CN113902946A - Power system fault direction judging method and device, terminal equipment and storage medium - Google Patents
Power system fault direction judging method and device, terminal equipment and storage medium Download PDFInfo
- Publication number
- CN113902946A CN113902946A CN202111172769.8A CN202111172769A CN113902946A CN 113902946 A CN113902946 A CN 113902946A CN 202111172769 A CN202111172769 A CN 202111172769A CN 113902946 A CN113902946 A CN 113902946A
- Authority
- CN
- China
- Prior art keywords
- fault
- neural network
- power system
- data set
- convolutional neural
- 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
Images
Classifications
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention belongs to the technical field of power system fault detection, and discloses a method and a device for judging a fault direction of a power system, terminal equipment and a storage medium. The method comprises the following steps: constructing a system simulation model of the power system to be detected, and acquiring a fault data set; performing data preprocessing on the fault data set to obtain an electric power training data set and an electric power test data set; constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the electric power training data set to obtain a convolutional neural network model to be tested; inputting the power test data set into a convolutional neural network model to be tested for testing, and taking the convolutional neural network model to be tested as a fault test model when a test result meets a preset test condition; and judging the fault direction of the power system to be detected according to the fault test model. According to the method, the fault direction in the power system is quickly judged by utilizing the convolutional neural network model.
Description
Technical Field
The invention relates to the technical field of power system fault detection, in particular to a method and a device for judging fault direction of a power system, terminal equipment and a storage medium.
Background
The power grid can safely and stably operate and has great significance to national economy. Nowadays, the size and coverage of power distribution networks are increasing, and the complexity of their operation is also increasing, and as power distribution networks are approaching their maximum load during operation, the frequency and risk of failures are increasing. The circuit fault in the power distribution network seriously affects the stability and safety of system operation, causes huge loss to social safety production, and further affects the daily life and work of the people. If the fault position is rapidly judged when a fault occurs and the fault line section is isolated for troubleshooting, the loss can be timely stopped, the influence on the rest normal parts of the system can be prevented, the fault recovery time can be obviously shortened, the economic loss is reduced, and the reliability of the power system is improved. Therefore, a quick, reliable and accurate fault azimuth determination method is very important.
The existing fault azimuth determination methods are roughly classified into two types: the first type is a traditional fault azimuth determination method, which can be mainly divided into an impedance method, an injection method and a traveling wave method. The method mainly considers the physical parameters such as voltage, node impedance, traveling wave signals and the like in the power grid, but the calculation is relatively complex, and in recent years, the distribution lines are more and more complex, and more branches are provided, so that the orientation determination and positioning efficiency is low. The second type is an automatic fault azimuth judgment method which mainly comprises a matrix algorithm, a bat algorithm, a genetic algorithm, a particle swarm algorithm, a neural network algorithm and the like, wherein the algorithms consider different structural characteristics and fault conditions of the power distribution network, and the fault tolerance and the accuracy are improved to a certain extent. However, the method still has limitations and is easy to fall into the situation of local optimization, and meanwhile, how to construct a proper objective function and a proper switching function is the bottleneck of solving the model.
The above is only for the purpose of assisting understanding of the technical aspects of the present invention, and does not represent an admission that the above is prior art.
Disclosure of Invention
The invention mainly aims to provide a method and a device for judging fault positions of a power system, a terminal device and a storage medium, and aims to solve the technical problem that the efficiency of judging fault positions of the power system is low in the prior art.
In order to achieve the above object, the present invention provides a method for determining fault location of an electrical power system, the method comprising the following steps:
constructing a system simulation model of a power system to be detected, and acquiring a fault data set of the power system to be detected;
performing data preprocessing on the fault data set, and acquiring a power training data set and a power test data set according to a processing result;
constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the electric power training data set to obtain a convolutional neural network model to be tested;
inputting the power test data set into the convolutional neural network model to be tested for testing, and taking the convolutional neural network model to be tested as a fault test model corresponding to the power system to be tested when a test result meets a preset test condition;
and judging the fault direction of the power system to be detected according to the fault test model.
Optionally, the step of constructing a system simulation model of the to-be-detected power system and acquiring a fault data set of the to-be-detected power system specifically includes:
the method comprises the steps of obtaining a system type of a power system to be detected, and establishing a system simulation model according to the system type;
acquiring a preset fault information set, and performing fault simulation on the system simulation model according to the preset fault information set;
and detecting fault data generated in the fault simulation process to obtain a fault data set.
Optionally, the step of performing data preprocessing on the fault data set and acquiring a power training data set and a power testing data set according to a processing result specifically includes:
acquiring abnormal data and normal data in the fault data set according to a preset numerical boundary, and performing numerical correction on the abnormal data to acquire corrected data;
carrying out data normalization processing on the corrected data and the normal data to obtain a fault data set preprocessing result;
and performing set division on the preprocessing result of the fault data set according to a preset set proportion to obtain an electric power training data set and an electric power test data set.
Optionally, the step of constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the power training data set to obtain a convolutional neural network model to be tested specifically includes:
and constructing a one-dimensional convolutional neural network model as an initial convolutional neural network model, inputting the electric power training data set into the initial convolutional neural network model, and training according to a preset optimization target to obtain the convolutional neural network model to be tested.
Optionally, the preset optimization objective is to minimize a cross entropy loss function as an objective.
Optionally, the step of judging the fault location of the power system to be detected according to the fault test model specifically includes:
acquiring current operation data of a power system to be detected, and inputting the current operation data into the fault detection model;
and when the output result is that a fault exists, determining the current fault position of the power system to be detected according to the fault detection model and the system simulation model.
Optionally, after the step of constructing an initial convolutional neural network model and training the initial convolutional neural network model according to the power training data set to obtain a convolutional neural network model to be tested, the method further includes:
and inputting the power test data set into the convolutional neural network model to be tested for testing, and performing optimization training on the convolutional neural network model to be tested when a test result does not meet a preset test condition.
In addition, in order to achieve the above object, the present invention further provides an apparatus for determining a fault location of an electrical power system, the apparatus including:
the first model building module is used for building a system simulation model of the power system to be detected and acquiring a fault data set of the power system to be detected;
the data processing module is used for carrying out data preprocessing on the fault data set and acquiring an electric power training data set and an electric power test data set according to a processing result;
the second model building module is used for building an initial convolutional neural network model and training the initial convolutional neural network model according to the electric power training data set so as to obtain a convolutional neural network model to be tested;
the model testing module is used for inputting the power testing data set into the convolutional neural network model to be tested for testing, and when a testing result meets a preset testing condition, the convolutional neural network model to be tested is used as a fault testing model corresponding to the power system to be tested;
and the fault detection module is used for judging the fault direction of the power system to be detected according to the fault test model.
In addition, in order to achieve the above object, the present invention further provides a terminal device, which includes a memory, a processor, and a power system fault orientation determination program stored on the memory and operable on the processor, wherein the power system fault orientation determination program is configured to implement the steps of the power system fault orientation determination method as described above.
In addition, in order to achieve the above object, the present invention further provides a storage medium having a power system fault orientation determination program stored thereon, wherein the power system fault orientation determination program, when executed by a processor, implements the steps of the power system fault orientation determination method as described above.
The method comprises the steps of constructing a system simulation model of the power system to be detected, and acquiring a fault data set of the power system to be detected; performing data preprocessing on the fault data set, and acquiring a power training data set and a power test data set according to a processing result; constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the electric power training data set to obtain a convolutional neural network model to be tested; inputting the power test data set into the convolutional neural network model to be tested for testing, and taking the convolutional neural network model to be tested as a fault test model corresponding to the power system to be tested when a test result meets a preset test condition; and judging the fault direction of the power system to be detected according to the fault test model. The method is based on the deep learning technology, and can effectively judge the fault direction in the power system by establishing a model for the power system and training a detection model according to the data of the power system, thereby greatly improving the efficiency and the accuracy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the structures shown in the drawings without creative efforts.
Fig. 1 is a schematic structural diagram of a terminal device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart illustrating a first embodiment of a method for determining a fault location of an electrical power system according to the present invention;
FIG. 3 is a schematic diagram of a convolutional neural network of the fault direction determination method of the power system of the present invention;
FIG. 4 is a flowchart illustrating a second embodiment of the method for determining a fault location of an electrical power system according to the present invention;
fig. 5 is a block diagram illustrating a first embodiment of a fault location determination apparatus for an electric power system according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a terminal device in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the terminal device may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a Random Access Memory (RAM) Memory, or may be a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the terminal device and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include therein an operating system, a network communication module, a user interface module, and a power system fault location determination program.
In the terminal device shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the terminal device of the present invention may be provided in the terminal device, and the terminal device calls the power system fault orientation determination program stored in the memory 1005 through the processor 1001 and executes the power system fault orientation determination method provided by the embodiment of the present invention.
An embodiment of the present invention provides a method for determining a fault location of an electrical power system, and referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of the method for determining a fault location of an electrical power system according to the present invention.
In this embodiment, the method for determining the fault location of the power system includes the following steps:
step S10: the method comprises the steps of constructing a system simulation model of the electric power system to be detected, and acquiring a fault data set of the electric power system to be detected.
It should be noted that the system simulation model may be constructed by using a visual simulation tool Simulink to obtain system data of the power system to be detected, which includes but is not limited to: and constructing the power system to be detected by branch impedance, load of each node, interconnection switch parameters and the like. The fault data set comprises various parameters of the power system to be detected when the power system to be detected is in fault, and the various parameters can be historical fault parameters of the power system to be detected or fault parameters obtained by fault simulation through a system simulation model.
Further, step S10 specifically includes: the method comprises the steps of obtaining a system type of a power system to be detected, and establishing a system simulation model according to the system type; acquiring a preset fault information set, and performing fault simulation on the system simulation model according to the preset fault information set; and detecting fault data generated in the fault simulation process to obtain a fault data set.
In this embodiment, the system type of the power system to be detected is an IEEE standard 33 node as an example for explanation (in a specific implementation, the method may also be used in power systems of other system types, and this embodiment does not limit an actual application scenario of the present invention). In specific implementations, for example: an IEEE33 node power distribution network model is built by using MATLAB/Simulink, and modules such as Simscape, Sources and Sinks in the Simulink are compared with a standard 33 node power distribution network to perform assembly.
Further, to obtain a failure data set, based on the above example: adding a monitoring point module between preset nodes in a simulation model of an IEEE33 node, wherein the nodes can be between 8 and 9 nodes; setting four random value model parameters of short circuit type, short circuit occurrence position, short circuit occurrence time (initial phase) and short circuit resistance value, and carrying out simulation on short circuit fault; furthermore, in order to acquire more fault data, an Asynchronous motor module of Asynchronous Machine can be added, and the motor is started statically to cause voltage sag and simulate load impact fault; the circuit data on the monitoring points are collected, and the circuit data comprise voltage, current and power data of each node, data of a branch where a fault is located and the like, and the data are added into a fault data set.
Step S20: performing data preprocessing on the fault data set, and acquiring a power training data set and a power test data set according to a processing result;
it is easy to understand that for effective model training, some abnormal data in the fault data need to be deleted, so the fault data set is subjected to data preprocessing. Processing the data anomaly points: during the simulation of the circuit, the recorded circuit data may be abnormal due to environmental and operational problems, and the data points are unreasonable, and if the data points are not processed, the subsequent result may be deviated.
Step S20 specifically includes: acquiring abnormal data and normal data in the fault data set according to a preset numerical boundary, and performing numerical correction on the abnormal data to acquire corrected data; carrying out data normalization processing on the corrected data and the normal data to obtain a fault data set preprocessing result; and performing set division on the preprocessing result of the fault data set according to a preset set proportion to obtain an electric power training data set and an electric power test data set.
In specific implementation, the preset numerical value boundary can be implemented by box plot analysis, that is, the numerical value larger than or smaller than the upper and lower bounds set by the box plot is an abnormal value, and then the identified abnormal value is corrected by the average value of the upper and lower neighbor points of the abnormal value, so as to obtain corrected data.
Further, since the types of input parameters are different and the numerical difference is large, the main solution to such problems is to normalize the input parameter data, and the most commonly adopted method for normalizing the data is min-max normalization, which is also called dispersion normalization, and is a linear transformation on the original data, so that the result value is mapped between [0,1], and the formula is as follows:
wherein XmaxMaximum value of single-column characteristic data, XminIs the minimum value of the single column of characteristic data.
Further, the preset ratio may be 7: the implementation may include a power training data set 700 and a power test data set 300.
Step S30: constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the electric power training data set to obtain a convolutional neural network model to be tested;
step S30 specifically includes: and constructing a one-dimensional convolutional neural network model as an initial convolutional neural network model, inputting the electric power training data set into the initial convolutional neural network model, and training according to a preset optimization target to obtain the convolutional neural network model to be tested.
It should be noted that, a one-dimensional CNN convolutional neural network model is built, an input layer, a hidden layer and an output layer of the neural network are determined, the number of neurons of the input layer corresponds to input parameter characteristics, and network weights are initialized, the one-dimensional CNN convolutional neural network designed by the method has 9 layers in total, and comprises 1 input layer, 4 convolutional layers and 2 pooling layers, 1 discarding layer and 1 fully-connected layer output layer, and the output layer is a position predicted by the neural network to have a fault;
it should be noted that, referring to fig. 3, fig. 3 is a schematic diagram of a convolutional neural network according to the method for determining a fault location of a power system of the present invention, where the preset optimization target is a target of minimizing a cross entropy loss function. The method comprises the steps of inputting a training data set into a 1D CNN-based neural network model for training, using minimization of a cross entropy loss function as an optimization target, adopting an Adam optimizer to update and adjust model parameters, reducing prediction errors, using a test set for testing to obtain an ideal prediction model, storing a network model architecture and network parameters with ideal effects, and using a cross entropy loss function loss and accuracy as evaluation indexes.
Where M denotes the number of categories, yicRepresents a symbolic function (0 or 1), if the true class of sample i is equal to c, then yicGet 1, otherwise get 0, picRepresenting the predicted probability that sample i belongs to class c.
Wherein, TP represents the number of correct positive case predictions, FP represents the number of incorrect negative case predictions, TN represents the number of correct negative case predictions, and FN represents the number of incorrect positive case predictions.
It should be noted that, the specific network structure parameters are set according to actual requirements, and are not explained one by one in this embodiment. The meaning of each layer of the convolutional neural network model to be tested is as follows: (1) an input layer: one acquisition interval is 0.00001 second, only data of a fault occurrence period is acquired, and 13000 sampling points are total. After the data are preprocessed, each group of data comprises three-phase voltage values Ua, Ub and Uc and three-phase current values Ia, Ib and Ic, and in the neural network, the data are spread into 13000 multiplied by 6 vectors and then transmitted into the neural network.
(2) First 1D CNN layer: the first layer defines a filter with the convolution kernel size of 100, the step length is 1, in order to learn more data features, 100 filters are defined, multidimensional extraction is carried out, and 100 different characteristics can be obtained through training in the first layer of the network.
(3) Second 1D CNN layer: the output of the first CNN will be input into the second CNN layer. 50 different filters are defined on this network layer for training, with the remaining parameters identical to the first 1D CNN layer.
(4) Maximum pooling layer: to reduce the complexity of the output and to prevent over-fitting of the data, pooling layers are used after the CNN layer, using a 3 × 1 window, with a step size of 1.
(5) Third 1D CNN layer: in order to learn the features of higher layers, a one-dimensional convolution network is continuously used for feature extraction, the size of a convolution kernel is 50 × 1, the step size is 1, the number of the convolution kernels is 160, and Relu is used as an activation function.
(6) Fourth 1D CNN layer: the parameters are the same as the third 1D CNN layer, and feature extraction is continued.
(7) Average pooling layer: one more pooling layer was added to further avoid the occurrence of overfitting. The mean pooling takes the mean of two weights in the neural network. The size of the output matrix is 1 × 160. Each feature detector has only one weight left in this layer of the neural network.
(8) Dropout layer: here a ratio of 0.5 is chosen, the remaining parameters being the same as for the first Dropout layer.
(9) Fully connected layer activated using Softmax: since there are 2 classes to predict (i.e., "upstream" and "downstream"), the last layer will reduce the length 160 vectors to length 2 vectors. Softmax is used as the activation function. It forces the sum of 2 output values of the neural network to one, the output value will represent the probability of occurrence of each of these 2 classes, and the higher the probability is the predicted classification result.
Step S40: inputting the power test data set into the convolutional neural network model to be tested for testing, and taking the convolutional neural network model to be tested as a fault test model corresponding to the power system to be tested when a test result meets a preset test condition;
it should be noted that, after the neural network model is trained, the training set and the validation set are evaluated respectively for their effects. Further explanation is made based on the above embodiments, for example: the cross entropy loss function loss of the training set is 0.0821, and the cross entropy loss function loss of the validation set is 0.0026, both values are very low, close to 0. And the accuracy rate reaches 1. The results of these two data sets show that: the constructed one-dimensional convolutional neural network model has high prediction precision and quite excellent predictability, and can be used as a fault test model. In the training process of this embodiment, the hyper-parameter setting may be: the batch size is 10, the iteration count epochs is 20, and the learning rate learning _ rate is 0.00001.
Step S50: and judging the fault direction of the power system to be detected according to the fault test model.
Step S50 specifically includes: acquiring current operation data of a power system to be detected, and inputting the current operation data into the fault detection model; and when the output result is that a fault exists, determining the current fault position of the power system to be detected according to the fault detection model and the system simulation model.
In specific implementation, when fault detection is performed, a pre-trained fault detection model is loaded, current operation data is input into the fault detection model, and an output result is obtained. When the output result shows that the fault exists, the fault caused by the parameters of the device is analyzed according to the output result, and the fault position can be quickly determined by combining a system simulation model.
According to the method, based on the deep learning technology, the fault direction in the power system can be effectively judged by carrying out model establishment on the power system and training the detection model according to the data of the power system, and the efficiency and the accuracy can be greatly improved.
Referring to fig. 4, fig. 4 is a schematic flow chart illustrating a method for determining a fault location of an electrical power system according to a second embodiment of the present invention. Based on the first embodiment, after step S30, the method for determining a fault location of a power system according to this embodiment further includes:
step S60: and inputting the power test data set into the convolutional neural network model to be tested for testing, and performing optimization training on the convolutional neural network model to be tested when a test result does not meet a preset test condition.
Based on the first embodiment, if the cross entropy loss function loss of the training set is not close to 0, and the accuracy rate does not reach 1. Training the convolutional neural network model to be tested continuously so as to optimize the model and improve the test accuracy.
Referring to fig. 5, fig. 5 is a block diagram illustrating a configuration of a fault location determination device of an electric power system according to a first embodiment of the present invention.
As shown in fig. 5, the apparatus includes: the first model building module 10 is configured to build a system simulation model of a to-be-detected power system, and acquire a fault data set of the to-be-detected power system;
it should be noted that the system simulation model may be constructed by using a visual simulation tool Simulink to obtain system data of the power system to be detected, which includes but is not limited to: and constructing the power system to be detected by branch impedance, load of each node, interconnection switch parameters and the like. The fault data set comprises various parameters of the power system to be detected when the power system to be detected is in fault, and the various parameters can be historical fault parameters of the power system to be detected or fault parameters obtained by fault simulation through a system simulation model.
Further, the first model building module 10 is specifically configured to obtain a system type of the power system to be detected, and build a system simulation model according to the system type; acquiring a preset fault information set, and performing fault simulation on the system simulation model according to the preset fault information set; and detecting fault data generated in the fault simulation process to obtain a fault data set.
In this embodiment, the system type of the power system to be detected is an IEEE standard 33 node as an example for explanation (in a specific implementation, the method may also be used in power systems of other system types, and this embodiment does not limit an actual application scenario of the present invention). In specific implementations, for example: an IEEE33 node power distribution network model is built by using MATLAB/Simulink, and modules such as Simscape, Sources and Sinks in the Simulink are compared with a standard 33 node power distribution network to perform assembly.
Further, to obtain a failure data set, based on the above example: adding a monitoring point module between preset nodes in a simulation model of an IEEE33 node, wherein the nodes can be between 8 and 9 nodes; setting four random value model parameters of short circuit type, short circuit occurrence position, short circuit occurrence time (initial phase) and short circuit resistance value, and carrying out simulation on short circuit fault; furthermore, in order to acquire more fault data, an Asynchronous motor module of Asynchronous Machine can be added, and the motor is started statically to cause voltage sag and simulate load impact fault; the circuit data on the monitoring points are collected, and the circuit data comprise voltage, current and power data of each node, data of a branch where a fault is located and the like, and the data are added into a fault data set.
The data processing module 20 is configured to perform data preprocessing on the fault data set, and obtain an electric power training data set and an electric power test data set according to a processing result;
it is easy to understand that for effective model training, some abnormal data in the fault data need to be deleted, so the fault data set is subjected to data preprocessing. Processing the data anomaly points: during the simulation of the circuit, the recorded circuit data may be abnormal due to environmental and operational problems, and the data points are unreasonable, and if the data points are not processed, the subsequent result may be deviated.
The data processing module 20 is specifically configured to obtain abnormal data and normal data in the fault data set according to a preset numerical boundary, and perform numerical correction on the abnormal data to obtain corrected data; carrying out data normalization processing on the corrected data and the normal data to obtain a fault data set preprocessing result; and performing set division on the preprocessing result of the fault data set according to a preset set proportion to obtain an electric power training data set and an electric power test data set.
In specific implementation, the preset numerical value boundary can be implemented by box plot analysis, that is, the numerical value larger than or smaller than the upper and lower bounds set by the box plot is an abnormal value, and then the identified abnormal value is corrected by the average value of the upper and lower neighbor points of the abnormal value, so as to obtain corrected data.
Further, since the types of input parameters are different and the numerical difference is large, the main solution to such problems is to normalize the input parameter data, and the most commonly adopted method for normalizing the data is min-max normalization, which is also called dispersion normalization, and is a linear transformation on the original data, so that the result value is mapped between [0,1], and the formula is as follows:
wherein XmaxMaximum value of single-column characteristic data, XminIs the minimum value of the single column of characteristic data.
Further, the preset ratio may be 7: the implementation may include a power training data set 700 and a power test data set 300.
The second model building module 30 is configured to build an initial convolutional neural network model, and train the initial convolutional neural network model according to the power training data set to obtain a convolutional neural network model to be tested;
the second model building module 30 is specifically configured to build a one-dimensional convolutional neural network model as an initial convolutional neural network model, and input the power training data set into the initial convolutional neural network model to train according to a preset optimization target, so as to obtain a convolutional neural network model to be tested.
It should be noted that, a one-dimensional CNN convolutional neural network model is built, an input layer, a hidden layer and an output layer of the neural network are determined, the number of neurons of the input layer corresponds to input parameter characteristics, and network weights are initialized, the one-dimensional CNN convolutional neural network designed by the method has 9 layers in total, and comprises 1 input layer, 4 convolutional layers and 2 pooling layers, 1 discarding layer and 1 fully-connected layer output layer, and the output layer is a position predicted by the neural network to have a fault;
it should be noted that, referring to fig. 3, fig. 3 is a schematic diagram of a convolutional neural network according to the method for determining a fault location of a power system of the present invention, where the preset optimization target is a target of minimizing a cross entropy loss function. The method comprises the steps of inputting a training data set into a 1D CNN-based neural network model for training, using minimization of a cross entropy loss function as an optimization target, adopting an Adam optimizer to update and adjust model parameters, reducing prediction errors, using a test set for testing to obtain an ideal prediction model, storing a network model architecture and network parameters with ideal effects, and using a cross entropy loss function loss and accuracy as evaluation indexes.
Where M denotes the number of categories, yicRepresents a symbolic function (0 or 1), if the true class of sample i is equal to c, then yicGet 1, otherwise get 0, picRepresenting the predicted probability that sample i belongs to class c.
Wherein, TP represents the number of correct positive case predictions, FP represents the number of incorrect negative case predictions, TN represents the number of correct negative case predictions, and FN represents the number of incorrect positive case predictions.
It should be noted that, the specific network structure parameters are set according to actual requirements, and are not explained one by one in this embodiment. The meaning of each layer of the convolutional neural network model to be tested is as follows: (1) an input layer: one acquisition interval is 0.00001 second, only data of a fault occurrence period is acquired, and 13000 sampling points are total. After the data are preprocessed, each group of data comprises three-phase voltage values Ua, Ub and Uc and three-phase current values Ia, Ib and Ic, and in the neural network, the data are spread into 13000 multiplied by 6 vectors and then transmitted into the neural network.
(2) First 1D CNN layer: the first layer defines a filter with the convolution kernel size of 100, the step length is 1, in order to learn more data features, 100 filters are defined, multidimensional extraction is carried out, and 100 different characteristics can be obtained through training in the first layer of the network.
(3) Second 1D CNN layer: the output of the first CNN will be input into the second CNN layer. 50 different filters are defined on this network layer for training, with the remaining parameters identical to the first 1D CNN layer.
(4) Maximum pooling layer: to reduce the complexity of the output and to prevent over-fitting of the data, pooling layers are used after the CNN layer, using a 3 × 1 window, with a step size of 1.
(5) Third 1D CNN layer: in order to learn the features of higher layers, a one-dimensional convolution network is continuously used for feature extraction, the size of a convolution kernel is 50 × 1, the step size is 1, the number of the convolution kernels is 160, and Relu is used as an activation function.
(6) Fourth 1D CNN layer: the parameters are the same as the third 1D CNN layer, and feature extraction is continued.
(7) Average pooling layer: one more pooling layer was added to further avoid the occurrence of overfitting. The mean pooling takes the mean of two weights in the neural network. The size of the output matrix is 1 × 160. Each feature detector has only one weight left in this layer of the neural network.
(8) Dropout layer: here a ratio of 0.5 is chosen, the remaining parameters being the same as for the first Dropout layer.
(9) Fully connected layer activated using Softmax: since there are 2 classes to predict (i.e., "upstream" and "downstream"), the last layer will reduce the length 160 vectors to length 2 vectors. Softmax is used as the activation function. It forces the sum of 2 output values of the neural network to one, the output value will represent the probability of occurrence of each of these 2 classes, and the higher the probability is the predicted classification result.
The model testing module 40 is configured to input the power testing data set into the convolutional neural network model to be tested for testing, and when a testing result meets a preset testing condition, use the convolutional neural network model to be tested as a fault testing model corresponding to the power system to be tested;
it should be noted that, after the neural network model is trained, the training set and the validation set are evaluated respectively for their effects. Further explanation is made based on the above embodiments, for example: the cross entropy loss function loss of the training set is 0.0821, and the cross entropy loss function loss of the validation set is 0.0026, both values are very low, close to 0. And the accuracy rate reaches 1. The results of these two data sets show that: the constructed one-dimensional convolutional neural network model has high prediction precision and quite excellent predictability, and can be used as a fault test model. In the training process of this embodiment, the hyper-parameter setting may be: the batch size is 10, the iteration count epochs is 20, and the learning rate learning _ rate is 0.00001.
And the fault detection module 50 is used for judging the fault direction of the power system to be detected according to the fault test model.
The fault detection module 50 is specifically configured to acquire current operation data of the power system to be detected, and input the current operation data into the fault detection model; and when the output result is that a fault exists, determining the current fault position of the power system to be detected according to the fault detection model and the system simulation model.
In specific implementation, when fault detection is performed, a pre-trained fault detection model is loaded, current operation data is input into the fault detection model, and an output result is obtained. When the output result shows that the fault exists, the fault caused by the parameters of the device is analyzed according to the output result, and the fault position can be quickly determined by combining a system simulation model.
According to the device, based on the deep learning technology, the fault direction in the power system can be effectively judged by carrying out model establishment on the power system and training the detection model according to the data of the power system, and the efficiency and the accuracy can be greatly improved.
Furthermore, an embodiment of the present invention further provides a storage medium, where a power system fault orientation determination program is stored, and the power system fault orientation determination program is executed by a processor to perform the steps of the power system fault orientation determination method as described above.
Since the storage medium adopts all technical solutions of all the embodiments, at least all the beneficial effects brought by the technical solutions of the embodiments are achieved, and no further description is given here.
It should be understood that the above is only an example, and the technical solution of the present invention is not limited in any way, and in a specific application, a person skilled in the art may set the technical solution as needed, and the present invention is not limited thereto.
It should be noted that the above-described work flows are only exemplary, and do not limit the scope of the present invention, and in practical applications, a person skilled in the art may select some or all of them to achieve the purpose of the solution of the embodiment according to actual needs, and the present invention is not limited herein.
In addition, the technical details that are not described in detail in this embodiment may be referred to a method for determining a fault location of an electrical power system provided in any embodiment of the present invention, and are not described herein again.
Further, it is to be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention or portions thereof that contribute to the prior art may be embodied in the form of a software product, where the computer software product is stored in a storage medium (e.g. Read Only Memory (ROM)/RAM, magnetic disk, optical disk), and includes several instructions for enabling a terminal device (e.g. a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A method for judging fault location of a power system is characterized by comprising the following steps:
constructing a system simulation model of a power system to be detected, and acquiring a fault data set of the power system to be detected;
performing data preprocessing on the fault data set, and acquiring a power training data set and a power test data set according to a processing result;
constructing an initial convolutional neural network model, and training the initial convolutional neural network model according to the electric power training data set to obtain a convolutional neural network model to be tested;
inputting the power test data set into the convolutional neural network model to be tested for testing, and taking the convolutional neural network model to be tested as a fault test model corresponding to the power system to be tested when a test result meets a preset test condition;
and judging the fault direction of the power system to be detected according to the fault test model.
2. The method for distinguishing fault orientations of an electric power system according to claim 1, wherein the step of constructing a system simulation model of the electric power system to be detected and acquiring the fault data set of the electric power system to be detected specifically comprises:
the method comprises the steps of obtaining a system type of a power system to be detected, and establishing a system simulation model according to the system type;
acquiring a preset fault information set, and performing fault simulation on the system simulation model according to the preset fault information set;
and detecting fault data generated in the fault simulation process to obtain a fault data set.
3. The method for determining a fault location of an electrical power system according to claim 2, wherein the step of preprocessing the fault data set and obtaining an electrical training data set and an electrical testing data set according to the processing result specifically includes:
acquiring abnormal data and normal data in the fault data set according to a preset numerical boundary, and performing numerical correction on the abnormal data to acquire corrected data;
carrying out data normalization processing on the corrected data and the normal data to obtain a fault data set preprocessing result;
and performing set division on the preprocessing result of the fault data set according to a preset set proportion to obtain an electric power training data set and an electric power test data set.
4. The method for judging the fault location of the power system as claimed in claim 3, wherein the step of constructing an initial convolutional neural network model and training the initial convolutional neural network model according to the power training data set to obtain the convolutional neural network model to be tested specifically comprises:
and constructing a one-dimensional convolutional neural network model as an initial convolutional neural network model, inputting the electric power training data set into the initial convolutional neural network model, and training according to a preset optimization target to obtain the convolutional neural network model to be tested.
5. The method according to claim 4, wherein the predetermined optimization objective is to minimize a cross entropy loss function.
6. The method for judging the fault location of the power system according to claim 5, wherein the step of judging the fault location of the power system to be detected according to the fault test model specifically comprises:
acquiring current operation data of a power system to be detected, and inputting the current operation data into the fault detection model;
and when the output result is that a fault exists, determining the current fault position of the power system to be detected according to the fault detection model and the system simulation model.
7. The method for determining a fault location of a power system according to claim 6, wherein after the steps of constructing an initial convolutional neural network model and training the initial convolutional neural network model according to the power training data set to obtain a convolutional neural network model to be tested, the method further comprises:
and inputting the power test data set into the convolutional neural network model to be tested for testing, and performing optimization training on the convolutional neural network model to be tested when a test result does not meet a preset test condition.
8. An apparatus for determining a fault location of an electrical power system, the apparatus comprising:
the first model building module is used for building a system simulation model of the power system to be detected and acquiring a fault data set of the power system to be detected;
the data processing module is used for carrying out data preprocessing on the fault data set and acquiring an electric power training data set and an electric power test data set according to a processing result;
the second model building module is used for building an initial convolutional neural network model and training the initial convolutional neural network model according to the electric power training data set so as to obtain a convolutional neural network model to be tested;
the model testing module is used for inputting the power testing data set into the convolutional neural network model to be tested for testing, and when a testing result meets a preset testing condition, the convolutional neural network model to be tested is used as a fault testing model corresponding to the power system to be tested;
and the fault detection module is used for judging the fault direction of the power system to be detected according to the fault test model.
9. A terminal device, characterized in that the terminal device comprises: a memory, a processor and a power system fault orientation discrimination program stored on the memory and operable on the processor, the power system fault orientation discrimination program being configured to implement the steps of the power system fault orientation discrimination method of any one of claims 1 to 7.
10. A storage medium having stored thereon a power system fault orientation discrimination program which, when executed by a processor, implements the steps of the power system fault orientation discrimination method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111172769.8A CN113902946A (en) | 2021-10-08 | 2021-10-08 | Power system fault direction judging method and device, terminal equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111172769.8A CN113902946A (en) | 2021-10-08 | 2021-10-08 | Power system fault direction judging method and device, terminal equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113902946A true CN113902946A (en) | 2022-01-07 |
Family
ID=79190448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111172769.8A Pending CN113902946A (en) | 2021-10-08 | 2021-10-08 | Power system fault direction judging method and device, terminal equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113902946A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114742108A (en) * | 2022-04-20 | 2022-07-12 | 中科航迈数控软件(深圳)有限公司 | Method and system for detecting fault of bearing of numerical control machine tool |
| CN114861880A (en) * | 2022-05-06 | 2022-08-05 | 清华大学 | Industrial equipment fault prediction method and device based on cavity convolutional neural network |
| CN117200203A (en) * | 2023-09-07 | 2023-12-08 | 航电所(成都)科技有限公司 | Operation optimization method and system applied to power system |
| CN117584136A (en) * | 2024-01-18 | 2024-02-23 | 泰山学院 | Robot fault detection method and system based on artificial intelligence |
-
2021
- 2021-10-08 CN CN202111172769.8A patent/CN113902946A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114742108A (en) * | 2022-04-20 | 2022-07-12 | 中科航迈数控软件(深圳)有限公司 | Method and system for detecting fault of bearing of numerical control machine tool |
| CN114742108B (en) * | 2022-04-20 | 2022-12-20 | 中科航迈数控软件(深圳)有限公司 | Method and system for detecting fault of bearing of numerical control machine tool |
| CN114861880A (en) * | 2022-05-06 | 2022-08-05 | 清华大学 | Industrial equipment fault prediction method and device based on cavity convolutional neural network |
| CN114861880B (en) * | 2022-05-06 | 2024-04-12 | 清华大学 | Industrial equipment fault prediction method and device based on cavity convolutional neural network |
| CN117200203A (en) * | 2023-09-07 | 2023-12-08 | 航电所(成都)科技有限公司 | Operation optimization method and system applied to power system |
| CN117584136A (en) * | 2024-01-18 | 2024-02-23 | 泰山学院 | Robot fault detection method and system based on artificial intelligence |
| CN117584136B (en) * | 2024-01-18 | 2024-03-29 | 泰山学院 | Robot fault detection method and system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111027058B (en) | Method for detecting data attack of power system, computer equipment and storage medium | |
| Appiah et al. | Long short-term memory networks based automatic feature extraction for photovoltaic array fault diagnosis | |
| CN109635928B (en) | Voltage sag reason identification method based on deep learning model fusion | |
| CN113902946A (en) | Power system fault direction judging method and device, terminal equipment and storage medium | |
| Tong et al. | Detection and classification of transmission line transient faults based on graph convolutional neural network | |
| CN111368890A (en) | Fault detection method and device and information physical fusion system | |
| Mohagheghi et al. | Condition monitoring of power electronic circuits using artificial neural networks | |
| CN111833583B (en) | Training method, device, equipment and medium for power data anomaly detection model | |
| Guo et al. | A novel radial basis function neural network principal component analysis scheme for PMU-based wide-area power system monitoring | |
| CN114386537A (en) | Lithium battery fault diagnosis method and device based on Catboost and electronic equipment | |
| Bayrak et al. | Deep learning-based multi-model ensemble method for classification of PQDs in a hydrogen energy-based microgrid using modified weighted majority algorithm | |
| CN111738348A (en) | Power data anomaly detection method and device | |
| CN117394529A (en) | SCADA-based auxiliary decision method and system for main distribution network loop-closing reverse power supply control conditions | |
| Chen et al. | Real‐time recognition of power quality disturbance‐based deep belief network using embedded parallel computing platform | |
| CN116720095A (en) | Electrical characteristic signal clustering method for optimizing fuzzy C-means based on genetic algorithm | |
| CN111881159A (en) | Fault detection method and device based on cost-sensitive extreme random forest | |
| CN114740730A (en) | SVG parameter optimization identification method based on convolutional neural network | |
| Bhuiyan et al. | A deep learning through DBN enabled transmission line fault transient classification framework for multimachine microgrid systems | |
| Firos et al. | Fault detection in power transmission lines using AI model | |
| Zhu et al. | Structure-Aware Recurrent Learning Machine for Short-Term Voltage Trajectory Sensitivity Prediction | |
| CN115936926A (en) | SMOTE-GBDT-based unbalanced electricity stealing data classification method and device, computer equipment and storage medium | |
| CN113435575B (en) | Gate graph neural network transient stability evaluation method based on unbalanced data | |
| CN115865421A (en) | Intrusion detection method and system for power distribution network information system | |
| CN115275975A (en) | Method and device for determining electric power data matching degree of optical storage charging station | |
| CN110542851A (en) | Fault diagnosis method and device for circuit breaker operating mechanism, computer and storage medium |
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 |