CN116894212A - High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm - Google Patents
High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm Download PDFInfo
- Publication number
- CN116894212A CN116894212A CN202310540863.7A CN202310540863A CN116894212A CN 116894212 A CN116894212 A CN 116894212A CN 202310540863 A CN202310540863 A CN 202310540863A CN 116894212 A CN116894212 A CN 116894212A
- Authority
- CN
- China
- Prior art keywords
- circuit breaker
- voltage circuit
- fault detection
- fault
- breaker fault
- 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
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 106
- 238000005457 optimization Methods 0.000 title claims description 13
- 241000282461 Canis lupus Species 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 25
- 238000012706 support-vector machine Methods 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000013145 classification model Methods 0.000 claims abstract description 13
- 238000012216 screening Methods 0.000 claims description 26
- 238000013480 data collection Methods 0.000 claims description 19
- 239000011159 matrix material Substances 0.000 claims description 18
- 238000012360 testing method Methods 0.000 claims description 11
- 241000282421 Canidae Species 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 9
- 230000006870 function Effects 0.000 description 12
- 238000000605 extraction Methods 0.000 description 6
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical group [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/327—Testing of circuit interrupters, switches or circuit-breakers
- G01R31/3271—Testing of circuit interrupters, switches or circuit-breakers of high voltage or medium voltage devices
- G01R31/3275—Fault detection or status indication
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The application discloses a high-voltage circuit breaker fault detection method based on an optimized SVM algorithm, which is characterized in that the current characteristics of a switching-on/off coil of the high-voltage circuit breaker are extracted at multiple angles, key characteristics are screened out by adopting a Laplace score method, the dimensionality of a fault characteristic set is reduced, and the fault judgment and diagnosis are facilitated; optimizing kernel function parameters and penalty factors of a support vector machine by using a gray wolf algorithm, and performing SVM modeling according to the optimized kernel function parameters and penalty factors to obtain a GWO-SVM classification model; and training GWO-SVM classification models by using an optimal fault feature set, constructing an efficient and accurate high-voltage circuit breaker fault detection model, realizing high-voltage circuit breaker fault detection by using the high-voltage circuit breaker fault detection model, improving the precision and efficiency of the high-voltage circuit breaker fault detection, and effectively ensuring the safe and stable operation of a power system.
Description
Technical Field
The application relates to the technical field of power equipment fault detection, in particular to a high-voltage circuit breaker fault detection method and system based on an optimized SVM algorithm.
Background
High voltage circuit breakers are critical and expensive equipment for an electrical power grid, and their proper operation is related to safe and reliable operation of the entire grid. As a result of investigation, it was found that: the ratio of faults of the high-voltage breaker operating mechanism and the control circuit is more than 45% of the total faults. Because the operating mechanism opening and closing coil current contains rich information, the operating state of the operating mechanism can be effectively represented, and the acquisition mode is non-invasive, the high-voltage circuit breaker operating mechanism opening and closing coil current waveform is widely used for evaluating the operating state of the circuit breaker operating mechanism.
In recent years, with the continuous development of theoretical techniques such as artificial intelligence and machine learning, machine learning methods such as artificial neural networks, K-nearest neighbor analysis and support vector machines are widely applied in the field of fault diagnosis of high-voltage circuit breakers, and good effects are obtained. The switching-on/off coil current reserves abundant breaker running state information, but the current coil current characteristic extraction method has the problems that the extracted characteristic quantity is redundant, the potential effective characteristic information of the coil current is not fully extracted, and the like. Under the influence of actual working conditions, running environments and the like, the characteristics are extracted from the aspects of coil current peak values, key time points and the like, and the method has certain limitations. Therefore, it is necessary to extract current characteristics of the opening and closing coil at multiple angles, and a characteristic selection method is used for screening a characteristic subset with low dimensionality and strong fault resolution to establish a fault diagnosis type of the high-voltage circuit breaker.
Disclosure of Invention
Aiming at the problems that the extracted characteristic quantity is redundant, the potential effective characteristic information of the coil current is not fully extracted and the like in the conventional coil current characteristic extraction method during fault diagnosis of the high-voltage circuit breaker, the application provides the high-voltage circuit breaker fault detection method and system based on an optimized SVM algorithm, so that the precision and efficiency of fault detection of the high-voltage circuit breaker are improved, and the safe and stable operation of a power system is effectively ensured; the current characteristics of the switching-on and switching-off coils of the high-voltage circuit breaker are extracted at multiple angles, key characteristics are screened out by adopting a Laplace score method, the dimensionality of a fault characteristic set is reduced, and fault discrimination and diagnosis are facilitated; and a support vector machine is optimized by using a gray wolf algorithm, so that the high-efficiency and accurate fault detection model of the high-voltage circuit breaker can be constructed.
In order to achieve the above object, the present application adopts the following technical scheme.
A high-voltage circuit breaker fault detection method based on an optimization SVM algorithm comprises the following steps:
step S1: extracting characteristics from the current of a switching-on/off coil of a historical high-voltage circuit breaker, and establishing an initial fault characteristic set;
step S2: screening key features from the initial fault feature set by using a Laplace score method to obtain an optimal fault feature set;
step S3: optimizing kernel function parameters and penalty factors of a support vector machine by using a gray wolf algorithm, and performing SVM modeling according to the optimized kernel function parameters and penalty factors to obtain a GWO-SVM classification model;
step S4: training GWO-SVM classification models by using the optimal fault feature set to obtain a high-voltage circuit breaker fault detection model;
step S5: collecting and processing the current of a switching-on/off coil of the real-time high-voltage circuit breaker, inputting processed real-time data (namely key characteristics corresponding to the current of the switching-on/off coil of the real-time high-voltage circuit breaker) into a high-voltage circuit breaker fault detection model, and outputting a high-voltage circuit breaker fault detection result by the high-voltage circuit breaker fault detection model;
the application provides a fault detection method of a high-voltage circuit breaker based on an optimization SVM algorithm, which is characterized in that the current characteristics of an opening and closing coil of the high-voltage circuit breaker are extracted at multiple angles, and the characteristics with low dimensionality and strong fault resolution are screened out by adopting a Laplacian score method to serve as an optimal fault characteristic set, so that irrelevant characteristics are removed, redundant characteristics are deleted, the dimensionality of the fault characteristic set is reduced, and the problems of redundancy of extracted characteristic quantity, insufficient extraction of potential effective characteristic information of coil current and the like in the current coil current characteristic extraction method are solved; the gray wolf algorithm is used for optimizing the kernel function parameters and the penalty factors of the support vector machine, so that the construction of a high-efficiency and accurate high-voltage circuit breaker fault detection model is facilitated, the precision and efficiency of high-voltage circuit breaker fault detection are improved, and the safe and stable operation of a power system is effectively ensured.
Preferably, in step S1, the features include a key time point, a current peak, and a relative time span of the core movement, a skewness Skew, a root mean square RMS, a crest factor CF, and a shape factor SF. According to the application, the current of the switching-on/off coil of the high-voltage circuit breaker is taken as a research object, and on the basis of a common current peak value and a key time point, the statistical characteristics and the duration time of different stages of the movement of the iron core are introduced to construct fault characteristics, so that the current characteristics of the switching-on/off coil of the high-voltage circuit breaker are fully extracted at multiple angles.
Preferably, the specific process of step S2 includes the following steps:
step S21: constructing a neighbor graph G by using given m data sample points; the neighbor graph G is a graph describing the relation among samples, and the ith node corresponds to x i The j-th node corresponds to x j ;x i And x j There are two cases of a bordered connection and a borderless connection, the bordered connection representing x i And x j The distance between them is short, and the borderless connection represents x i And x j The distance between the two is far;
step S22: if the data sample point x i And x j Is in close proximity toOtherwise S ij =0, where σ is a constant and S is a weighting matrix;
step S23: solving a Laplace matrix L=D-S corresponding to the matrix S through the matrix S, wherein D is a diagonal matrix obtained according to the matrix S;
step S24: computing Laplace score for each featureWherein f i Representing data sample points x i Taking the value on the characteristic f, wherein Var (f) is the estimated variance of the characteristic f; the lower the score, the better the feature f, and the partial geometric result information is reserved, which is beneficial to fault discrimination and diagnosis;
step S25: arranging Laplace scores of the features from small to large, and selecting a plurality of features arranged at the forefront to form an optimal fault feature set;
the Laplace score method is an unsupervised feature selection algorithm, and is based on Laplace feature value mapping and local preserving projection, and local geometric information of features is preserved.
Preferably, in step S3, a specific process of optimizing kernel function parameters and penalty factors of the support vector machine using a wolf algorithm includes the following steps:
step A1: initializing a gray wolf population N, randomly initializing the position of the gray wolves, wherein the maximum iteration number is T, initializing the iteration number T, enabling T to be 1, and randomly initializing a target fitness value;
step A2: the position of the predation position of the gray wolves is updated according to the Lewy flight;
step A3: calculating the fitness value of each gray wolf individual; if the fitness value of the new individual is better than that of the old individual, updating the new generation of individual, replacing the original position by the new individual, and updating the fitness value; otherwise, the old individuals are reserved, and the original fitness value is kept unchanged;
step A4: calculating and updating the gray wolf algorithm parameters according to the updated fitness value;
step A5: if the current iteration number is greater than the maximum iteration number, outputting a kernel function parameter g and a penalty factor C; otherwise, t=t+1, returning to step A2;
according to the application, the kernel function parameters and the penalty factors of the support vector machine are optimized by using the gray wolf algorithm, so that the construction of a high-efficiency and accurate fault detection model of the high-voltage circuit breaker is facilitated, the precision and efficiency of fault detection of the high-voltage circuit breaker are improved, and the safe and stable operation of the power system is effectively ensured.
Preferably, in step S4, the optimal fault feature set is divided into a training set and a test set, and the training set is used for training the GWO-SVM classification model; the test set is used for testing the high-voltage circuit breaker fault detection model.
The high-voltage circuit breaker fault detection system based on the optimization SVM algorithm comprises a data collection unit, wherein the input end of the data collection unit is connected to a high-voltage circuit breaker opening and closing coil, the output end of the data collection unit is connected with the input end of a fault feature screening unit, the output end of the fault feature screening unit is connected with the input end of a high-voltage circuit breaker fault detection model generating unit, the output end of the high-voltage circuit breaker fault detection model generating unit is connected with the input end of the high-voltage circuit breaker fault detection unit, and the input end of the high-voltage circuit breaker fault detection unit is also connected with the output end of the fault feature screening unit. The application provides a high-voltage circuit breaker fault detection system based on an optimization SVM algorithm, which comprises a data collection unit, a fault feature screening unit, a high-voltage circuit breaker fault detection model generation unit and a high-voltage circuit breaker fault detection unit which are sequentially connected, wherein the data collection unit is also connected with a high-voltage circuit breaker switching-on/off coil, and the high-voltage circuit breaker fault detection unit is also connected with the fault feature screening unit.
Preferably, the data collection unit is used for collecting the current of the opening and closing coil of the high-voltage circuit breaker and extracting the characteristics from the current, so as to establish an initial fault characteristic set. The data collection unit selects key time points, current peaks, relative time span of iron core movement, skewness Shew, root mean square RMS, crest factor CF, shape factor SF and other characteristic quantities from the current curve of the opening and closing coil, and establishes a 16-dimensional initial fault characteristic set.
Preferably, the fault feature screening unit screens key features from the initial fault feature set by using a Laplace score method to obtain an optimal fault feature set. And calculating the current characteristic scores (namely importance) of the switching-on and switching-off coils by using a Laplacian score method based on the established initial fault characteristic set, sequencing, and sequentially establishing a low-dimensional optimal fault characteristic set according to the characteristic importance.
Preferably, in the high-voltage circuit breaker fault detection model generation unit, a gray wolf algorithm is used for optimizing a support vector machine, and the optimized support vector machine is trained by using an optimal fault feature set to generate the high-voltage circuit breaker fault detection model. Taking the reserved important characteristics (key characteristics extracted from the historical opening and closing coil current) as input, and establishing a fault detection model by using a training set; optimizing the model by adopting a gray wolf algorithm to obtain optimal parameters C and g, thereby establishing an optimal classification model; the test set is then used to verify the modeled fault diagnosis performance.
Preferably, in the high-voltage circuit breaker fault detection unit, an optimal fault feature set corresponding to the real-time high-voltage circuit breaker opening and closing coil current output by the fault feature screening unit is used as input of a high-voltage circuit breaker fault detection model, and a high-voltage circuit breaker fault detection result is output. After the high-voltage circuit breaker fault detection model is output by the high-voltage circuit breaker fault detection model generating unit, the high-voltage circuit breaker fault detection unit inputs the obtained important characteristics of the current of the real-time opening and closing coil into the high-voltage circuit breaker fault detection model to obtain a fault detection result, and real-time, accurate and efficient high-voltage circuit breaker fault detection is realized.
Therefore, the application has the advantages that:
(1) The method comprises the steps of extracting current characteristics of a switching-on/off coil of a high-voltage circuit breaker at multiple angles, screening out key characteristics with low dimensionality and strong fault resolution by using a Laplace score method as an optimal fault characteristic set, removing irrelevant characteristics, deleting redundant characteristics, reducing the dimensionality of the fault characteristic set, and solving the problems of redundancy of extracted characteristic quantity, insufficient extraction of potential effective characteristic information of coil current and the like existing in the current coil current characteristic extraction method;
(2) The gray wolf algorithm is used for optimizing the kernel function parameters and the penalty factors of the support vector machine, so that the construction of a high-efficiency and accurate high-voltage circuit breaker fault detection model is facilitated, the precision and efficiency of high-voltage circuit breaker fault detection are improved, and the safe and stable operation of a power system is effectively ensured.
Drawings
Fig. 1 is a flowchart of a high-voltage circuit breaker fault detection method based on an optimized SVM algorithm in an embodiment of the present application.
Fig. 2 is a schematic structural diagram of a fault detection system of a high-voltage circuit breaker based on an optimized SVM algorithm in an embodiment of the present application.
1. The device comprises a data collection unit 2, a fault characteristic screening unit 3, a high-voltage circuit breaker fault detection model generation unit 4 and a high-voltage circuit breaker fault detection unit.
Detailed Description
The application is further described below with reference to the drawings and detailed description.
A high-voltage circuit breaker fault detection method based on an optimized SVM algorithm, as shown in FIG. 1, comprises the following steps:
step S1: extracting characteristics from the current of a switching-on/off coil of a historical high-voltage circuit breaker, and establishing an initial fault characteristic set;
step S2: screening key features from the initial fault feature set by using a Laplace score method to obtain an optimal fault feature set;
step S3: optimizing kernel function parameters and penalty factors of a support vector machine by using a gray wolf algorithm, and performing SVM modeling according to the optimized kernel function parameters and penalty factors to obtain a GWO-SVM classification model;
step S4: training GWO-SVM classification models by using the optimal fault feature set to obtain a high-voltage circuit breaker fault detection model;
step S5: collecting and processing the current of a switching-on/off coil of the real-time high-voltage circuit breaker, inputting the processed real-time data (namely key characteristics corresponding to the current of the switching-on/off coil of the real-time high-voltage circuit breaker) into a high-voltage circuit breaker fault detection model, and outputting a high-voltage circuit breaker fault detection result by the high-voltage circuit breaker fault detection model;
the embodiment provides a high-voltage circuit breaker fault detection method based on an optimization SVM algorithm, which comprises the steps of extracting current characteristics of a switching-on/off coil of the high-voltage circuit breaker at multiple angles, screening out characteristics with low dimensionality and strong fault resolution as an optimal fault characteristic set by adopting a Laplacian score method, removing irrelevant characteristics, and deleting redundant characteristics; and optimizing kernel function parameters and penalty factors of the support vector machine by using a gray wolf algorithm, and constructing an efficient and accurate high-voltage circuit breaker fault detection model.
In step S1, the features include the key time points, the current peaks, the relative time span of the core movement, the skewness Skew, the root mean square RMS, the crest factor CF, and the shape factor SF. In the embodiment, the current of the opening and closing coil of the high-voltage circuit breaker is taken as a research object, and the statistical characteristics and the duration of different stages of the movement of the iron core are introduced to construct fault characteristics on the basis of a common current peak value and a key time point.
The specific process of step S2 includes the following steps:
step S21: constructing a neighbor graph G by using given m data sample points; the neighbor graph G is a graph describing the relation among samples, and the ith node corresponds to x i The j-th node corresponds to x j ;x i And x j There are two cases of a bordered connection and a borderless connection, the bordered connection representing x i And x j The distance between them is short, and the borderless connection represents x i And x j The distance between the two is far;
step S22: if the data sample point x i And x j Is in close proximity toOtherwise S ij =0, where σ is a constant and S is a weighting matrix;
step S23: solving a Laplace matrix L=D-S corresponding to the matrix S through the matrix S, wherein D is a diagonal matrix obtained according to the matrix S;
step S24: computing Laplace score for each featureWherein f i Representing data sample points x i Taking the value on the characteristic f, wherein Var (f) is the estimated variance of the characteristic f; the lower the score, the better the feature f, and the partial geometric result information is reserved, which is beneficial to fault discrimination and diagnosis;
step S25: and arranging the Laplace scores of the features from small to large, and selecting a plurality of features arranged at the forefront to form an optimal fault feature set.
In step S3, a specific process of optimizing kernel function parameters and penalty factors of the support vector machine by using a wolf algorithm includes the following steps:
step A1: initializing a gray wolf population N, randomly initializing the position of the gray wolves, wherein the maximum iteration number is T, initializing the iteration number T, enabling T to be 1, and randomly initializing a target fitness value;
step A2: the position of the predation position of the gray wolves is updated according to the Lewy flight;
step A3: calculating the fitness value of each gray wolf individual; if the fitness value of the new individual is better than that of the old individual, updating the new generation of individual, replacing the original position by the new individual, and updating the fitness value; otherwise, the old individuals are reserved, and the original fitness value is kept unchanged;
step A4: calculating and updating the gray wolf algorithm parameters according to the updated fitness value;
step A5: if the current iteration number is greater than the maximum iteration number, outputting a kernel function parameter g and a penalty factor C; otherwise, t=t+1, returning to step A2.
In the step S4, the optimal fault characteristic set is divided into a training set and a testing set, wherein the training set is used for training a GWO-SVM classification model; the test set is used for testing a fault detection model of the high-voltage circuit breaker.
The high-voltage circuit breaker fault detection system based on the optimization SVM algorithm adopts the high-voltage circuit breaker fault detection method based on the optimization SVM algorithm, as shown in FIG. 2, the high-voltage circuit breaker fault detection system comprises a data collection unit 1, wherein the input end of the data collection unit 1 is connected to a high-voltage circuit breaker switching-on/switching-off coil, the output end of the data collection unit 1 is connected with the input end of a fault feature screening unit 2, the output end of the fault feature screening unit 2 is connected with the input end of a high-voltage circuit breaker fault detection model generating unit 3, the output end of the high-voltage circuit breaker fault detection model generating unit 3 is connected with the input end of a high-voltage circuit breaker fault detection unit 4, and the input end of the high-voltage circuit breaker fault detection unit 4 is also connected with the output end of the fault feature screening unit 2. The embodiment provides a high-voltage circuit breaker fault detection system based on an optimization SVM algorithm, which comprises a data collection unit 1, a fault feature screening unit 2, a high-voltage circuit breaker fault detection model generation unit 3 and a high-voltage circuit breaker fault detection unit 4 which are sequentially connected, wherein the data collection unit 1 is further connected with a high-voltage circuit breaker opening and closing coil, the high-voltage circuit breaker fault detection unit 4 is further connected with the fault feature screening unit 2, the high-voltage circuit breaker fault detection model is obtained based on key feature training of high-voltage circuit breaker opening and closing coil current, and real-time, accurate and efficient high-voltage circuit breaker fault detection is achieved by using the high-voltage circuit breaker fault detection model.
The data collection unit 1 is used for collecting the current of the opening and closing coil of the high-voltage circuit breaker and extracting the characteristics from the current, and an initial fault characteristic set is established. The data collection unit 1 selects key time points, current peaks, relative time span of iron core movement, skewness Shew, root mean square RMS, crest factor CF, shape factor SF and other characteristic quantities from the current curve of the opening and closing coil, and establishes a 16-dimensional initial fault characteristic set.
The fault feature screening unit 2 screens key features from the initial fault feature set by using a Laplace score method to obtain an optimal fault feature set. And calculating the current characteristic scores (namely importance) of the switching-on and switching-off coils by using a Laplacian score method based on the established initial fault characteristic set, sequencing, and sequentially establishing a low-dimensional optimal fault characteristic set according to the characteristic importance.
In the high-voltage circuit breaker fault detection model generation unit 3, a gray wolf algorithm is used for optimizing a support vector machine, and the optimized support vector machine is trained by using an optimal fault feature set to generate a high-voltage circuit breaker fault detection model. Taking the reserved important characteristics (key characteristics extracted from the historical opening and closing coil current) as input, and establishing a fault detection model by using a training set; optimizing the model by adopting a gray wolf algorithm to obtain optimal parameters C and g, thereby establishing an optimal classification model; the test set is then used to verify the modeled fault diagnosis performance.
In the high-voltage circuit breaker fault detection unit 4, an optimal fault feature set corresponding to the real-time high-voltage circuit breaker opening and closing coil current output by the fault feature screening unit 2 is used as input of a high-voltage circuit breaker fault detection model, and a high-voltage circuit breaker fault detection result is output. After the high-voltage circuit breaker fault detection model is output by the high-voltage circuit breaker fault detection model generating unit 3, the high-voltage circuit breaker fault detection unit 4 inputs the obtained important real-time opening and closing coil current characteristics into the high-voltage circuit breaker fault detection model to obtain a fault detection result.
The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the application is subject to the protection scope of the claims.
Claims (10)
1. The high-voltage circuit breaker fault detection method based on the optimization SVM algorithm is characterized by comprising the following steps of:
step S1: extracting characteristics from the current of a switching-on/off coil of a historical high-voltage circuit breaker, and establishing an initial fault characteristic set;
step S2: screening key features from the initial fault feature set by using a Laplace score method to obtain an optimal fault feature set;
step S3: optimizing kernel function parameters and penalty factors of a support vector machine by using a gray wolf algorithm, and performing SVM modeling according to the optimized kernel function parameters and penalty factors to obtain a GWO-SVM classification model;
step S4: training GWO-SVM classification models by using the optimal fault feature set to obtain a high-voltage circuit breaker fault detection model;
step S5: collecting and processing the current of a switching-on/off coil of the real-time high-voltage circuit breaker, inputting the processed real-time data into a high-voltage circuit breaker fault detection model, and outputting a high-voltage circuit breaker fault detection result by the high-voltage circuit breaker fault detection model.
2. The method according to claim 1, wherein in step S1, the features include a critical moment point, a current peak value, and a relative time span of the core movement, a skewness Skew, a root mean square RMS, a crest factor CF, and a shape factor SF.
3. The high-voltage circuit breaker fault detection method based on the optimized SVM algorithm according to claim 1, wherein the specific process of step S2 comprises the following steps:
step S21: constructing a neighbor graph G by using given m data sample points;
step S22: if the data sample point x i And x j Is in close proximity toOtherwise S ij =0, where σ is a constant and S is a weighting matrix;
step S23: solving a Laplace matrix L=D-S corresponding to the matrix S through the matrix S, wherein D is a diagonal matrix obtained according to the matrix S;
step S24: computing Laplace score for each featureWherein f i Representing data sample points x i Taking the value on the characteristic f, wherein Var (f) is the estimated variance of the characteristic f;
step S25: and arranging the Laplace scores of the features from small to large, and selecting a plurality of features arranged at the forefront to form an optimal fault feature set.
4. The method for detecting faults of a high voltage circuit breaker according to claim 1, 2 or 3, wherein in step S3, a specific process of optimizing kernel function parameters and penalty factors of a support vector machine by using a gray wolf algorithm comprises the following steps:
step A1: initializing a gray wolf population N, randomly initializing the position of the gray wolves, wherein the maximum iteration number is T, initializing the iteration number T, enabling T to be 1, and randomly initializing a target fitness value;
step A2: the position of the predation position of the gray wolves is updated according to the Lewy flight;
step A3: calculating the fitness value of each gray wolf individual; if the fitness value of the new individual is better than that of the old individual, updating the new generation of individual, replacing the original position by the new individual, and updating the fitness value; otherwise, the old individuals are reserved, and the original fitness value is kept unchanged;
step A4: calculating and updating the gray wolf algorithm parameters according to the updated fitness value;
step A5: if the current iteration number is greater than the maximum iteration number, outputting a kernel function parameter g and a penalty factor C; otherwise, t=t+1, returning to step A2.
5. A high voltage circuit breaker failure detection method based on optimized SVM algorithm according to claim 1 or 3, characterized in that in step S4, the optimal failure feature set is divided into a training set and a test set, the training set is used for training the GWO-SVM classification model; the test set is used for testing the high-voltage circuit breaker fault detection model.
6. A high-voltage circuit breaker fault detection system based on an optimized SVM algorithm, and a high-voltage circuit breaker fault detection method based on an optimized SVM algorithm according to any one of claims 1-5, characterized in that the system comprises a data collection unit, an input end of the data collection unit is connected to a high-voltage circuit breaker opening and closing coil, an output end of the data collection unit is connected to an input end of a fault feature screening unit, an output end of the fault feature screening unit is connected to an input end of a high-voltage circuit breaker fault detection model generating unit, an output end of the high-voltage circuit breaker fault detection model generating unit is connected to an input end of the high-voltage circuit breaker fault detection unit, and an input end of the high-voltage circuit breaker fault detection unit is also connected to an output end of the fault feature screening unit.
7. The optimized SVM algorithm based high voltage circuit breaker fault detection system of claim 6, wherein the data collection unit is configured to collect and extract characteristics of the high voltage circuit breaker opening and closing coil current and establish an initial fault characteristic set.
8. The high-voltage circuit breaker failure detection system based on the optimization SVM algorithm according to claim 7, wherein the failure feature screening unit adopts the Laplace score method to screen out key features from the initial failure feature set, and obtains the optimal failure feature set.
9. The high-voltage circuit breaker fault detection system based on the optimization SVM algorithm according to claim 8, wherein in the high-voltage circuit breaker fault detection model generation unit, a gray-wolf algorithm is used for optimizing a support vector machine, and the optimized support vector machine is trained by using an optimal fault feature set to generate a high-voltage circuit breaker fault detection model.
10. The high-voltage circuit breaker fault detection system based on the optimized SVM algorithm according to claim 8 or 9, wherein in the high-voltage circuit breaker fault detection unit, an optimal fault feature set corresponding to the real-time high-voltage circuit breaker opening and closing coil current output by the fault feature screening unit is used as input of a high-voltage circuit breaker fault detection model, and a high-voltage circuit breaker fault detection result is output.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310540863.7A CN116894212A (en) | 2023-05-15 | 2023-05-15 | High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310540863.7A CN116894212A (en) | 2023-05-15 | 2023-05-15 | High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116894212A true CN116894212A (en) | 2023-10-17 |
Family
ID=88311355
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310540863.7A Pending CN116894212A (en) | 2023-05-15 | 2023-05-15 | High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116894212A (en) |
-
2023
- 2023-05-15 CN CN202310540863.7A patent/CN116894212A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111157899B (en) | Method for estimating SOC of battery based on model fusion idea | |
CN105891707A (en) | Opening-closing fault diagnosis method for air circuit breaker based on vibration signals | |
CN110726898B (en) | Power distribution network fault type identification method | |
CN108680348A (en) | A kind of breaker mechanical fault diagnosis method and system based on random forest | |
Wang et al. | Few-shot transfer learning with attention mechanism for high-voltage circuit breaker fault diagnosis | |
CN116524200A (en) | High-voltage circuit breaker fault diagnosis method based on image recognition | |
CN105372528B (en) | A kind of state maintenance method of Power Transformer Internal Faults and New Transformer | |
CN109936113B (en) | Protection action intelligent diagnosis method and system based on random forest algorithm | |
CN109409444B (en) | Multivariate power grid fault type discrimination method based on prior probability | |
CN112288147B (en) | Method for predicting insulation state of generator stator by BP-Adaboost strong predictor | |
CN110646708B (en) | 10kV single-core cable early state identification method based on double-layer long-and-short-term memory network | |
CN110457758A (en) | Prediction technique, device, system and the storage medium in Instability of Rock Body stage | |
CN112364756A (en) | GIS fault classification method based on longicorn stigma and random forest | |
CN110609477A (en) | Electric power system transient stability discrimination system and method based on deep learning | |
CN114757110B (en) | Circuit breaker fault diagnosis method based on sliding window detection and current extraction signals | |
CN116401532B (en) | Method and system for recognizing frequency instability of power system after disturbance | |
CN115879048A (en) | Series arc fault identification method and system based on WRFMDA model | |
CN116418117A (en) | Equipment detection system for intelligent power grid | |
CN116894212A (en) | High-voltage circuit breaker fault detection method and system based on optimization SVM algorithm | |
CN111079647A (en) | Circuit breaker defect identification method | |
CN111506636A (en) | System and method for analyzing residential electricity consumption behavior based on autoregressive and neighbor algorithm | |
CN115113038B (en) | Circuit breaker fault diagnosis method based on current signal phase space reconstruction | |
CN113988205B (en) | Method and system for judging electric precipitation working condition | |
CN115795360A (en) | Cable fault detection method based on artificial neural network | |
CN109193579B (en) | Method and system for determining an operating overvoltage in a direct current network line |
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 |