CN117290679A - Running state detection method and device of current transformer and electronic equipment - Google Patents

Running state detection method and device of current transformer and electronic equipment Download PDF

Info

Publication number
CN117290679A
CN117290679A CN202311190142.4A CN202311190142A CN117290679A CN 117290679 A CN117290679 A CN 117290679A CN 202311190142 A CN202311190142 A CN 202311190142A CN 117290679 A CN117290679 A CN 117290679A
Authority
CN
China
Prior art keywords
current transformer
output data
data
imf
sample entropy
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.)
Pending
Application number
CN202311190142.4A
Other languages
Chinese (zh)
Inventor
韩程鹏
何飒
胡玉春
黎彦林
王毅博
赵赟
乔建云
姜飞
张鑫
王新
叶魁
徐廷云
牛翔
王天平
方江
张新
刘磊
程熙晔
李晨晨
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Xinjiang Electric Power Co Ltd Changji Power Supply Co
Original Assignee
State Grid Xinjiang Electric Power Co Ltd Changji Power Supply Co
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by State Grid Xinjiang Electric Power Co Ltd Changji Power Supply Co filed Critical State Grid Xinjiang Electric Power Co Ltd Changji Power Supply Co
Priority to CN202311190142.4A priority Critical patent/CN117290679A/en
Publication of CN117290679A publication Critical patent/CN117290679A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
    • G01R35/02—Testing or calibrating of apparatus covered by the other groups of this subclass of auxiliary devices, e.g. of instrument transformers according to prescribed transformation ratio, phase angle, or wattage rating
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Abstract

The invention provides a method and a device for detecting the running state of a current transformer and electronic equipment, wherein the method comprises the following steps: acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data; performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component; determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components; inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the detection model for detection to obtain an operation state detection result of the current transformer. The invention can realize the purpose of accurately evaluating the operation difference state of the transformer under the actual working condition.

Description

Running state detection method and device of current transformer and electronic equipment
Technical Field
The invention relates to the technical field of electric power metering monitoring, in particular to a method and a device for detecting the running state of a current transformer and electronic equipment.
Background
The current transformer is used as a key device in a distribution network line and is applied to the line in a large scale, and the main working principle is to measure a current signal in the line through electromagnetic induction. Due to the intelligent power grid and the rapid development of ultra-high voltage power transmission, the working environment of the current transformer in the distribution network is increasingly bad, and the factors influencing the stable operation of the current transformer are increased, so that a certain hidden danger is brought to the safe and stable operation of the power grid. The realization of the quick, accurate and efficient on-line identification technology of the running state of the current transformer in the distribution network is a key of the economic and intelligent construction of the power system.
In order to improve the detection precision of the running state of a single current transformer, a sigmoid function is utilized to perform relation fitting on the current transformer value and the current effective value on the bus, and a function center point and an uncertain domain are selected as feature vectors to perform fault diagnosis. However, the single mathematical statistics method is difficult to process the output data of the transformer with complex distribution characteristics, so that the existing method cannot accurately evaluate the operation difference state of the transformer under the actual working condition.
Disclosure of Invention
In view of the foregoing, it is necessary to provide a method and an apparatus for detecting an operating state of a current transformer, and an electronic device, so as to achieve the purpose of accurately evaluating the operating state of the transformer under actual working conditions.
In order to achieve the above object, the present invention provides a method for detecting an operation state of a current transformer, including:
acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component;
determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
Further, the obtaining the output data of the current transformer of the transformer substation includes:
acquiring output data of a current transformer of a transformer substation acquired at equal time intervals; the output data includes normal state data and abnormal state data.
Further, the preprocessing the output data of the current transformer to obtain denoising data includes:
determining standard deviation of the output data of the current transformer based on time series data corresponding to the output data of the current transformer;
and screening abnormal data in the output data of the current transformer based on the standard deviation to obtain denoising data.
Further, the performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components includes:
constructing each IMF component corresponding to the denoising data, the center frequency of each IMF component and a corresponding problem expression of the denoising data;
based on a preset secondary penalty factor and a Lagrange factor, and combining the variation problem expression, constructing a Lagrange expansion expression;
and carrying out iterative computation on the Lagrangian expansion expression based on a multiplication operator alternating algorithm to obtain a plurality of IMF components.
Further, the determining the weight coefficient of each IMF component based on the influence of the sample entropy of each IMF component on the running state of the current transformer includes:
k-means clustering is carried out on a feature vector matrix formed by sample entropy of each IMF component, and a clustered contour coefficient is obtained;
and determining the weight coefficient corresponding to each IMF component based on the clustered contour coefficient and the clustered contour coefficient corresponding to each IMF component removed from the plurality of IMF components.
Further, the clustering contour coefficients corresponding to the IMF components after being removed are calculated based on the following formula:
wherein w is k Clusters representing the kth group of sample clusters, u k True tags representing samples of the k-th group, P k Representing a clustering contour coefficient corresponding to a kth group of samples, wherein N represents the total number of IMF components; the kth group of samples are samples corresponding to the kth IMF component removed from the plurality of IMF components;
the weight coefficient corresponding to each IMF component is calculated based on the following formula:
wherein c k The weight coefficient corresponding to the kth IMF component.
Further, the eigenvector matrix is determined by calculating the following formula:
wherein alpha represents a margin index, beta represents a waveform index,represents root mean square value, F IMFN Representing sample entropy corresponding to the Nth IMF component;
representing the eigenvector matrix, a 1 And a 2 Respectively representing different eigenvectors in the eigenvector matrix.
The invention also provides an operation state detection device of the current transformer, which comprises:
the preprocessing module is used for acquiring output data of a current transformer of the transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
the decomposition module is used for carrying out empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating sample entropy of each IMF component;
the determining module is used for determining weight coefficients of the IMF components based on the influence of sample entropy of the IMF components on the running state of the current transformer and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
the detection module is used for inputting the characteristic vector matrix into a preset BP neural network model, inputting the characteristic vector matrix corresponding to the output data for training into the preset BP neural network model for training to obtain a current transformer state detection model, inputting the characteristic vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
The invention also provides an electronic device comprising a memory and a processor, wherein,
the memory is used for storing programs;
the processor is coupled to the memory, and is configured to execute the program stored in the memory, so as to implement the steps in the method for detecting an operation state of a current transformer according to any one of the foregoing claims.
The invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method of detecting the operational state of a current transformer as described in any of the above.
The beneficial effects of the implementation mode are that: according to the running state detection method, the running state detection device and the electronic equipment of the current transformer, through processing, namely preprocessing, the abnormal value of the output data of the current transformer of the transformer substation, denoising data is obtained, the weight coefficient of each IMF component is determined based on the influence degree of the sample entropy of each IMF component on the running state of the current transformer, the characteristic distinction degree of each state feature is more obvious in a unique characteristic weight coefficient selection mode, a feature vector matrix is constructed as the input of a BP neural network model, the running state detection result (state type) of the current transformer is taken as the output, the detection of the on-line running state of the current transformer is realized, and the running difference state of the transformer is accurately estimated under the actual working condition.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following description will briefly explain the drawings needed in the description of the embodiments, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of an embodiment of a method for detecting an operation state of a current transformer according to the present invention;
FIG. 2 is a schematic diagram of a BP neural network model provided by the invention;
FIG. 3 is a flowchart of another embodiment of a method for detecting an operation state of a current transformer according to the present invention;
FIG. 4 is a schematic diagram of the identification effect of BP neural network without feature weight selection provided by the invention;
FIG. 5 is a schematic diagram of the identification effect of the BP neural network after feature weight selection provided by the invention;
FIG. 6 is a block diagram of an embodiment of a current transformer operating state detection device provided by the present invention;
fig. 7 is a schematic structural diagram of an embodiment of an electronic device according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. 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.
In the description of the embodiments of the present application, unless otherwise indicated, the meaning of "a plurality" is two or more.
The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or device that comprises a list of steps or modules is not necessarily limited to those steps or modules that are expressly listed or inherent to such process, method, article, or device.
The naming or numbering of the steps in the embodiments of the present invention does not mean that the steps in the method flow must be executed according to the time/logic sequence indicated by the naming or numbering, and the named or numbered flow steps may change the execution order according to the technical purpose to be achieved, so long as the same or similar technical effects can be achieved.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
The invention provides a method and a device for detecting the running state of a current transformer and electronic equipment, and the method and the device are respectively described below.
As shown in fig. 1, the present invention provides a method for detecting an operation state of a current transformer, including:
step 110, obtaining output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
step 120, performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component;
130, determining a weight coefficient of each IMF component based on the sample entropy of each IMF component (Intrinsic Mode Functions, connotation mode component), and determining a feature vector matrix based on the sample entropy of each IMF component and the weight coefficient;
step 140, inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
It will be appreciated that, as shown in fig. 2, the BP (Back Propagation) neural network model has a decisive influence on the weight distribution and output result of the model due to the number and properties of the input variables of the model, so that selecting an appropriate input variable is a key step for detecting the state of the current transformer. According to the method, the output signal of the current transformer is subjected to feature construction according to a feature matrix construction method, the calculated feature matrix is used as network input, and the corresponding current transformer is used as output, wherein the running state of the current transformer is normal, fixed deviation, drift deviation, transformation ratio deviation and precision distortion:
y= [ fixed variation ratio variation drift deviation of normal accuracy distortion ]
In some embodiments, the obtaining output data of the current transformer of the substation includes:
acquiring output data of a current transformer of a transformer substation acquired at equal time intervals; the output data includes normal state data and abnormal state data.
It can be understood that the real output data of the current transformer of the transformer substation is collected as a test sample, and the test sample is divided into an abnormal state and a normal state according to the state of the transformer, wherein the abnormal state is further divided into precision distortion, fixed error, transformation ratio error and drift error. And the number of data samples in the normal state and the abnormal state is 200 respectively, the data length N is 1000, the data of each fault type is fully covered, and the acquisition time interval is one hour for acquiring the output data of the current transformer. The normal state data refers to data in which the current transformer is in a normal state, and the abnormal state data refers to data in which the current transformer is in an abnormal state.
In some embodiments, the preprocessing the output data of the current transformer to obtain denoising data includes:
determining standard deviation of the output data of the current transformer based on time series data corresponding to the output data of the current transformer;
and screening abnormal data in the output data of the current transformer based on the standard deviation to obtain denoising data.
It can be understood that the data preprocessing is performed according to the collected time series data, abnormal points in the data are removed by utilizing the 3 sigma principle, the standard deviation sigma of the whole signal is calculated, whether each data point is within the + -3 sigma interval of the average value is counted, and the abnormal points are screened. And then, performing empirical mode decomposition on the processed data through a VMD (Variational Modal Decomposition) algorithm to obtain a plurality of IMF components, and calculating the sample entropy of each IMF as a characteristic quantity, so that the characteristic quantity lays a foundation for the subsequent state detection more obviously, the influence of signal noise is avoided, and the characteristics among the states of the current transformer are more highlighted.
In some embodiments, the performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components includes:
constructing each IMF component corresponding to the denoising data, the center frequency of each IMF component and a corresponding problem expression of the denoising data;
based on a preset secondary penalty factor and a Lagrange factor, and combining the variation problem expression, constructing a Lagrange expansion expression;
and carrying out iterative computation on the Lagrangian expansion expression based on a multiplication operator alternating algorithm to obtain a plurality of IMF components.
It can be understood that the invention adopts the VMD algorithm to perform empirical mode decomposition on the denoising data, the VMD algorithm decomposition is a nonlinear signal decomposition method for decomposing an original signal into a series of eigenmode functions around each center frequency by setting the center frequency of bandwidth limitation, and the algorithm is mainly divided into two parts, namely the construction of a variation problem and the solution of the variation problem.
Construction of the variation problem:
firstly, constructing a variation problem, and if each intrinsic mode component is the bandwidth in the central frequency range, treating the variation problem as solving k mode functions, and solving the minimum value of the sum of estimated bandwidths of the mode functions, wherein the sum of modes obtained by decomposition is an original signal f, and the specific steps are as follows:
calculating a single-side frequency spectrum of each modal function by using Hilbert transformation:
where δ (t) represents a pulse function, j represents an imaginary symbol, and t represents time.
Center frequency according to each modal componentEstimating, namely modulating the frequency spectrum of each modal component by taking a baseband as a standard;
bandwidth estimation is carried out on each modal component according to the square norm of the gradient of the demodulation signal, and the constraint problem is that:
wherein f represents the original signal, w k Represents the center frequency of each modal component, u k Represents the decomposed modal components, and δ (t) represents the pulse function.
Solving the variational problem:
the constraint problem between the secondary punishment factors and the Lagrange multiplicators is converted into the non-constraint problem, the secondary punishment factors can ensure the reconstruction accuracy under the condition that the Gaussian noise exists in the signals, the Lagrange multiplicators can ensure the strictness of constraint conditions, and the Lagrange expansion expression is as follows:
where a represents a penalty factor and λ represents a lagrangian factor.
Cross-over by multiplicationIterative computation is performed on the substitution algorithm (Alternate Direction Method of Multipliers, ADMM), forAnd carrying out iterative updating, and calculating saddle points of the extended Lagrangian expression. />The expression formula is:
the conversion to the frequency domain is done using a Parseval/Planchrel Fourier equidistant transform:
the variable conversion mode is expressed as a non-negative frequency interval integral form, and the formula is updated as
According to the same stepsIs an updated formula of (c)
Wherein u is k Representing the decomposed kth eigenmode sequence component; wk represents the center frequency of the kth eigenmode component; f represents the original signal.
In some embodiments, the determining the weight coefficient of each IMF component based on the influence of the sample entropy of each IMF component on the running state of the current transformer includes:
k-means clustering is carried out on a feature vector matrix formed by sample entropy of each IMF component, and a clustered contour coefficient is obtained;
and determining the weight coefficient corresponding to each IMF component based on the clustered contour coefficient and the clustered contour coefficient corresponding to each IMF component removed from the plurality of IMF components.
It can be understood that, according to the IMF components obtained by decomposing the secondary current VMD output by the current transformer, the sample entropy of each IMF component is calculated respectively, so as to obtain a relation curve between the sample entropy of each IMF component and each type of signal, wherein the horizontal axis corresponds to the IMF component serial number, and the vertical axis corresponds to the sample entropy of each IMF component. Firstly, carrying out K-means clustering calculation on feature vectors formed by sample entropy of each IMF component obtained through decomposition to obtain clustered profile coefficients beta, and then carrying out clustering profile coefficient calculation beta after removing each IMF component i And comparing the two values to distribute corresponding weight factors, and then carrying out aggregation processing on each IMF sample entropy after the weight coefficient processing to obtain the feature vector.
In some embodiments, the cluster contour coefficients corresponding to the IMF components are calculated based on the following formula:
wherein w is k Clusters representing the kth group of sample clusters, u k True tags representing samples of the k-th group, P k Representing a clustering contour coefficient corresponding to a kth group of samples, wherein N represents the total number of IMF components; the kth group of samples are samples corresponding to the kth IMF component removed from the plurality of IMF components;
the weight coefficient corresponding to each IMF component is calculated based on the following formula:
wherein c k The weight coefficient corresponding to the kth IMF component.
In some embodiments, the eigenvector matrix is determined by calculating the following formula:
wherein alpha represents a margin index, beta represents a waveform index,represents root mean square value, F IMFN Representing sample entropy corresponding to the Nth IMF component;
representing the eigenvector matrix, a 1 And a 2 Respectively representing different eigenvectors in the eigenvector matrix.
It can be understood that the corresponding weight coefficient is calculated for each IMF component obtained by decomposition according to the above steps, and then the multidimensional feature matrix is subjected to dimension reduction by using the integration method, so as to obtain a final two-dimensional feature matrix, namely a feature vector matrix.
In other embodiments, a flow chart of a method for detecting an operation state of a current transformer is shown in fig. 3, after a BP neural network identification effect without feature weight selection is shown in fig. 4, an output signal of the current transformer is subjected to feature extraction through a trained BP neural network model and is used as an input quantity to obtain the operation state (state is normal, fixed deviation, drift deviation, transformation ratio deviation and accuracy distortion) of the current transformer, and each group of data is listed in fig. 4 below; the identification effect of the BP neural network after the feature weight selection is shown in fig. 5. The method provided by the invention has higher accuracy in identifying the state of the current transformer, meanwhile, in the method provided by the invention, the difference among the state features is more obvious after the feature weights are selected, the state identification accuracy is obviously improved, the method has high calculation speed and high accuracy, and meanwhile, the influence of external interference signals on the detection accuracy can be reduced, so that the method meets the practical engineering application environment.
In summary, the method for detecting the operation state of the current transformer provided by the invention comprises the following steps: acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data; performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component; determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components; inputting a characteristic vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a characteristic vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer.
According to the invention, abnormal values of output data of the current transformer of the transformer substation are processed, namely preprocessed, denoising data is obtained, the weight coefficient of each IMF component is determined based on the influence degree of sample entropy of each IMF component on the running state of the current transformer, the characteristic distinction degree of each state is more obvious in a unique characteristic weight coefficient selection mode, a characteristic vector matrix is constructed as input of a BP neural network model, the running state detection result (state type) of the current transformer is taken as output, the on-line running state detection of the current transformer is realized, and the running difference state of the transformer is accurately estimated under the actual working condition.
Further, the method selects sample entropy as a quantization characteristic of the complexity of the time sequence, carries out modal decomposition on the output signal of the current transformer through a VMD algorithm, screens related components according to weight factors of the components, combines a common time-frequency domain index as a feature vector, carries out dimension reduction processing on the feature vector by using a weight coefficient sum method, and detects the state of the current transformer by using the dimension reduced data as the input of the BP neural network.
As shown in fig. 6, the present invention further provides an operation state detection device 600 of a current transformer, including:
the preprocessing module 610 is configured to obtain output data of a current transformer of a transformer substation, and perform preprocessing on the output data of the current transformer to obtain denoising data;
the decomposition module 620 is configured to perform empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculate a sample entropy of each IMF component;
the determining module 630 is configured to determine a weight coefficient of each IMF component based on an influence of sample entropy of each IMF component on an operation state of the current transformer, and determine a feature vector matrix based on the sample entropy and the weight coefficient of each IMF component;
the detection module 640 is configured to input a feature vector matrix corresponding to output data for training to a preset BP neural network model for training to obtain a current transformer state detection model, and input a feature vector matrix corresponding to output data to be detected to the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
The operation state detection device for the current transformer provided in the foregoing embodiment may implement the technical solution described in the operation state detection method embodiment for the current transformer, and the specific implementation principle of each module or unit may refer to the corresponding content in the operation state detection method embodiment for the current transformer, which is not described herein.
As shown in fig. 7, the present invention further provides an electronic device 700 accordingly. The electronic device 700 includes a processor 701, a memory 702, and a display 703. Fig. 7 shows only some of the components of the electronic device 700, but it should be understood that not all of the illustrated components are required to be implemented and that more or fewer components may be implemented instead.
The memory 702 may be an internal storage unit of the electronic device 700 in some embodiments, such as a hard disk or memory of the electronic device 700. The memory 702 may also be an external storage device of the electronic device 700 in other embodiments, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the electronic device 700.
Further, the memory 702 may also include both internal storage units and external storage devices of the electronic device 700. The memory 702 is used for storing application software and various types of data for installing the electronic device 700.
The processor 701 may in some embodiments be a central processing unit (Central Processing Unit, CPU), microprocessor or other data processing chip for executing program code or processing data stored in the memory 702, such as the current transformer operating state detection method of the present invention.
The display 703 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like in some embodiments. The display 703 is used for displaying information on the electronic device 700 and for displaying a visual user interface. The components 701-703 of the electronic device 700 communicate with each other over a system bus.
In some embodiments of the present invention, when the processor 701 executes the running state detection program of the current transformer in the memory 702, the following steps may be implemented:
acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component;
determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
It should be understood that: the processor 701 may perform other functions in addition to the above functions when executing the running state detection program of the current transformer in the memory 702, and in particular, reference may be made to the description of the corresponding method embodiments above.
Further, the type of the electronic device 700 is not particularly limited, and the electronic device 700 may be a portable electronic device such as a mobile phone, a tablet computer, a personal digital assistant (personal digitalassistant, PDA), a wearable device, a laptop (laptop), etc. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices that carry IOS, android, microsoft or other operating systems. The portable electronic device described above may also be other portable electronic devices, such as a laptop computer (laptop) or the like having a touch-sensitive surface, e.g. a touch panel. It should also be appreciated that in other embodiments of the invention, the electronic device 700 may not be a portable electronic device, but rather a desktop computer having a touch-sensitive surface (e.g., a touch panel).
In yet another aspect, the present invention further provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the method for detecting an operation state of a current transformer provided by the above methods, the method comprising:
acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component;
determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
Those skilled in the art will appreciate that all or part of the flow of the methods of the embodiments described above may be accomplished by way of a computer program that instructs associated hardware, and that the program may be stored in a computer readable storage medium. The computer readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory.
The method, the device and the electronic equipment for detecting the running state of the current transformer provided by the invention are described in detail, and specific examples are applied to the description of the principle and the implementation mode of the invention, and the description of the above examples is only used for helping to understand the method and the core idea of the invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in light of the ideas of the present invention, the present description should not be construed as limiting the present invention.

Claims (10)

1. A method for detecting an operating state of a current transformer, comprising:
acquiring output data of a current transformer of a transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
performing empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating to obtain sample entropy of each IMF component;
determining weight coefficients of the IMF components based on influence of sample entropy of the IMF components on the running state of the current transformer, and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
inputting a feature vector matrix corresponding to output data for training into a preset BP neural network model for training to obtain a current transformer state detection model, and inputting a feature vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
2. The method for detecting an operation state of a current transformer according to claim 1, wherein the obtaining output data of the current transformer of the substation comprises:
acquiring output data of a current transformer of a transformer substation acquired at equal time intervals; the output data includes normal state data and abnormal state data.
3. The method for detecting an operation state of a current transformer according to claim 1, wherein the preprocessing the output data of the current transformer to obtain denoising data comprises:
determining standard deviation of the output data of the current transformer based on time series data corresponding to the output data of the current transformer;
and screening abnormal data in the output data of the current transformer based on the standard deviation to obtain denoising data.
4. The method for detecting an operation state of a current transformer according to claim 1, wherein the performing empirical mode decomposition on the denoised data to obtain a plurality of IMF components includes:
constructing each IMF component corresponding to the denoising data, the center frequency of each IMF component and a corresponding problem expression of the denoising data;
based on a preset secondary penalty factor and a Lagrange factor, and combining the variation problem expression, constructing a Lagrange expansion expression;
and carrying out iterative computation on the Lagrangian expansion expression based on a multiplication operator alternating algorithm to obtain a plurality of IMF components.
5. The method for detecting an operation state of a current transformer according to claim 1, wherein the determining the weight coefficient of each IMF component based on the influence of the sample entropy of each IMF component on the operation state of the current transformer comprises:
k-means clustering is carried out on a feature vector matrix formed by sample entropy of each IMF component, and a clustered contour coefficient is obtained;
and determining the weight coefficient corresponding to each IMF component based on the clustered contour coefficient and the clustered contour coefficient corresponding to each IMF component removed from the plurality of IMF components.
6. The method for detecting an operation state of a current transformer according to claim 5, wherein the cluster contour coefficients corresponding to the IMF components are calculated based on the following formula:
wherein w is k Clusters representing the kth group of sample clusters, u k True tags representing samples of the k-th group, P k Representing a clustering contour coefficient corresponding to a kth group of samples, wherein N represents the total number of IMF components; the kth group of samples are samples corresponding to the kth IMF component removed from the plurality of IMF components;
the weight coefficient corresponding to each IMF component is calculated based on the following formula:
wherein c k The weight coefficient corresponding to the kth IMF component.
7. The method of claim 6, wherein the eigenvector matrix is determined by calculating the following formula:
wherein alpha represents a margin index, beta represents a waveform index,represents root mean square value, F IMFN Representing sample entropy corresponding to the Nth IMF component;
representing the eigenvector matrix, a 1 And a 2 Respectively representing different eigenvectors in the eigenvector matrix.
8. An operational state detection device of a current transformer, comprising:
the preprocessing module is used for acquiring output data of a current transformer of the transformer substation, and preprocessing the output data of the current transformer to obtain denoising data;
the decomposition module is used for carrying out empirical mode decomposition on the denoising data to obtain a plurality of IMF components, and calculating sample entropy of each IMF component;
the determining module is used for determining weight coefficients of the IMF components based on the influence of sample entropy of the IMF components on the running state of the current transformer and determining a feature vector matrix based on the sample entropy and the weight coefficients of the IMF components;
the detection module is used for inputting the characteristic vector matrix into a preset BP neural network model, inputting the characteristic vector matrix corresponding to the output data for training into the preset BP neural network model for training to obtain a current transformer state detection model, inputting the characteristic vector matrix corresponding to the output data to be detected into the current transformer state detection model for detection to obtain an operation state detection result of the current transformer;
the output data of the current transformer comprises the output data for training and the output data to be identified.
9. An electronic device comprising a memory and a processor, wherein,
the memory is used for storing programs;
the processor is coupled to the memory for executing the program stored in the memory to implement the steps in the method for detecting an operation state of a current transformer according to any one of claims 1 to 7.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the method of detecting the operational state of a current transformer according to any of claims 1 to 7.
CN202311190142.4A 2023-09-13 2023-09-13 Running state detection method and device of current transformer and electronic equipment Pending CN117290679A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311190142.4A CN117290679A (en) 2023-09-13 2023-09-13 Running state detection method and device of current transformer and electronic equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311190142.4A CN117290679A (en) 2023-09-13 2023-09-13 Running state detection method and device of current transformer and electronic equipment

Publications (1)

Publication Number Publication Date
CN117290679A true CN117290679A (en) 2023-12-26

Family

ID=89238194

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311190142.4A Pending CN117290679A (en) 2023-09-13 2023-09-13 Running state detection method and device of current transformer and electronic equipment

Country Status (1)

Country Link
CN (1) CN117290679A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117477495A (en) * 2023-12-28 2024-01-30 国网山西省电力公司太原供电公司 Current transformer state monitoring system and method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117477495A (en) * 2023-12-28 2024-01-30 国网山西省电力公司太原供电公司 Current transformer state monitoring system and method
CN117477495B (en) * 2023-12-28 2024-03-12 国网山西省电力公司太原供电公司 Current transformer state monitoring system and method

Similar Documents

Publication Publication Date Title
CN108318129B (en) Method for discriminating true and false of modal parameter of bridge structure and terminal equipment
CN108593260B (en) Optical cable line fault positioning and detecting method and terminal equipment
Zhang et al. Modulation classification method for frequency modulation signals based on the time–frequency distribution and CNN
Gao et al. Spare optimistic based on improved ADMM and the minimum entropy de-convolution for the early weak fault diagnosis of bearings in marine systems
CN106505557B (en) Remote measurement error identification method and device
CN103020459B (en) A kind of cognitive method of various dimensions electricity consumption behavior and system
CN109886464B (en) Low-information-loss short-term wind speed prediction method based on optimized singular value decomposition generated feature set
CN109975673B (en) Method for identifying fault arc at direct current side of photovoltaic microgrid
CN104966161B (en) A kind of power quality recorder data calculation and analysis methods based on gauss hybrid models
CN111080105A (en) Transformer area user-to-user relationship identification method and system based on voltage time sequence data
CN117290679A (en) Running state detection method and device of current transformer and electronic equipment
CN110990121B (en) Kubernetes scheduling strategy based on application portraits
CN107730097B (en) Bus load prediction method and device and computing equipment
CN108564569B (en) A kind of distress in concrete detection method and device based on multicore classification learning
CN112596016A (en) Transformer fault diagnosis method based on integration of multiple one-dimensional convolutional neural networks
CN110222098A (en) Electric power high amount of traffic abnormality detection based on flow data clustering algorithm
CN110188322A (en) A kind of wave-shape amplitude uncertainty determines method and system
CN116776181A (en) Terminal side load identification method, medium and system based on improved fuzzy clustering
CN114897032B (en) Current transformer fault diagnosis method and device based on width learning
CN117079005A (en) Optical cable fault monitoring method, system, device and readable storage medium
CN110991774A (en) Electric quantity load prediction method and device
CN116826735A (en) Broadband oscillation identification method and device for new energy station
CN108764301B (en) A kind of distress in concrete detection method based on reversed rarefaction representation
Gao et al. Feature extraction method of ball mill load based on adaptive variational mode decomposition and improved power spectrum analysis
Su et al. PSR-LSTM model for weak pulse signal detection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication