltage is 16.6%. All monitoring data sampling periods are 24h. The embodiment of the application is simulated by taking MATLAB as a working platform, oil chromatographic monitoring data of the main transformer from 10 months, 19 days, 2006 to 9 months, 18 days, 2008 are selected as sample data, and the sample data is preprocessed to obtain a sample data set. The transformer sends out first-level early warning in 3 and 2 days of 2008, adopts a non-coding ratio method for analysis, And (3) withThe ratio of (2) is less than 1, and the low-temperature overheat fault is diagnosed.
S2, constructing a health state theoretical value prediction model aiming at each key time sequence variable, acquiring a historical monitoring value of each sampling time of each key time sequence variable in a previous sampling period, which is acquired by an oil chromatography online monitoring system when the oil immersed power transformer is in a health state, inputting the historical monitoring value into the health state theoretical value prediction model, acquiring a predicted value of each sampling time of each key time sequence variable in a current sampling period, and constructing a predicted sample data set based on the predicted value.
In a specific embodiment, the actual monitoring value and the historical monitoring value are both normalized data.
In a specific embodiment, the state of health theoretical value prediction model comprises a trained long-term and short-term memory neural network.
Specifically, the measured sample data set and the predicted sample data set include normalized data obtained by preprocessing data of a plurality of key time sequence variables, the preprocessing mode is to select the data of the key time sequence variables in a period as sample data, the sample data includes an actual monitoring value and a historical monitoring value, the sample data is normalized, and all the sample data are normalized to be in a range from 0 to 1 according to an equal proportion relation, so that the normalized data can be obtained.
Further, the long-period memory neural network is trained by utilizing the historical monitoring value collected by the oil-immersed power transformer oil chromatography on-line monitoring system under the healthy state, and the trained long-period memory neural network is obtained, namely the healthy state theoretical value prediction model. The long-short-term memory neural network has regression prediction capability, a single-output health state theoretical value prediction model is respectively established for each key time sequence variable, input data of the health state theoretical value prediction model is a historical monitoring value of the key time sequence variable before a current sampling period (namely, in a previous sampling period), the input data is output as a predicted value of each sampling point of the key time sequence variable before the current sampling period, and a predicted sample data set is constructed based on the predicted value of each sampling point of each key time sequence variable before the current sampling period.
S3, constructing a reference sample matrix and an actual measurement sample matrix under each period based on the predicted sample data set and the actual measurement sample data set, fitting each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain reference distribution and disturbance distribution, calculating the inconsistency index score of each key time sequence variable according to the reference distribution and the disturbance distribution by utilizing a Jansen shannon divergence algorithm, and calculating the global comprehensive inconsistency index score of each period according to the inconsistency index score of each key time sequence variable.
In a specific embodiment, constructing the reference sample matrix and the measured sample matrix at each period based on the predicted sample data set and the measured sample data set specifically includes:
dividing the predicted sample data set and the measured sample data set into N time periods, wherein each time period comprises N sampling points;
based on n sets of predicted data within the t-th period Constructing a reference sample matrix X shown in the following formula, wherein/>Predictive data composed of predictive values of n sampling points in a t-th period for key time sequence variables of m dimensions;
；
Wherein, Representing key timing variables/>Predicted value at j-th sample point,/>，M represents the total number of key timing variables;
Based on n measured data sets in the t-th period Constructing a measured sample matrix X' shown in the following formula, wherein/>Actual monitoring data formed by actual monitoring values of n sampling points of the key time sequence variable of m dimensions in the t-th period;
；
Wherein, Representing key timing variables/>Actual monitored value at the j-th sampling point.
In a specific embodiment, fitting to obtain a reference distribution and a disturbance distribution for each key timing variable based on a reference sample matrix and a measured sample matrix specifically includes:
According to critical timing variables in a matrix X of reference samples Calculates the mean/>, of n predictorsAnd standard deviation/>According to the mean/>And standard deviation/>Calculating key timing variables/>, in a reference sample matrix XPredicted value/>, at j-th sample pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a reference distribution P:
；
According to critical timing variables in the measured sample matrix X Calculated mean value of n actual monitored values of (2)And standard deviation/>According to the mean/>And standard deviation/>Calculating key time sequence variable/>, in measured sample matrix XActual monitoring value/>, at the j-th sampling pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a disturbance distribution Q:
。
In a specific embodiment, calculating an inconsistency index score of each key time sequence variable by using a jensen shannon divergence algorithm according to a reference distribution and a disturbance distribution, and calculating a global comprehensive inconsistency index score of each period according to the inconsistency index score of each key time sequence variable, wherein the method specifically comprises the following steps:
calculating the inconsistency index score of each key time sequence variable in the t period by adopting the following steps:
；
；
Wherein, Representing the i-th dimension row vector in the reference distribution,Representing the i-th dimension row vector in the disturbance distribution,Representing the probabilities of the ith row and jth column in the reference distribution,Representing the probability of the ith row and jth column in the disturbance distribution,Representing critical timing variables at time tIs used for the value of the KL divergence,Is shown inTime period critical timing variableIs a non-uniformity index score for (1);
Calculating a global comprehensive inconsistency index score for a period t using ：
。
Specifically, based on the predicted sample data set and the measured sample data set, time windows are divided, and a reference sample matrix and a measured sample matrix under each period are constructed. According to the probability of each sample in the reference sample matrix and the actually measured sample matrix under Gaussian distribution; fitting the reference sample matrix and the measured sample matrix to a reference distribution and a disturbance distribution, respectively. Wherein the critical timing variables in the matrix of reference samplesMean/>And standard deviation/>The calculation formulas of (a) are respectively as follows:
；
；
actual measurement of critical timing variables in a sample matrix Mean/>And standard deviation/>The calculation formulas of (a) are respectively as follows:
；
。/>
Calculating inconsistent index scores and overall comprehensive inconsistent index scores of each key time sequence variable by utilizing a Jansen shannon divergence algorithm according to the reference distribution and the disturbance distribution, comparing the reference distribution P and the disturbance distribution Q of each period, and calculating Inconsistent indicator score for each key time sequence variable of time periodI.e. the jensen shannon divergence between two probability distributions, to measure the dynamic difference of variable data between reference sample and measured sample, and to score by using the inconsistency index of each variableCalculating global comprehensive inconsistency index score of t-period system，The range of the values is as follows. And taking the global comprehensive inconsistent index score as an index for triggering alarm, and comprehensively considering the inconsistent index score of each key time sequence variable and the global comprehensive inconsistent index score to determine the defect type.
S4, determining whether to trigger an alarm according to the overall comprehensive inconsistent index score in the continuous time period, and performing fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
In a specific embodiment, step S4 specifically includes:
constructing the corresponding relation between the defect type and the main characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tIf the difference value of the (2) is larger than the threshold value, generating an alarm signal, sequencing the inconsistent index scores of the key time sequence variables in the t+1 time period from large to small, and enabling the inconsistent index score to be larger than the global comprehensive inconsistent index score/>, of the t+1 time periodRegarding key time sequence variables of (2) as main characteristic gases, and scoring the inconsistency index to be smaller than the global comprehensive inconsistency index score/> of the t+1 periodThe key time sequence variable of the (1) is regarded as secondary characteristic gas, and the defect type corresponding to the oil immersed power transformer in the t+1 period is determined according to the corresponding relation between the defect type and the primary characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tAnd if the difference value of the current voltage is smaller than or equal to the threshold value, determining that the oil-immersed power transformer is in a healthy state, and repeating the steps by taking the t+1 period as the t period.
Specifically, repeating the steps for the period t+1, and calculating the global comprehensive inconsistent index score of the period t+1. When the increase of the global comprehensive inconsistent index score between the continuous time periods is smaller than or equal to the threshold value, repeating the steps S2-S4 by taking the t+1 time period as the t time period.
Specifically, a threshold value gamma is set for the global comprehensive inconsistency index score in the continuous period, whenTriggering an alarm; if not, the health state is judged. Using critical timing variables within an abnormal period of time triggering an alarmDefect identification by score-combined feature gas method, wherein key time sequence variableScore greater thanThe dissolved gas variable in the oil of (a) forms a dynamic network marker, is regarded as the main characteristic gas of the defect, and is used for controlling the key time sequence variableScore less than or equal toThe dissolved gas variables in the oil of (a) are regarded as secondary characteristic gases, and the correspondence relationship between the defect types and the characteristic variables is shown in table 1.
TABLE 1
If the transformer is in a healthy state, returning to the step S2, repeating the operations from the steps S2 to S4 with the t+1 period as the t period.
The above steps S1-S4 do not necessarily represent the order between steps, but the step symbols indicate that the order between steps is adjustable.
The technical scheme of the embodiment of the application is further described by specific examples.
Example 1: embodiment 1 of the present application corresponds to a normal situation, i.e. the transformer is in a healthy state. Based on data of 100 sampling time points of real-time monitoring of dissolved gas in each oil of an oil chromatography on-line monitoring system of the transformer in a healthy state at 21 st of 05 month in 2007 to at 08 nd 28 th in 2007, setting each 10 continuous sampling points as a period, and calculating the inconsistency index score of the dissolved gas in the oil in each period for 10 periodsGlobal comprehensive inconsistent index score/>. As can be seen from FIGS. 3 and 4, during this run time, the similarity of the reference profile to the disturbance profile is high due to the small variation of dissolved gas in each oil over different time periods,/>The whole is in a lower interval range. /(I)Although there is fluctuation, the fluctuation is kept within a certain range,/>The whole body is kept stable, and no abrupt and large rise occurs. />, dissolved gas in oilAlthough the fluctuation exists, the whole is kept stable and is in a lower interval range, which indicates that no abnormal condition exists and accords with maintenance records.
Example 2: example 2 of the present application corresponds to a low temperature overheat fault analysis. Based on the data of 100 sampling time points of real-time monitoring of dissolved gas in each oil of the transformer from 10 months in 2007 to 10 days in 2008 to 01 months in 17 days, each 10 continuous sampling points are set as a period, 10 periods are taken as a total, and the inconsistency index score of the dissolved gas in the oil in each period is calculatedGlobal comprehensive inconsistent index score/>. From FIGS. 5 and 6, it can be seen that during this run time, each period/>Is increased by the overall fluctuation range of/>Overall too high, setting the threshold γ to 0.03, then at time period 5,/>A sharp rise is occurring and the rise is,Indicating that the system reaches near the critical point of state transition, meaning that the operating state of the transformer has changed critically during period 5, from steady state/>Critical state/>Conversion to defective State/>Therefore, the period 5 is an abnormal period, and an early warning signal is sent out in the period 5.
Further analysis of faults, the dissolved gas in each oil in the period 5 stateThe sorting is performed in order from large to small as shown in fig. 7. When the transformer is in the critical state of period 5,/>/>, Of three characteristic gasesThe values are all greater than the peak value 0.073 of the global comprehensive inconsistency index score for time period 5. As can be seen in conjunction with FIG. 6, the/>, of the three feature gases during the time period around time period 5The values are higher than other characteristic gases in the same time period, and the three characteristic gases are proved to be three characteristic variables with the largest fluctuation in the critical state of the system, so the three characteristic variables form dynamic network markers in the critical state of the system, are a variable subset with the highest contribution degree to the fault, are main characteristic gases in the fault process, and/>/>Although the value is lower than/>But is much higher than/>Thus constituting a secondary characteristic gas during the fault. By combining with a characteristic gas method for analysis, when oil paper of the transformer is overheated, a large amount of CO gas is generated by overheat of the solid insulating material, when the overheat part reaches a certain temperature, cellulose is gradually carbonized and the oil temperature of the overheat part is increased, so that/>The isogas increases greatly. It can be inferred that the fault occurring at this time in the system is a oilpaper overheat fault. Comparing maintenance records, the transformer sends out first-level early warning in 3/2/2008, and adopts a coding-free ratio method for analysis,/>And/>The ratio of (2) is less than 1, and the low-temperature overheat fault is judged to be less than 300 ℃. Compared with the traditional threshold method, the method for early warning and identifying the defects of the oil immersed power transformer can effectively early warn and identify the early defects of the transformer in time.
The case analysis result shows that when the transformer is changed from the normal operation state to the critical state, the systemA change occurs. When the transformer is in a normal operating state,/>The method is low and has small variation amplitude and is in a stable state; when the transformer reaches a critical state before a catastrophic failure with the development of internal potential defects, the system/>A sharp rise to a peak occurs. Analysis of each characteristic gas in abnormal period by combining characteristic gas methodThe identification of the transformer defects can be realized. The defect early warning and identifying method for the oil immersed power transformer can send out early warning signals before faults occur, and identify fault types, so that the method has a certain engineering practical value.
With further reference to fig. 8, as an implementation of the method shown in the foregoing drawings, the present application provides an embodiment of an apparatus for pre-warning and identifying defects of an oil-immersed power transformer, where the embodiment of the apparatus corresponds to the embodiment of the method shown in fig. 2, and the apparatus may be specifically applied to various electronic devices.
The embodiment of the application provides an oil immersed power transformer defect early warning and identifying device, which comprises the following components:
The actual measurement module 1 is configured to take the content of dissolved gas in oil with high correlation with the running state of the oil immersed power transformer as a key time sequence variable in the variables monitored by the oil chromatograph on-line monitoring system, acquire actual monitoring values of the key time sequence variable at each sampling time in the current sampling period, and construct an actual measurement sample data set based on the actual monitoring values;
The prediction module 2 is configured to construct a health state theoretical value prediction model for each key time sequence variable, acquire historical monitoring values of sampling times of each key time sequence variable in a previous sampling period, acquired by the oil chromatography on-line monitoring system when the oil immersed power transformer is in a health state, input the historical monitoring values into the health state theoretical value prediction model, acquire predicted values of sampling times of each key time sequence variable in a current sampling period, and construct a predicted sample data set based on the predicted values;
An index score calculation module 3, configured to construct a reference sample matrix and an actual measurement sample matrix under each period based on the predicted sample data set and the actual measurement sample data set, fit each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain a reference distribution and a disturbance distribution, calculate the inconsistent index score of each key time sequence variable according to the reference distribution and the disturbance distribution by using the jensen shannon divergence algorithm, and calculate the global comprehensive inconsistent index score of each period according to the inconsistent index score of each key time sequence variable;
The monitoring and identifying module 4 is configured to determine whether to trigger an alarm according to the overall comprehensive inconsistent index score in the continuous time period, and perform fault identification by utilizing the inconsistent index score of each key time sequence variable of the time period triggering the alarm and combining a characteristic gas method to determine the defect type.
Referring now to fig. 9, there is illustrated a schematic diagram of a computer apparatus 900 suitable for use in an electronic device (e.g., a server or terminal device as illustrated in fig. 1) for implementing an embodiment of the present application. The electronic device shown in fig. 9 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present application.
As shown in fig. 9, the computer apparatus 900 includes a Central Processing Unit (CPU) 901 and a Graphics Processor (GPU) 902, which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 903 or a program loaded from a storage section 909 into a Random Access Memory (RAM) 904. In the RAM 904, various programs and data required for the operation of the computer device 900 are also stored. The CPU 901, GPU902, ROM 903, and RAM 904 are connected to each other by a bus 905. An input/output (I/O) interface 906 is also connected to bus 905.
The following components are connected to the I/O interface 906: an input section 907 including a keyboard, a mouse, and the like; an output portion 908 including a speaker, such as a Liquid Crystal Display (LCD), or the like; a storage section 909 including a hard disk or the like; and a communication section 910 including a network interface card such as a LAN card, a modem, or the like. The communication section 910 performs communication processing via a network such as the internet. The drive 911 may also be connected to the I/O interface 906 as needed. A removable medium 912 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed as needed on the drive 911 so that a computer program read out therefrom is installed into the storage section 909 as needed.
In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flowcharts. In such embodiments, the computer program may be downloaded and installed from a network via the communication portion 910, and/or installed from the removable medium 912. The above-described functions defined in the method of the present application are performed when the computer program is executed by a Central Processing Unit (CPU) 901 and a Graphics Processor (GPU) 902.
It should be noted that the computer readable medium according to the present application may be a computer readable signal medium or a computer readable medium, or any combination of the two. The computer readable medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor apparatus, device, or means, or a combination of any of the foregoing. More specific examples of the computer-readable medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution apparatus, device, or apparatus. In the present application, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may be any computer readable medium that is not a computer readable medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution apparatus, device, or apparatus. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based devices which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules involved in the embodiments of the present application may be implemented in software or in hardware. The described modules may also be provided in a processor.
As another aspect, the present application also provides a computer-readable medium that may be contained in the electronic device described in the above embodiment; or may exist alone without being incorporated into the electronic device. The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: taking the content of dissolved gas in oil with high correlation with the running state of an oil immersed power transformer as a key time sequence variable in the variables monitored by an oil chromatograph on-line monitoring system, acquiring actual monitoring values of the key time sequence variable at each sampling time in a current sampling period, and constructing an actual measurement sample data set based on the actual monitoring values; constructing a health state theoretical value prediction model aiming at each key time sequence variable, acquiring a historical monitoring value of each sampling time of each key time sequence variable in a previous sampling period, which is acquired by an oil chromatography online monitoring system when the oil immersed power transformer is in a health state, inputting the historical monitoring value into the health state theoretical value prediction model to obtain a predicted value of each sampling time of each key time sequence variable in a current sampling period, and constructing a predicted sample data set based on the predicted value; constructing a reference sample matrix and an actual measurement sample matrix under each period based on a predicted sample data set and an actual measurement sample data set, fitting each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain reference distribution and disturbance distribution, calculating the inconsistent index score of each key time sequence variable by using a Jansen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and calculating the global comprehensive inconsistent index score of each period according to the inconsistent index score of each key time sequence variable; and determining whether to trigger an alarm according to the global comprehensive inconsistent index score in the continuous time period, and performing fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
The above description is only illustrative of the preferred embodiments of the present application and of the principles of the technology employed. It will be appreciated by persons skilled in the art that the scope of the application referred to in the present application is not limited to the specific combinations of the technical features described above, but also covers other technical features formed by any combination of the technical features described above or their equivalents without departing from the inventive concept described above. Such as the above-mentioned features and the technical features disclosed in the present application (but not limited to) having similar functions are replaced with each other.
Claims (10)
1. The defect early warning and identifying method for the oil immersed power transformer is characterized by comprising the following steps of:
Taking the content of dissolved gas in oil with high correlation with the running state of an oil immersed power transformer as a key time sequence variable in the variables monitored by an oil chromatograph on-line monitoring system, acquiring actual monitoring values of the key time sequence variable at each sampling time in a current sampling period, and constructing an actual measurement sample data set based on the actual monitoring values;
Constructing a health state theoretical value prediction model aiming at each key time sequence variable, acquiring a historical monitoring value of each sampling time of each key time sequence variable in a previous sampling period, which is acquired by an oil chromatography online monitoring system when the oil immersed power transformer is in a health state, inputting the historical monitoring value into the health state theoretical value prediction model to obtain a predicted value of each sampling time of each key time sequence variable in a current sampling period, and constructing a predicted sample data set based on the predicted value;
Constructing a reference sample matrix and an actual measurement sample matrix under each time period based on the predicted sample data set and the actual measurement sample data set, fitting each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain reference distribution and disturbance distribution, calculating the inconsistency index score of each key time sequence variable by using a Jasen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and calculating the global comprehensive inconsistency index score of each time period according to the inconsistency index score of each key time sequence variable;
And determining whether to trigger an alarm according to the global comprehensive inconsistent index score in the continuous time period, and performing fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
2. The method for early warning and identifying a defect of an oil immersed power transformer according to claim 1, wherein the actual monitoring value and the historical monitoring value are normalized data.
3. The method for pre-warning and identifying defects of an oil-immersed power transformer according to claim 1, wherein the theoretical state of health value prediction model comprises a trained long-short-term memory neural network.
4. The method for early warning and identifying a defect of an oil immersed power transformer according to claim 1, wherein the constructing a reference sample matrix and an actual measurement sample matrix under each period based on the predicted sample data set and the actual measurement sample data set specifically comprises:
dividing the predicted sample data set and the measured sample data set into N time periods, wherein each time period comprises N sampling points;
based on n sets of predicted data within the t-th period Constructing a reference sample matrix X shown in the following formula, wherein/>Predictive data composed of predictive values of n sampling points in a t-th period for key time sequence variables of m dimensions;
；
Wherein, Representing key timing variables/>Predicted value at j-th sample point,/>，/>M represents the total number of key timing variables;
Based on n measured data sets in the t-th period Constructing a measured sample matrix X' shown in the following formula, wherein/>Actual monitoring data formed by actual monitoring values of n sampling points of the key time sequence variable of m dimensions in the t-th period;
；
Wherein, Representing key timing variables/>Actual monitored value at the j-th sampling point.
5. The method for early warning and identifying a defect of an oil immersed power transformer according to claim 4, wherein the fitting of the reference sample matrix and the actually measured sample matrix to each of the key time sequence variables to obtain a reference distribution and a disturbance distribution specifically comprises:
According to the key time sequence variable in the reference sample matrix X Calculates the mean/>, of n predictorsAnd standard deviation/>According to the mean/>And standard deviation/>Calculating a key timing variable/>, in the reference sample matrix XPredicted value/>, at j-th sample pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a reference distribution P:
；
according to the key time sequence variable in the actual measurement sample matrix X Calculated mean value of n actual monitored values of (2)And standard deviation/>According to the mean/>And standard deviation/>Calculating a key timing variable/>, in the measured sample matrix XActual monitoring value/>, at the j-th sampling pointCumulative probability distribution Density/>The following formula is shown:
；
will accumulate probability distribution density Conversion to key timing variable/>Probability/>, under gaussian distribution, of predicted value at jth sampling pointThe following formula is shown:
；
Probability of predicted value under Gaussian distribution at jth sampling point based on each key timing variable Constructing a disturbance distribution Q:
。
6. The method for early warning and identifying the defects of the oil-immersed power transformer according to claim 1, wherein the calculating the inconsistency index score of each key time sequence variable by using a jensen shannon divergence algorithm according to the reference distribution and the disturbance distribution, and calculating the global comprehensive inconsistency index score of each period according to the inconsistency index score of each key time sequence variable, specifically comprises:
Calculating the inconsistency index score of each key time sequence variable in the t period by adopting the following steps:
；
；
Wherein, Representing the i-th dimension row vector in the reference distribution,/>Representing the i-th dimension row vector in the disturbance distribution,Representing the probability of the ith row, jth column in the reference distribution,/>Representing the probability of the ith row, jth column in the disturbance distribution,/>Represents the critical timing variable/>, at time tKL divergence value of/>Expressed at/>Time period critical timing variable/>Is a non-uniformity index score for (1);
Calculating a global comprehensive inconsistency index score for a period t using ：
。
7. The method for early warning and identifying the defects of the oil-immersed power transformer according to claim 1, wherein the determining whether to trigger an alarm according to the global comprehensive inconsistency index score in the continuous time period, and performing fault identification by using the inconsistency index score of each key time sequence variable in the time period of triggering the alarm in combination with a characteristic gas method, specifically comprises:
constructing the corresponding relation between the defect type and the main characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tIf the difference value of the key time sequence variable is larger than the threshold value, generating an alarm signal, sequencing the inconsistent index scores of the key time sequence variables in the t+1 time period from large to small, and enabling the inconsistent index score to be larger than the global comprehensive inconsistent index score/>, of the t+1 time periodRegarding key time sequence variables of (2) as main characteristic gases, and scoring the inconsistency index to be smaller than the global comprehensive inconsistency index score/> of the t+1 periodThe key time sequence variable of the oil immersed power transformer is regarded as secondary characteristic gas, and the defect type corresponding to the oil immersed power transformer in the t+1 period is determined according to the corresponding relation between the defect type and the primary characteristic gas and the secondary characteristic gas;
Global integrated inconsistency indicator score in response to determining a t+1 period Global composite inconsistency index score/>, with period tAnd if the difference value of the current voltage of the current transformer is smaller than or equal to the threshold value, determining that the oil-immersed power transformer is in a healthy state, and repeating the steps by taking the t+1 period as the t period.
8. The utility model provides an oily formula power transformer defect early warning and identification device which characterized in that includes:
The actual measurement module is configured to take the content of dissolved gas in oil with high correlation with the running state of the oil immersed power transformer as a key time sequence variable in the variables monitored by the oil chromatograph on-line monitoring system, obtain actual monitoring values of the key time sequence variable at each sampling time in the current sampling period, and construct an actual measurement sample data set based on the actual monitoring values;
The prediction module is configured to construct a health state theoretical value prediction model for each key time sequence variable, acquire historical monitoring values of each sampling time of each key time sequence variable in a previous sampling period, which are acquired by the oil chromatography online monitoring system when the oil immersed power transformer is in a health state, input the historical monitoring values into the health state theoretical value prediction model, acquire predicted values of each sampling time of each key time sequence variable in a current sampling period, and construct a prediction sample data set based on the predicted values;
The index score calculation module is configured to construct a reference sample matrix and an actual measurement sample matrix under each time period based on the prediction sample data set and the actual measurement sample data set, fit each key time sequence variable based on the reference sample matrix and the actual measurement sample matrix to obtain a reference distribution and a disturbance distribution, calculate the inconsistent index score of each key time sequence variable according to the reference distribution and the disturbance distribution by using a jensen shannon divergence algorithm, and calculate the global comprehensive inconsistent index score of each time period according to the inconsistent index score of each key time sequence variable;
The monitoring and identifying module is configured to determine whether to trigger an alarm according to the overall comprehensive inconsistent index score in the continuous time period, and perform fault identification by utilizing the inconsistent index score of each key time sequence variable in the time period of triggering the alarm and combining a characteristic gas method to determine the defect type.
9. An electronic device, comprising:
One or more processors;
Storage means for storing one or more programs,
When executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-7.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410402329.4A CN117992860B (en) | 2024-04-03 | 2024-04-03 | Method and device for early warning and identifying defects of oil immersed power transformer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410402329.4A CN117992860B (en) | 2024-04-03 | 2024-04-03 | Method and device for early warning and identifying defects of oil immersed power transformer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117992860A true CN117992860A (en) | 2024-05-07 |
| CN117992860B CN117992860B (en) | 2024-06-25 |
Family
ID=90902426
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410402329.4A Active CN117992860B (en) | 2024-04-03 | 2024-04-03 | Method and device for early warning and identifying defects of oil immersed power transformer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117992860B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109490685A (en) * | 2018-10-30 | 2019-03-19 | 华侨大学 | A kind of transformer early defect method for early warning based on oil dissolved gas on-line monitoring |
| CN109632975A (en) * | 2018-12-22 | 2019-04-16 | 武汉新运维光电科技股份有限公司 | A kind of analysis of transformer oil chromatographic aggregation of data and failure prediction system and method |
| CN110188309A (en) * | 2019-05-31 | 2019-08-30 | 华侨大学 | Oil-immersed power transformer defect method for early warning based on Hidden Markov Model |
| CN111579731A (en) * | 2020-06-28 | 2020-08-25 | 华侨大学 | Early transformer defect early warning method based on combined model |
| CN115792681A (en) * | 2022-03-01 | 2023-03-14 | 浙江凌骁能源科技有限公司 | Single battery consistency detection algorithm based on Internet of vehicles big data platform |
| CN116432524A (en) * | 2023-03-23 | 2023-07-14 | 国网浙江省电力有限公司瑞安市供电公司 | Transformer oil temperature prediction method, device, equipment and storage medium |
-
2024
- 2024-04-03 CN CN202410402329.4A patent/CN117992860B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109490685A (en) * | 2018-10-30 | 2019-03-19 | 华侨大学 | A kind of transformer early defect method for early warning based on oil dissolved gas on-line monitoring |
| CN109632975A (en) * | 2018-12-22 | 2019-04-16 | 武汉新运维光电科技股份有限公司 | A kind of analysis of transformer oil chromatographic aggregation of data and failure prediction system and method |
| CN110188309A (en) * | 2019-05-31 | 2019-08-30 | 华侨大学 | Oil-immersed power transformer defect method for early warning based on Hidden Markov Model |
| CN111579731A (en) * | 2020-06-28 | 2020-08-25 | 华侨大学 | Early transformer defect early warning method based on combined model |
| CN115792681A (en) * | 2022-03-01 | 2023-03-14 | 浙江凌骁能源科技有限公司 | Single battery consistency detection algorithm based on Internet of vehicles big data platform |
| CN116432524A (en) * | 2023-03-23 | 2023-07-14 | 国网浙江省电力有限公司瑞安市供电公司 | Transformer oil temperature prediction method, device, equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 张燕;方瑞明;: "基于油中溶解气体动态网络标志物模型的变压器缺陷预警与辨识", 电工技术学报, no. 09, 10 May 2020 (2020-05-10) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117992860B (en) | 2024-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108375715B (en) | Power distribution network line fault risk day prediction method and system | |
| CN111080072A (en) | Distribution transformer health index evaluation method, device and system | |
| Gadhavi et al. | Student final grade prediction based on linear regression | |
| CN114863651A (en) | Intelligent early warning method for monitoring state of auxiliary machine | |
| CN117490948A (en) | Cable-stayed bridge health monitoring method and system based on dynamic threshold | |
| CN116720324A (en) | Traction substation key equipment fault early warning method and system based on prediction model | |
| Tang et al. | Prediction of bearing performance degradation with bottleneck feature based on LSTM network | |
| CN110134079B (en) | Process parameter early warning method and system based on slope analysis | |
| CN113112123B (en) | Method for diagnosing and evaluating faults of aircraft avionics system based on incidence matrix | |
| Zhu et al. | Temperature prediction of aluminum reduction cell based on integration of dual attention LSTM for non-stationary sub-sequence and ARMA for stationary sub-sequences | |
| CN117992860B (en) | Method and device for early warning and identifying defects of oil immersed power transformer | |
| Zhang et al. | Research on modeling in operator mental workload based on VACP method | |
| CN105741184A (en) | Transformer state evaluation method and apparatus | |
| CN111367255A (en) | Performance evaluation test system and method for multi-variable control system | |
| Shojaee et al. | Applying SVSSI sampling scheme to np-chart to decrease the time of detecting shifts using Markov chain approach and Monte Carlo simulation | |
| CN116736214A (en) | CVT stability evaluation method and system | |
| CN110083131A (en) | Technological parameter on-line early warning method and readable storage medium storing program for executing based on amplitude of variation | |
| JP2019191880A (en) | Equipment management support system | |
| Sarma et al. | A long short-term memory based prediction model for transformer fault diagnosis using dissolved gas analysis with digital twin technology | |
| Hu et al. | A multi-feature-based fault diagnosis method based on the weighted timeliness broad learning system | |
| CN114254904A (en) | Method and device for evaluating operation health degree of engine room of wind turbine generator | |
| Zhao et al. | Intelligent anti-corrosion expert system based on big data analysis | |
| CN113724795A (en) | Method and device for calculating concentration of raffinate hydrogen peroxide, storage medium and processor | |
| Miao et al. | Extracting procedures of key data from a structural maintenance database | |
| da Silva Filho et al. | Paraconsistent artificial neural network for structuring statistical process control in electrical engineering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |