CN116087692B - Distribution network tree line discharge fault identification method, system, terminal and medium - Google Patents
Distribution network tree line discharge fault identification method, system, terminal and medium Download PDFInfo
- Publication number
- CN116087692B CN116087692B CN202310384768.2A CN202310384768A CN116087692B CN 116087692 B CN116087692 B CN 116087692B CN 202310384768 A CN202310384768 A CN 202310384768A CN 116087692 B CN116087692 B CN 116087692B
- Authority
- CN
- China
- Prior art keywords
- line discharge
- tree line
- tree
- distribution network
- physical quantity
- 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.)
- Active
Links
Images
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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
-
- 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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
-
- 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
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
- G01R31/1263—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation
- G01R31/1272—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation of cable, line or wire insulation, e.g. using partial discharge measurements
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- 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
Abstract
The invention discloses a method, a system, a terminal and a medium for identifying tree discharge faults of a power distribution network, which relate to the technical field of power distribution networks and have the technical scheme that: establishing a basic characteristic sample library of tree line discharge faults of the low-current grounding system; randomly combining physical quantities in a basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination; clustering analysis is carried out on the physical quantity combinations through a K-means method, and the physical quantity combinations with fitness not less than a set threshold value are selected as optimized feature samples; training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model; and inputting the physical quantity data acquired from the power distribution network in real time into a tree line discharge fault identification model to carry out fault identification, so as to obtain a fault identification result. The invention can realize early identification of the tree line discharge fault in the complex interference environment.
Description
Technical Field
The invention relates to the technical field of power distribution networks, in particular to a power distribution network tree line discharge fault identification method, a system, a terminal and a medium.
Background
In recent years, higher requirements are put forward on the safety of the power distribution network in China, and in order to ensure the reliability and economy of power supply, the power distribution network line in China mostly adopts a small current grounding mode, and single-phase grounding faults are a main fault mode. The distance between the distribution network line passing through the forest zone and the tall trees is relatively short, and the tree line discharge can occur under the action of external forces such as lightning stroke, strong wind and the like, so that single-phase grounding faults are caused, and the distribution network line which is not grounded or is grounded through an arc suppression coil can continue to supply power for 1-2 hours under the condition of single-phase grounding due to relatively small fault current. However, a single-phase fault which is not handled in time may develop into an arc grounding fault, and the arc may ignite branches and leaves to induce a forest fire due to inflammables such as tree bamboos and the like around the fault point, so that monitoring and identifying the fault of the power distribution network are very necessary.
The existing single-phase earth fault identification method for the power distribution network mainly extracts and analyzes transient characteristic quantity and steady characteristic quantity of faults, and traditional power distribution automation equipment is used as an important carrier for quickly identifying and processing faults of the power distribution network, so that the principle and effect of fault classification are large in difference, and the accuracy rate cannot meet the working requirements of a power system. In recent years, the development of artificial intelligence technology is rapid, intelligent algorithms such as neural networks and cluster analysis are gradually applied to classification and identification of single-phase earth faults of a power distribution network, and research hotspots are focused on solving problems through combination of multiple methods. The genetic algorithm is superior to other intelligent algorithms in many aspects of convergence, robustness, efficiency and the like by virtue of excellent expansibility and potential parallelism; the BP neural network is widely applied in fault diagnosis due to strong linear mapping capability and self-learning self-adaptation capability.
However, the single-phase ground fault features are complex, various and mutually interwoven, and the overlapping repetition between different fault types can be caused by classifying by only the single-dimensional features, so that the multi-angle and multi-level classification method is gradually paid attention to and applied to the classification work of the fault ground tissues in the power distribution network, and the defects of unreasonable classification, weak generalization capability, insufficient interpretability and the like still exist. Therefore, how to study and design a distribution network tree line discharge fault identification method, system, terminal and medium capable of overcoming the defects is a problem which needs to be solved at present.
Disclosure of Invention
In order to solve the defects in the prior art, the invention aims to provide a method, a system, a terminal and a medium for identifying tree line discharge faults of a power distribution network, which adopt a genetic algorithm to screen a large amount of data, reduce the dimension of the processed data, then perform cluster analysis on the data by using an artificial neural network algorithm, and distinguish whether the operation data is in a fault state or a normal operation state, thereby realizing early identification of the tree line discharge faults under a complex interference environment.
The technical aim of the invention is realized by the following technical scheme:
in a first aspect, a method for identifying a tree discharge fault of a power distribution network is provided, which includes the following steps:
establishing a basic characteristic sample library of tree line discharge faults of the low-current grounding system;
randomly combining physical quantities in a basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination;
clustering analysis is carried out on the physical quantity combinations through a K-means method, and the physical quantity combinations with fitness not less than a set threshold value are selected as optimized feature samples;
training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model;
and inputting the physical quantity data acquired from the power distribution network in real time into a tree line discharge fault identification model to carry out fault identification, so as to obtain a fault identification result.
Further, the establishing process of the basic characteristic sample library specifically comprises the following steps:
acquiring high-dimensional tree line discharge fault feature quantity through a laboratory tree line discharge simulation experiment to obtain a first feature quantity;
analyzing the tree line discharge fault characteristic quantity through a simulation platform to obtain a second characteristic quantity;
and integrating the first characteristic quantity and the second characteristic quantity, and then establishing a basic characteristic sample library.
Further, the obtaining process of the second characteristic quantity specifically includes:
establishing a tree line discharge distribution model on a Comsol simulation platform, abstracting the tree line discharge into needle-plate discharge, wherein the tree is a needle electrode, and the line is abstracted into a plate electrode;
establishing distribution parameter simulation of tree line discharge through a Comsol simulation platform, and calculating a charge transfer and transport process in the tree line discharge process to obtain voltage and current waveforms generated by the tree line discharge;
and inputting the voltage and current waveforms into a lumped parameter power distribution network model in the Simulink platform to obtain a second characteristic quantity.
Further, the process of randomly combining physical quantities in the basic feature sample library by adopting the genetic algorithm specifically comprises the following steps:
randomly generating P groups of physical quantity combination vectors formed by initial tree line discharge fault characteristics as a first generation, intersecting Q with minimum adaptability, wherein the sub-vector elements are respectively from two male parent vectors, and Q is smaller than P;
let the crossover rate be eta c ，0＜η c < 1, randomly generating a number i between 0 and 1 before crossing, if i < eta c Performing cross operation; otherwise, selecting any one of the two male parent vectors as a next generation child vector;
let the mutation rate be eta m ，0＜η m < 1, randomly generating a number j between 0 and 1 for each element of the sub-vector when performing cross operation, if j < eta m Performing mutation operation;
and (3) ending one iteration, judging whether a physical quantity combination with the fitness not smaller than a set threshold exists in the sub-vector, if so, ending the iteration, and if not, continuing the iteration.
Further, the combination of physical quantities is represented by a plurality of sets of vectors containing only 0/1; wherein 0 indicates that the corresponding physical quantity is not considered when the tree line discharge fault identification is performed, and 1 indicates that the corresponding physical quantity is considered when the tree line discharge fault identification is performed.
Further, the genetic algorithm is an optimized genetic algorithm optimized by dynamic programming, and the dynamic programming optimization process specifically comprises the following steps:
the number of non-zero elements in the vector is used as a state quantity S, and different S represent different stages;
judging whether vectors with fitness not less than a set threshold value exist from S=1 in all vectors generated by the limited iteration times; if not, the search is continued in the s=2 state until a satisfactory vector is found.
Further, the fitness is the difference between the sample center of the K-means clustering analysis of the data set in the physical quantity combination and the sample center of the actual classification.
In a second aspect, a distribution network tree line discharge fault identification system is provided, including:
the sample construction module is used for establishing a basic characteristic sample library of the tree line discharge fault of the low-current grounding system;
the random combination module is used for carrying out random combination on physical quantities in the basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination;
the clustering analysis module is used for carrying out clustering analysis on the physical quantity combinations through a K-means method, and selecting the physical quantity combinations with fitness not smaller than a set threshold as optimized feature samples;
the model training module is used for training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model;
the fault recognition module is used for inputting the physical quantity data acquired from the power distribution network in real time into the tree line discharge fault recognition model to perform fault recognition, and a fault recognition result is obtained.
In a third aspect, a computer terminal is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements a method for identifying a tree discharge fault of a power distribution network according to any one of the first aspects when the program is executed.
In a fourth aspect, a computer readable medium is provided, on which a computer program is stored, the computer program being executable by a processor to implement a method for identifying a tree discharge fault of a power distribution network according to any one of the first aspects.
Compared with the prior art, the invention has the following beneficial effects:
1. according to the power distribution network tree line discharge fault identification method provided by the invention, a genetic algorithm is adopted to screen a large amount of data, the dimension of the processed data is reduced, then clustering analysis is carried out on the data, and whether the operation data is in a fault state or a normal operation state is distinguished, so that early identification of the tree line discharge fault in a complex interference environment is realized;
2. the invention adopts the K-means clustering algorithm to effectively process the large data set and accelerate the convergence rate;
3. according to the invention, dynamic programming is added in the selection process, and compared with the traditional static programming, the dynamic programming reflects the relation and characteristics of the evolution of the dynamic process, so that the solving efficiency is improved, and the global optimal solution can be obtained;
4. the method adopts the optimized genetic algorithm to obtain the key physical quantity data after data screening, so that the BP neural network can realize accurate cluster analysis; compared with the method which only adopts the BP neural network, the genetic algorithm is added to carry out the data screening step, on one hand, because the genetic algorithm is a directed random algorithm, the method adopts directed search to avoid the problem of local minimization possibly occurring in the traditional BP neural network; on the other hand, the selected key physical quantity participates in cluster analysis, so that the operation time is reduced, and the operation speed is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention. In the drawings:
FIG. 1 is a flow chart in an embodiment of the invention;
fig. 2 is a system block diagram in an embodiment of the invention.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
Example 1: a power distribution network tree line discharge fault identification method, as shown in figure 1, comprises the following steps:
step S1: establishing a basic characteristic sample library of tree line discharge faults of the low-current grounding system;
step S2: randomly combining physical quantities in a basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination;
step S3: clustering analysis is carried out on the physical quantity combinations through a K-means method, and the physical quantity combinations with fitness not less than a set threshold value are selected as optimized feature samples;
step S4: training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model;
step S5: and inputting the physical quantity data acquired from the power distribution network in real time into a tree line discharge fault identification model to carry out fault identification, so as to obtain a fault identification result.
According to the method, the physical quantity combination is randomly generated, the characteristic data is reserved, and the variables with low association degree are removed, so that the database is simplified.
In step S1, since the tree line discharge fault presents different fault characteristics in a complex line environment, in order to accurately identify the tree line discharge fault, the tree line discharge fault characteristic quantity is obtained through a laboratory simulation experiment and a means of Comsol/Simulink simulation, the distribution parameter discharge simulation and the circuit lumped parameter simulation are coupled, and the equivalent impedance model of the tree line discharge fault is established according to the circuit principle. And processing fault characteristic data obtained through experiments and simulations by adopting a database integration method, and establishing a basic characteristic sample library of the tree line discharge faults of the low-current grounding system.
And for a simulation experiment, a 10kV distribution line tree line discharge simulation experiment platform is built, and a tree line discharge fault is constructed. The leakage current of the fault point is grounded through a sampling resistor, a voltage waveform is measured at the high-voltage side of the sampling resistor, the waveform is connected to an oscilloscope, the leakage current waveform is obtained according to ohm's law, and the first characteristic quantity can be obtained.
For simulation analysis, a tree line discharge distribution model is established on a Comsol simulation platform, tree line discharge is abstracted into needle-plate discharge, trees are needle electrodes, and the lines can be abstracted into plate electrodes relative to the size of the tips of the trees; establishing distribution parameter simulation of tree line discharge through a Comsol simulation platform, and calculating a charge transfer and transport process in the tree line discharge process to obtain voltage and current waveforms generated by the tree line discharge; and inputting the voltage and current waveforms into a lumped parameter power distribution network model in the Simulink platform to obtain a second characteristic quantity.
The characteristic amounts in the present embodiment include, but are not limited to, an electric signal such as current, voltage, impedance, and the like, a physical signal such as vibration, and a temperature signal.
In step S2, using a genetic algorithm, a plurality of physical quantity combinations of the tree line discharge fault characteristics can be randomly generated from all the physical quantities contained in the sample library, and the physical quantity combinations are tested for the accuracy of the cluster analysis and selected, so that the physical quantity combinations capable of ensuring the fault identification accuracy can be screened.
In the present embodiment, the physical quantity combination of the tree line discharge fault characteristics is represented by a plurality of sets of vectors including only 0/1, where 0 represents that the physical quantity is not considered when the tree line discharge fault identification is performed, and 1 represents that the physical quantity is considered when the tree line discharge fault identification is performed.
In order to screen out excellent combinations, the invention creatively designs an adaptability function for measuring the quality of the combinations. The fitness determined by the fitness function is the difference between the sample center of the K-means clustering analysis of the data set in the physical quantity combination and the sample center of the actual classification.
Specifically, the specific operation flow of the genetic algorithm application method provided by the invention is as follows:
a: randomly generating 10 groups of physical quantity combination vectors formed by initial tree line discharge fault characteristics as a first generation, taking 5 groups with minimum adaptability for crossing, wherein the sub-vector elements are respectively from two male parent vectors;
b: to ensure algorithm convergence, a crossover rate is introduced. Let the crossover rate be eta c ，0＜η c < 1, randomly generating a number i between 0 and 1 before crossing, if i < eta c Performing cross operation; otherwise, selecting any one of the two male parent vectors as a next generation child vector; the crossover rate is typically set at 80% -90%.
C: to increase the randomness of the combination, mutation operators are added. Let the mutation rate be eta m ，0＜η m < 1, randomly generating a number j between 0 and 1 for each element of the sub-vector when performing cross operation, if j < eta m Then, a mutation operation is performed, i.e. a non-operation is performed on the element. The variability can increase the randomness of the algorithm, and is preferably not too large, but 0.5% -1%.
D: and (3) ending one iteration, judging whether a physical quantity combination with the fitness not smaller than a set threshold exists in the sub-vector, if so, ending the iteration, and if not, continuing the iteration.
In order to ensure that the least physical quantity is considered in the identification of the tree line discharge fault characteristics, the genetic algorithm is an optimized genetic algorithm which is optimized through dynamic programming, and the process of the dynamic programming optimization specifically comprises the following steps: the number of non-zero elements in the vector is used as a state quantity S, and different S represent different stages; judging whether vectors with fitness not less than a set threshold value exist from S=1 in all vectors generated by the limited iteration times; if not, the search is continued in the s=2 state until a satisfactory vector is found.
In step S3, since the tree line discharge fault identification does not have reliable experience and basis yet, the classification is directly completed on the mathematical level by the method of the cluster analysis, and the classified data is corresponding to the running state (whether fault, fault type) of the line, so as to realize the fault identification.
Based on the simplified characteristic sample library, the self-learning clustering analysis is realized by adopting an artificial neural network method. And training the artificial neural network by utilizing the characteristic sample library data, and adjusting the connection weight of the neurons by self-learning according to the training process by the artificial neural network until a good classification effect is realized, wherein the neural network model can be used for directly carrying out cluster analysis on the operation data after the training is finished.
In step S4, the BP neural network training applied in the present invention is divided into two stages. The first stage is forward transmission of signals from an input layer to an hidden layer and from the hidden layer to an output layer; the second stage is reverse transmission of errors, and the errors are transmitted back to the hidden layer from the output layer and finally transmitted back to the input layer, and weights and biases of the hidden layer, the output layer and the input layer are sequentially adjusted. The number of the input layer nodes is identical to the number of the characteristic physical quantities, the number of the output layer nodes is identical to the number of the running state types (the number of the fault types+the normal running), and the number of the hidden layer nodes can be obtained through experiments.
The learning process of the BP neural network is divided into eight steps:the first step: network initialization, namely respectively assigning a random number with a range (-1, 1) to each connection weight, setting an error function e, setting a calculation precision value e and a maximum training frequency M; and a second step of: randomly selecting a kth input sample x (k) and a corresponding output expectation; and a third step of: calculating the input and output of each neuron of the hidden layer; fourth step: calculating partial derivatives d of error function on each unit of output layer by using expected output and actual output of neural network 0 (k) The method comprises the steps of carrying out a first treatment on the surface of the Fifth step: d using the neurons of the output layer 0 (k) And modifying the connection weight w by implicit layer neuron outputs h0 (k) Typically the learning rate h is selected to be within a range of (0, 1); sixth step: d using hidden layers of neurons h (k) And input of each neuron of the input layer to correct the connection weight; seventh step: calculating a global error; eighth step: judging whether the network error meets the requirement, if the error reaches the preset precision or the training times is larger than the preset maximum times, ending the training, otherwise, selecting the next training sample and the corresponding expected output, returning to the third step, and entering the next training round.
Example 2: a power distribution network tree line discharge fault identification system is used for realizing the power distribution network tree line discharge fault identification method described in the embodiment 1, and comprises a sample construction module, a random combination module, a cluster analysis module, a model training module and a fault identification module as shown in fig. 2.
The system comprises a sample construction module, a sampling module and a sampling module, wherein the sample construction module is used for establishing a basic characteristic sample library of the tree line discharge fault of the low-current grounding system; the random combination module is used for carrying out random combination on physical quantities in the basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination; the clustering analysis module is used for carrying out clustering analysis on the physical quantity combinations through a K-means method, and selecting the physical quantity combinations with fitness not smaller than a set threshold as optimized feature samples; the model training module is used for training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model; the fault recognition module is used for inputting the physical quantity data acquired from the power distribution network in real time into the tree line discharge fault recognition model to perform fault recognition, and a fault recognition result is obtained.
Working principle: the invention adopts a genetic algorithm to screen a large amount of data, reduces the dimension of the processed data, then carries out cluster analysis on the data, and distinguishes whether the operation data is in a fault state or a normal operation state, thereby realizing early identification of tree line discharge faults under a complex interference environment; in addition, the K-means clustering algorithm is adopted to effectively process a large data set, so that the convergence rate can be increased; in addition, dynamic programming is added in the selection process, and compared with the traditional static programming, the dynamic programming reflects the relation and characteristics of the evolution of the dynamic process, improves the solving efficiency and can obtain the global optimal solution.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing detailed description of the invention has been presented for purposes of illustration and description, and it should be understood that the invention is not limited to the particular embodiments disclosed, but is intended to cover all modifications, equivalents, alternatives, and improvements within the spirit and principles of the invention.
Claims (10)
1. The power distribution network tree line discharge fault identification method is characterized by comprising the following steps of:
establishing a basic characteristic sample library of tree line discharge faults of the low-current grounding system;
randomly combining physical quantities in a basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination;
clustering analysis is carried out on the physical quantity combinations through a K-means method, and the physical quantity combinations with fitness not less than a set threshold value are selected as optimized feature samples;
training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model;
and inputting the physical quantity data acquired from the power distribution network in real time into a tree line discharge fault identification model to carry out fault identification, so as to obtain a fault identification result.
2. The method for identifying a tree discharge fault of a power distribution network according to claim 1, wherein the establishing process of the basic characteristic sample library is specifically as follows:
acquiring high-dimensional tree line discharge fault feature quantity through a laboratory tree line discharge simulation experiment to obtain a first feature quantity;
analyzing the tree line discharge fault characteristic quantity through a simulation platform to obtain a second characteristic quantity;
and integrating the first characteristic quantity and the second characteristic quantity, and then establishing a basic characteristic sample library.
3. The method for identifying a tree discharge fault of a power distribution network according to claim 2, wherein the obtaining process of the second characteristic quantity specifically comprises the following steps:
establishing a tree line discharge distribution model on a Comsol simulation platform, abstracting the tree line discharge into needle-plate discharge, wherein the tree is a needle electrode, and the line is abstracted into a plate electrode;
establishing distribution parameter simulation of tree line discharge through a Comsol simulation platform, and calculating a charge transfer and transport process in the tree line discharge process to obtain voltage and current waveforms generated by the tree line discharge;
and inputting the voltage and current waveforms into a lumped parameter power distribution network model in the Simulink platform to obtain a second characteristic quantity.
4. The method for identifying a tree line discharge fault of a power distribution network according to claim 1, wherein the process of randomly combining physical quantities in a basic characteristic sample library by adopting a genetic algorithm is specifically as follows:
randomly generating P groups of physical quantity combination vectors formed by initial tree line discharge fault characteristics as a first generation, intersecting Q with minimum adaptability, wherein the sub-vector elements are respectively from two male parent vectors, and Q is smaller than P;
let the crossover rate be eta c ，0＜η c < 1, randomly generating a number i between 0 and 1 before crossing, if i < eta c Performing cross operation; otherwise, selecting any one of the two male parent vectors as a next generation child vector;
let the mutation rate be eta m ，0＜η m < 1, randomly generating a number j between 0 and 1 for each element of the sub-vector when performing cross operation, if j < eta m Performing mutation operation;
and (3) ending one iteration, judging whether a physical quantity combination with the fitness not smaller than a set threshold exists in the sub-vector, if so, ending the iteration, and if not, continuing the iteration.
5. The method for identifying a tree discharge fault of a power distribution network according to claim 1, wherein the combination of physical quantities is represented by a plurality of sets of vectors containing only 0/1; wherein 0 indicates that the corresponding physical quantity is not considered when the tree line discharge fault identification is performed, and 1 indicates that the corresponding physical quantity is considered when the tree line discharge fault identification is performed.
6. The method for identifying the tree line discharge faults of the power distribution network according to claim 1, wherein the genetic algorithm is an optimized genetic algorithm optimized by dynamic programming, and the process of dynamic programming optimization is specifically as follows:
the number of non-zero elements in the vector is used as a state quantity S, and different S represent different stages;
judging whether vectors with fitness not less than a set threshold value exist from S=1 in all vectors generated by the limited iteration times; if not, the search is continued in the s=2 state until a satisfactory vector is found.
7. The method for identifying a tree discharge fault of a power distribution network according to claim 1, wherein the fitness is a difference between a sample center of a K-means clustering analysis of a data set in a physical quantity combination and a sample center of an actual classification.
8. The utility model provides a distribution network tree line fault identification system that discharges which characterized in that includes:
the sample construction module is used for establishing a basic characteristic sample library of the tree line discharge fault of the low-current grounding system;
the random combination module is used for carrying out random combination on physical quantities in the basic characteristic sample library by adopting a genetic algorithm to obtain a physical quantity combination;
the clustering analysis module is used for carrying out clustering analysis on the physical quantity combinations through a K-means method, and selecting the physical quantity combinations with fitness not smaller than a set threshold as optimized feature samples;
the model training module is used for training the artificial neural network according to all the optimized characteristic samples to obtain a tree line discharge fault identification model;
the fault recognition module is used for inputting the physical quantity data acquired from the power distribution network in real time into the tree line discharge fault recognition model to perform fault recognition, and a fault recognition result is obtained.
9. A computer terminal comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements a method for identifying a tree discharge fault in a power distribution network according to any one of claims 1-7 when executing the program.
10. A computer readable medium having stored thereon a computer program, wherein the computer program is executable by a processor to implement a method for identifying a tree discharge fault in a power distribution network according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310384768.2A CN116087692B (en) | 2023-04-12 | 2023-04-12 | Distribution network tree line discharge fault identification method, system, terminal and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310384768.2A CN116087692B (en) | 2023-04-12 | 2023-04-12 | Distribution network tree line discharge fault identification method, system, terminal and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116087692A CN116087692A (en) | 2023-05-09 |
| CN116087692B true CN116087692B (en) | 2023-06-23 |
Family
ID=86212459
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310384768.2A Active CN116087692B (en) | 2023-04-12 | 2023-04-12 | Distribution network tree line discharge fault identification method, system, terminal and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116087692B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015158198A1 (en) * | 2014-04-17 | 2015-10-22 | 北京泰乐德信息技术有限公司 | Fault recognition method and system based on neural network self-learning |
| CN111060815A (en) * | 2019-12-17 | 2020-04-24 | 西安工程大学 | GA-Bi-RNN-based high-voltage circuit breaker fault diagnosis method |
| CN111208436A (en) * | 2020-02-21 | 2020-05-29 | 河海大学 | Energy storage battery overcharge diagnosis method based on IMF energy moment and genetic algorithm optimization SVM |
| CN111596167A (en) * | 2020-05-14 | 2020-08-28 | 中国南方电网有限责任公司 | Fault section positioning method and device based on fuzzy c-means clustering algorithm |
| CN112684295A (en) * | 2020-12-31 | 2021-04-20 | 国网河南省电力公司电力科学研究院 | Power distribution network fault line selection method and system under high permeability based on similarity separation degree |
| CN113447766A (en) * | 2021-08-17 | 2021-09-28 | 广东电网有限责任公司东莞供电局 | Method, device, equipment and storage medium for detecting high-resistance ground fault |
| CN114019296A (en) * | 2021-09-26 | 2022-02-08 | 广西电网有限责任公司电力科学研究院 | Distribution line ground fault identification method based on BP neural network |
| WO2023019601A1 (en) * | 2021-08-16 | 2023-02-23 | 苏州大学 | Signal modulation recognition method for complex-valued neural network based on structure optimization algorithm |
Family Cites Families (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6879917B2 (en) * | 2002-06-14 | 2005-04-12 | Progress Energy Carolinas Inc. | Double-ended distance-to-fault location system using time-synchronized positive-or negative-sequence quantities |
| JP4142608B2 (en) * | 2004-04-07 | 2008-09-03 | 株式会社日立製作所 | Tree contact monitoring device for distribution lines |
| US8041571B2 (en) * | 2007-01-05 | 2011-10-18 | International Business Machines Corporation | Application of speech and speaker recognition tools to fault detection in electrical circuits |
| CN107490760A (en) * | 2017-08-22 | 2017-12-19 | 西安工程大学 | The circuit breaker failure diagnostic method of fuzzy neural network is improved based on genetic algorithm |
| CN109459671A (en) * | 2018-09-27 | 2019-03-12 | 国网浙江省电力有限公司湖州供电公司 | A kind of switch cabinet state monitoring method of the improvement neural network based on genetic algorithm |
| CN109444667B (en) * | 2018-12-17 | 2021-02-19 | 国网山东省电力公司电力科学研究院 | Power distribution network early fault classification method and device based on convolutional neural network |
| CN111141794A (en) * | 2020-01-15 | 2020-05-12 | 合肥工业大学 | FPGA welding point fault online state monitoring method |
| JP2022085487A (en) * | 2020-11-27 | 2022-06-08 | 三菱電機株式会社 | Accident sign detection system, accident sign detection method, and master and slave stations of accident sign detection system |
| CN113033837A (en) * | 2021-03-05 | 2021-06-25 | 国网电力科学研究院武汉南瑞有限责任公司 | Artificial intelligence fault identification system and method based on transient waveform of power transmission line |
| CN113945796B (en) * | 2021-09-29 | 2023-10-24 | 集美大学 | Power distribution network fault positioning method, terminal equipment and storage medium |
| CN114186590A (en) * | 2021-12-10 | 2022-03-15 | 广西电网有限责任公司钦州供电局 | Power distribution network single-phase earth fault identification method based on wavelet and deep learning |
| CN114487805A (en) * | 2022-01-19 | 2022-05-13 | 西安零壹智能电器有限公司 | Circuit breaker fault diagnosis method based on genetic algorithm and clustering algorithm |
| CN114675132A (en) * | 2022-03-31 | 2022-06-28 | 云南电网有限责任公司电力科学研究院 | Tree line fault identification method, simulation device, system, computer device and medium |
| CN115017444A (en) * | 2022-04-18 | 2022-09-06 | 上海交通大学 | Positioning and identifying method for early faults of tree line |
| CN114757291B (en) * | 2022-04-26 | 2023-05-23 | 国网四川省电力公司电力科学研究院 | Single-phase fault identification optimization method, system and equipment based on machine learning algorithm |
| CN115144702A (en) * | 2022-07-26 | 2022-10-04 | 山东科汇电力自动化股份有限公司 | Ground fault type identification method based on self-organizing competitive neural network |
| CN115758634A (en) * | 2022-10-25 | 2023-03-07 | 昆明能讯科技有限责任公司 | Method for initiating tree combustion process space-time evolution through circuit discharge and storage medium |
-
2023
- 2023-04-12 CN CN202310384768.2A patent/CN116087692B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015158198A1 (en) * | 2014-04-17 | 2015-10-22 | 北京泰乐德信息技术有限公司 | Fault recognition method and system based on neural network self-learning |
| CN111060815A (en) * | 2019-12-17 | 2020-04-24 | 西安工程大学 | GA-Bi-RNN-based high-voltage circuit breaker fault diagnosis method |
| CN111208436A (en) * | 2020-02-21 | 2020-05-29 | 河海大学 | Energy storage battery overcharge diagnosis method based on IMF energy moment and genetic algorithm optimization SVM |
| CN111596167A (en) * | 2020-05-14 | 2020-08-28 | 中国南方电网有限责任公司 | Fault section positioning method and device based on fuzzy c-means clustering algorithm |
| CN112684295A (en) * | 2020-12-31 | 2021-04-20 | 国网河南省电力公司电力科学研究院 | Power distribution network fault line selection method and system under high permeability based on similarity separation degree |
| WO2023019601A1 (en) * | 2021-08-16 | 2023-02-23 | 苏州大学 | Signal modulation recognition method for complex-valued neural network based on structure optimization algorithm |
| CN113447766A (en) * | 2021-08-17 | 2021-09-28 | 广东电网有限责任公司东莞供电局 | Method, device, equipment and storage medium for detecting high-resistance ground fault |
| CN114019296A (en) * | 2021-09-26 | 2022-02-08 | 广西电网有限责任公司电力科学研究院 | Distribution line ground fault identification method based on BP neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116087692A (en) | 2023-05-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111046581A (en) | Power transmission line fault type identification method and system | |
| CN108616120A (en) | A kind of non-intrusive electrical load decomposition method based on RBF neural | |
| LU500551B1 (en) | Virtual load dominant parameter identification method based on incremental learning | |
| CN110969194B (en) | Cable early fault positioning method based on improved convolutional neural network | |
| CN111160241B (en) | Power distribution network fault classification method, system and medium based on deep learning | |
| CN106656357B (en) | Power frequency communication channel state evaluation system and method | |
| CN114169231A (en) | Method for obtaining distribution line fault classification, positioning and line selection deep learning model based on transfer learning | |
| CN112485622A (en) | Partial discharge pattern recognition method based on GA-BPNN | |
| CN109359665A (en) | A kind of family's electric load recognition methods and device based on support vector machines | |
| CN110866366A (en) | XGboost algorithm-based island detection method for photovoltaic microgrid containing PHEV | |
| CN114757286A (en) | Multi-class fault data generation method based on conditional countermeasure generation network | |
| CN112069723A (en) | Method and system for evaluating transient stability of power system | |
| CN114330486A (en) | Power system bad data identification method based on improved Wasserstein GAN | |
| CN114266396A (en) | Transient stability discrimination method based on intelligent screening of power grid characteristics | |
| CN116087692B (en) | Distribution network tree line discharge fault identification method, system, terminal and medium | |
| CN113379116A (en) | Cluster and convolutional neural network-based line loss prediction method for transformer area | |
| CN117272184A (en) | ISSA-RF algorithm-based photovoltaic power generation system fault classification method | |
| CN116882838A (en) | Short-term voltage stability assessment method based on Stacking integration strategy | |
| CN116401532A (en) | Method and system for recognizing frequency instability of power system after disturbance | |
| CN116756939A (en) | Transformer fault diagnosis method based on ANOVA-IWAA-SVM hybrid model | |
| CN115795360A (en) | Cable fault detection method based on artificial neural network | |
| CN115409317A (en) | Transformer area line loss detection method and device based on feature selection and machine learning | |
| CN114881176A (en) | Non-invasive load identification method based on self-adaptive optimization random forest | |
| CN114861792A (en) | Complex power grid key node identification method based on deep reinforcement learning | |
| CN114744631A (en) | Data driving voltage estimation method based on non-PMU power distribution network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |