CN115580526B - Communication network fault diagnosis method, system, electronic equipment and storage medium - Google Patents
Communication network fault diagnosis method, system, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115580526B CN115580526B CN202211209903.1A CN202211209903A CN115580526B CN 115580526 B CN115580526 B CN 115580526B CN 202211209903 A CN202211209903 A CN 202211209903A CN 115580526 B CN115580526 B CN 115580526B
- Authority
- CN
- China
- Prior art keywords
- clustering
- communication network
- density threshold
- data
- electronic equipment
- 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
- 238000004891 communication Methods 0.000 title claims abstract description 100
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000003745 diagnosis Methods 0.000 title claims abstract description 24
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 103
- 238000013528 artificial neural network Methods 0.000 claims abstract description 56
- 238000003064 k means clustering Methods 0.000 claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 24
- 238000012544 monitoring process Methods 0.000 claims abstract description 10
- 238000012360 testing method Methods 0.000 claims abstract description 8
- 238000009826 distribution Methods 0.000 claims description 22
- 230000004913 activation Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 230000007246 mechanism Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 11
- 230000006870 function Effects 0.000 description 60
- 238000004364 calculation method Methods 0.000 description 11
- 238000004458 analytical method Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 210000002569 neuron Anatomy 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000010606 normalization Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 238000007621 cluster analysis Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000010365 information processing Effects 0.000 description 3
- 238000011524 similarity measure Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006386 memory function Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000004540 process dynamic Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0677—Localisation of faults
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/142—Network analysis or design using statistical or mathematical methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Mathematical Optimization (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Medical Informatics (AREA)
- Mathematical Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pure & Applied Mathematics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Algebra (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention relates to a communication network fault diagnosis method, a system, electronic equipment and a storage medium, and relates to the field of electronic equipment communication, wherein the method comprises the following steps: acquiring historical electronic equipment communication network state data; clustering the historical electronic equipment communication network state data by using an improved K-means clustering algorithm to obtain a clustering result; determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model; and predicting the communication network state data of the real-time monitoring electronic equipment as a test set by using a fault prediction model to obtain a fault diagnosis model. The invention solves the problems of difficult fault positioning and difficult prediction of the electronic equipment communication network in the guarantee process.
Description
Technical Field
The present invention relates to the field of electronic equipment communications, and in particular, to a method, a system, an electronic device, and a storage medium for diagnosing a communication network failure.
Background
The electronic equipment communication network is complex in composition, information coupling relation among all components is tight, fault diagnosis and prediction are carried out on the electronic equipment communication network, and the electronic equipment communication network relates to a plurality of technologies such as computer technology, information processing technology and communication technology.
Disclosure of Invention
The invention aims to provide a communication network fault diagnosis method, a system, electronic equipment and a storage medium, which solve the problems of difficult fault positioning and difficult prediction of an electronic equipment communication network in the guarantee process.
In order to achieve the above object, the present invention provides the following solutions:
a communication network fault diagnosis method, comprising:
acquiring historical electronic equipment communication network state data;
clustering the historical electronic equipment communication network state data by using an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the clustering number;
determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model;
and predicting the communication network state data of the real-time monitoring electronic equipment by using the fault prediction model as a test set to obtain a fault diagnosis model.
Optionally, the improved K-means clustering algorithm utilizes a density partitioning algorithm to determine an initial clustering center, and specifically includes:
calculating density threshold values under different clustering number values by using the density threshold value coefficient and Euclidean distance;
and determining an initial clustering center according to the historical electronic equipment communication network state data and the density threshold value.
Optionally, the improved K-means clustering algorithm determines the number of clusters by using a K-value-SSE broken line image algorithm, and specifically comprises the following steps:
calculating error square sums under different clustering numbers according to the historical electronic equipment communication network state data;
determining a k value-SSE line graph according to the cluster number and the error square sum;
and determining the number of clusters according to the inflection point of the k value-SSE line graph.
Optionally, the activation function of the Elman neural network is a hyperbolic tangent S-shaped function, and the expression of the hyperbolic tangent S-shaped function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
The invention also provides a communication network fault diagnosis system, which comprises:
the acquisition module is used for acquiring historical electronic equipment communication network state data;
the clustering module is used for clustering the historical electronic equipment communication network state data by utilizing an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the clustering number;
the training module is used for determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model;
and the prediction module is used for predicting the communication network state data of the real-time monitoring electronic equipment as a test set by using the fault prediction model to obtain a fault diagnosis model.
Optionally, the initial cluster center determining sub-module in the cluster module specifically includes:
the density threshold determining unit is used for calculating density thresholds under different cluster number values by using the density threshold coefficient and the Euclidean distance;
and the initial cluster center determining unit is used for determining an initial cluster center according to the historical electronic equipment communication network state data and the density threshold value.
Optionally, the clustering number determining word module in the clustering module specifically includes:
the error square sum determining unit is used for calculating error square sums under different clustering numbers according to the historical electronic equipment communication network state data;
a k-value-SSE line graph determining unit configured to determine a k-value-SSE line graph from the number of clusters and the sum of squares of errors;
and the cluster number determining unit is used for determining the cluster number according to the inflection point of the k value-SSE line graph.
Optionally, the activation function of the Elman neural network is a hyperbolic tangent S-shaped function, and the expression of the hyperbolic tangent S-shaped function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
The present invention also provides an electronic device including:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of the preceding claims.
The invention also provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a method as described in any of the above.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the method comprises the steps of acquiring historical electronic equipment communication network state data; clustering the historical electronic equipment communication network state data by using an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the clustering number; determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model; and predicting the communication network state data of the real-time monitoring electronic equipment by using the fault prediction model as a test set to obtain a fault diagnosis model. According to the invention, the communication network state data is analyzed by using a clustering algorithm, the traditional K-means algorithm is optimized based on an improved K-means clustering algorithm, the fault diagnosis of the electronic equipment communication network is completed, the fault prediction is performed on the electronic equipment communication network by using an Elman neural network algorithm, the problem of difficult selection of the Elman neural network structure is solved, and the structure of the Elman neural network hidden layer is determined by using a fitting error analysis method, so that the problems of difficult fault positioning and difficult prediction of the electronic equipment communication network in the guarantee process are solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a communication network fault diagnosis method provided by the invention;
FIG. 2 is a schematic diagram of the operation of the electronic equipment;
FIG. 3 is a schematic diagram of a conventional Elman neural network;
FIG. 4 is a flowchart of an Elman neural network prediction;
FIG. 5 is a schematic view of an experimental environment;
FIG. 6 is a k-value SSE line graph;
FIG. 7 is a schematic diagram of a clustering result;
FIG. 8 is a graph comparing clustering results;
FIG. 9 is a schematic diagram of a predicted outcome;
FIG. 10 is a comparison graph of prediction errors;
FIG. 11 is a schematic view of a sample area;
FIG. 12 is a schematic diagram of a threshold range when k is 1;
FIG. 13 is a schematic diagram of a threshold range when k is 2;
FIG. 14 is a schematic view of the threshold range when k is 3;
FIG. 15 is a schematic view of the threshold range when k is 4;
fig. 16 is a schematic view of the threshold range when k is 5.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a communication network fault diagnosis method, a system, electronic equipment and a storage medium, which solve the problems of difficult fault positioning and difficult prediction of an electronic equipment communication network in the guarantee process.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
As shown in fig. 1, the method for diagnosing a communication network fault provided by the present invention includes:
step 101: historical electronic equipment communication network status data is obtained.
Through the research on the operation mechanism of the electronic equipment communication network, as shown in fig. 2, each electronic equipment node comprises three subsystems, namely an information processing subsystem, a network control subsystem and a communication subsystem. The information processing subsystem comprises a network switch, a server and a client. The network control subsystem comprises network control equipment and network safety protection equipment. The communication subsystem comprises various wireless communication devices, wired communication devices, junction boxes and the like. The invention collects the state data of the data link layer of the communication network in the network control subsystem and the communication subsystem, uses a clustering algorithm to analyze and cluster the state data generated by the communication network so as to achieve the purposes of finding and rapidly positioning the faults, and predicts the equipment faults by using an Elman neural network algorithm so as to reduce the adverse effect of the equipment burst faults on the equipment operation.
Step 102: clustering the historical electronic equipment communication network state data by using an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K-value-SSE polyline image algorithm to determine the number of clusters.
The improved K-means clustering algorithm utilizes a density division algorithm to determine an initial clustering center, and specifically comprises the following steps:
and calculating the density threshold under different cluster number values by using the density threshold coefficient and the Euclidean distance.
And determining an initial clustering center according to the historical electronic equipment communication network state data and the density threshold value.
The improved K-means clustering algorithm utilizes a K value-error square sum SSE (Square Sum of Error, SSE) broken line image algorithm to determine the number of clusters, and specifically comprises the following steps:
and calculating the error square sum under different clustering numbers according to the historical electronic equipment communication network state data.
And determining a k value-SSE line graph according to the cluster number and the error square sum.
And determining the number of clusters according to the inflection point of the k value-SSE line graph.
Clustering is the process of assigning objects in a dataset to different clusters. Data objects within the same cluster are similar to each other with a higher degree of similarity, while data objects not within the same cluster are different from each other with a lower degree of similarity. The degree of similarity between data objects is determined according to the value of the description attribute of the data objects, and the degree of similarity between data objects is generally described by the distance between the data objects, where a large distance between data objects indicates that the data objects are different from each other, and vice versa. The basic guiding thought of clustering is to maximally realize that data objects in classes have the largest similarity with each other and data objects in classes have the smallest similarity with each other.
The K-means clustering algorithm belongs to a dynamic clustering algorithm, which is also called a step-by-step clustering method, and before the clustering algorithm iterates, the algorithm firstly randomly and sequentially selects K data objects from a data set as K initial clustering centers, then sequentially divides other data objects into the class where the closest clustering center is located, after the data objects are divided, calculates the central average value of each cluster as a new clustering center point, and iterates the clustering process. Until no more changes occur in the cluster center, i.e. the cluster criterion function values converge or the cluster criterion function successive values differ by less than a given threshold.
Aiming at the problems that the clustering result falls into a local optimal solution due to random selection of an initial clustering center and the clustering result is greatly influenced by the initial clustering center in the traditional K-means algorithm, the traditional K-means algorithm is improved by adopting the modes of state data normalization processing, similarity measurement formula and clustering criterion function selection, initial clustering center determination based on a density division algorithm and the like in combination with the characteristics of electronic equipment communication network state data.
(1) Normalization processing
Because each object contains 3 attributes, and the value range of each attribute is different, the data is normalized before cluster analysis, and the normalization has the following advantages: 1) After normalization, the speed of gradient descent to obtain the optimal solution in the clustering analysis process is increased; 2) Normalization can improve the accuracy of cluster analysis.
Using a m x n matrix to represent the running state data of the equipment, wherein the number of rows of the matrix is m, each sample is represented by m attributes, and m=3 according to the characteristics of the running state data of the equipment; the column number of the matrix is n, representing a total of n samples. Normalizing the data should also be normalizing the data of the same line. The normalization algorithm adopted by the invention is as follows:
wherein x is the number in the matrix to be normalized, y is the normalized number, x max Is the maximum value of the row where the matrix corresponding number x is located, x min Is the minimum value of the row in which the matrix corresponds to the number x.
(2) Similarity measurement formula selection
Let data set x= { X 1 ,x 2 ,…,x n There are two data objects x in } i And x j The method is specifically described as follows: x is x i ＝{x i1 ,x i2 ,…,x ir Sum x j ＝{x j1 ,x j2 ,…,x jr All of them are composed of r attribute values, which require similarity calculation for the two data objects before clustering, typically using d (x) i ，x j ) I.e. the distance between objects, determines the degree of similarity between the two data objects, generally d (x i ，x j ) The larger the description data object x i And x j The larger the difference, the smaller the opposite. In a clustering algorithm, the distance function used to calculate the distance between data objects generally meets the following requirements: non-negativity, symmetry, triangle inequality. Non-negativity is generally the similarity value (distance) d (x) between data objects i ，x j ) Symmetry between data objects is that symmetry requirements are satisfied between data objects, i.e., d (x) i ，x j )＝d(x j ，x i ) The final distance function needs to satisfy the requirement of triangle inequality, namely, satisfy the property that the sum of two sides is larger than the third side, expressed by a formula: d (x) i ，x j )<d(x i ，x k )+d(x k ，x j ). Based on the characteristics of the state data of the communication network of the electronic equipment, the Euclidean distance is mainly used as a state data similarity measurement formula, and the Euclidean distance calculation formula is as follows:
d(x i ，x j )＝(|x i1 -x j1 | 2 +|x i2 -x j2 | 2 +…+|x ir -x jr | 2 ) 1/2
(3) Clustering criterion function selection
The clustering algorithm is the key research content of the clustering analysis, wherein the similarity measure is the basis of the clustering algorithm, the similarity measure function is used for determining the dissimilarity between objects in the data set, but the similarity measure of the clustering algorithm is insufficient, a criterion function is also needed for evaluating the quality of the clustering result, the quality of the clustering result is influenced by the quality of the clustering criterion function, and the good clustering criterion function can often obtain a relatively correct clustering result. Currently, the common clustering functions are: based on the characteristics of the state data of the communication network of the electronic equipment, the invention mainly adopts the error square sum criterion function as a clustering criterion function for clustering analysis of the state data, and the calculation process of the error square sum criterion function is as follows:
let x= { X for data set X containing n data 1 ,x 2 ,…,x n K clusters are obtained after clustering, and are expressed as: w (W) 1 ,W 2 ,…,W k Wherein the number of data objects in each cluster is n 1 ,n 2 ,…,n k The method comprises the following steps: n is n 1 +n 2 +…+n k =n. Let m j Representing the j-th cluster W j The average value of all the objects in the system is similar to the running state data object of each device, m j Consisting of r attributes, then m j The h attribute calculation method is as follows:
the error sum of squares criterion function is defined as:
error square sum criterion function J c The value of (2) can be described as the sum of squares of errors of the data objects in all clusters and the cluster centers in the class in which they are located, according to their mathematical formulas, it is apparent that if the accuracy of one cluster result is relatively high, the phase between the data objects of each cluster in the cluster resultThe similarity is higher, because the data objects all have higher similarity with the clustering center, the value of the clustering criterion function, namely the value of the error square sum function, is smaller, and if the accuracy of the clustering result is lower, the similarity between the data objects of each cluster in the clustering result is lower, and because the data objects all have lower similarity with the clustering center, the value of the clustering criterion function, namely the value of the error square sum function, is larger. From the above analysis, the following can be concluded: the goal of clustering is to assign objects in the dataset to different classes, but it is also to some extent to find the clustering result with the smallest clustering criterion function value.
(4) Determining an initial cluster center based on a density partitioning algorithm
Inspired by the density-based clustering algorithm idea, the selection of the initial clustering center of the K-means clustering algorithm is optimized, and the selection result is close to the optimal solution as much as possible. In choosing the initial cluster center, the density threshold is used as a measure of distance to replace the traditional Euclidean distance measure. When each sample point is judged to be taken as the center, the total number of the sample points in a certain spherical space of the point is calculated through a defined density threshold value, and the total number is marked as the density of the sample points. The method comprises the steps of selecting a sample point with the maximum density as a first initial clustering center, and selecting a point with the maximum Euclidean distance with the first sample point as a second initial clustering center, wherein the purpose of the method is to avoid that the selected second initial clustering center is similar or too similar to the first clustering center.
Let data set x= { X 1 ,x 2 ,…,x n There are two data objects x in } i And x j The method is specifically described as follows: x is x i ＝{x i1 ,x i2 ,…,x ir Sum x j ＝{x j1 ,x j2 ,…,x jr The density threshold formula is:
wherein alpha is defined as densityAnd the threshold coefficient is used for reserving the density of the core area when the sample point is used as the center of the cluster for discrimination, and the density of the core area when the corresponding density threshold value is used for reserving the point as the center has more references, and k is the number of clusters. Density threshold y z Is defined as the maximum value of Euclidean distance between any two points of a sample point set multiplied by a coefficientIs a value of (2).
The selection of the density threshold coefficient has a decisive influence on the classification effect of the algorithm, so that the selection of the alpha value is optimized, the alpha value is suitable for the data distribution type of the experiment, and the clustering result is important to approach to the optimal solution. The thinking of selecting the density threshold coefficient alpha according to the characteristics of experimental data is as follows: according to the concept of normal distribution, a standard deviation range is used as a 'core region' of data, and the distribution of the corresponding 'core region' is selected according to different clustering numbers (k values), so that a density threshold coefficient alpha is obtained through calculation.
According to the experimental pretreatment result, the classification number k of the experimental data is less than 5, so that the density threshold coefficient alpha of k=1-5 is deduced to be selected.
The distance between any two points of the data sample is calculated to obtain a circular area with the maximum distance as the diameter, and the circular area can encompass substantially all sample points, as shown in fig. 11.
Let k=1 be the diameter at this maximum distance, which is the circle a, which is the range of all data distributions. According to the normal distribution, a numerical range of a standard deviation of 68.27% is taken as a core area of experimental data under the condition of k=1, a circle B is taken as a range of the core area, and the radius ratio of the circle A to the circle B is R: r=1:0.6827. As in fig. 12.
In FIG. 12, the density threshold y z Is the radius of circle B (single class core data range circle), according to the density threshold formulaThe value of the density threshold coefficient α is α= 2.9295 in the case of k=1, which is obtained from the geometric relationship.
Assuming that when k=2, the distribution ranges of the two types of data in the data distribution range a are represented by circles B and C, the numerical range of one standard deviation is 68.27% as the core region of the classified data, and the core region ranges of circles B and C are circles D and E. As shown in fig. 13. The value of the density threshold coefficient α is a= 2.9295 when k=2 is obtained from the geometric relationship according to the density threshold formula.
Similarly, when k=3, the classification distribution is as shown in fig. 14. When k=3 is calculated, the density threshold coefficient α= 2.1041.
When k=4, the classification distribution is as shown in fig. 15. When k=4 is calculated, the density threshold coefficient α= 1.7682.
When k=5, the classification distribution is as shown in fig. 16. When k=5 is calculated, the density threshold coefficient α=1.582.
In summary, when k=1 to 5, the values of the experimental data α are shown in table 1.
TABLE 1 Density threshold coefficient value Table
| k | α |
| 1 | 2.9295 |
| 2 | 2.9295 |
| 3 | 2.1041 |
| 4 | 1.7682 |
| 5 | 1.582 |
In the calculation of K mean value, density threshold values y with different sizes are obtained under the values of different K z Calculating the corresponding density threshold y of each point z After the density of the core area of the range is obtained, the point with the maximum corresponding density is selected as a first initial clustering center, euclidean distance between the point and other points is calculated, the point with the maximum distance is selected as a second initial clustering center, a third clustering center is selected as the point with the maximum Euclidean distance between the point and the first initial clustering center and the second initial clustering center, the selection of a fourth point and a fifth point is similar, and the like until k initial clustering centers meeting the condition are obtained. The improved accuracy of the clustering result of the data can be effectively observed.
(5) Determination of the number of clusters k
Aiming at another defect of the traditional K-means algorithm, namely that the K value of the clustering number cannot be determined, the K value is determined by a K value-SSE broken line image method. The specific method comprises the following steps: first, the error square sum J under different k values is calculated c The k value-SSE line graph is made, and the proper value of k is determined by searching the inflection point in the graph, because the SSE value is inevitably reduced along with the gradual increase of the k value, and the k value of a smooth part (also called an elbow part and a turning part in the middle of the gradual decrease of the curve) in the image can maximally reach a balance between the SSE and the k value.
(6) Improved algorithm flow
According to the characteristics of the state data of the communication network of the electronic equipment, the selection range of the k value of the clustering number is set to be 2-7, the initial clustering center is firstly determined through a density method aiming at each k value, then clustering processing is carried out based on a traditional k-means iterative process, and finally the square sum J of errors under the k value is calculated c . And finally, drawing a k value-SSE line graph, and determining a proper k value by searching an inflection point.
For each value of k, the specific iteration steps are as follows:
1. sequentially calculating electricityThe density of all sample points in the sub-equipment communication network state data set is selected as a first initial clustering center, and the sample point with the maximum density is marked as p 1 。
2. Sequentially calculating other sample points and p 1 Selecting the sample point with the largest distance as a second initial clustering center p 2 。
3. Sequentially calculating other sample points and p 1 Is the Euclidean distance d (x) n ,x 1 ) And p 2 Is the Euclidean distance d (x) n ,x 2 ) Selecting d (x n ,x 1 )+d(x n ,x 2 ) The largest data point of (2) is taken as the third initial cluster center and is marked as p 3 。
4. And so on,obtaining corresponding k initial cluster centers p through calculation k 。
5. And searching a clustering result of the clustering algorithm by using an iterative process in the traditional k-means clustering algorithm, and dividing the data objects into clusters represented by the nearest clustering centers one by one according to the nearest neighbor principle.
6. Respectively calculating the average value of all data objects in each cluster as a new center of each cluster, comparing the updated cluster center with the original cluster center, returning to the step 5 if the cluster center is changed, considering that the cluster center is selected to be ended when the position of the cluster center is not changed any more, and calculating the criterion function minimum error square sum J according to the calculated cluster center c 。
Step 103: and determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model.
Wherein, the activation function of the Elman neural network is a hyperbolic tangent Sigmoid function (Tan-Sigmoid function), and the expression of the hyperbolic tangent Sigmoid function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
Step 104: and predicting the communication network state data of the real-time monitoring electronic equipment by using the fault prediction model as a test set to obtain a fault diagnosis model.
The artificial neural network can perform associative memory and large-scale parallel processing, and can be divided into a feedforward neural network and a feedback neural network according to the flow direction of information in the network. Elman neural network is a relatively common feedback neural network, which introduces a receiving layer on the hidden layer of the feedforward network, belonging to the internal delay network. Elman neural networks can process dynamic information and reflect the system dynamic process.
The network topology of Elman neural network is composed of four layers: the first layer is an input layer and comprises a plurality of input nodes, so that the function of transmitting signals is achieved; the second layer is an hidden layer and comprises a plurality of hidden layer neurons, and receives input from an input layer node and feedback input of a receiving layer node; the third layer is a receiving layer, and the receiving layer unit respectively memorizes and stores the output value of the hidden layer neuron corresponding to the receiving layer unit at the previous moment and delays to feed back to the hidden layer neuron. The fourth layer is the output layer, and the neurons of the output layer play a role in linear weighting. The time delay feedback function of the receiving layer neurons enables the Elman neural network to have stronger sensitivity and dynamic memory function on historical data. The structure of the Elman neural network is shown in fig. 3.
The invention determines the structure of the Elman neural network by a fitting error analysis method, wherein the fitting error analysis method determines the thought of the structure of the Elman neural network to learn and train through training set data, acquires dynamic characteristics between input and output parameters, adopts an error correction learning rule, dynamically adjusts structural parameters of each layer, and finally acquires stable network parameters, as shown in fig. 4, and comprises the following specific processes:
(1) Training data through a training set, receiving data input through an input layer, processing through an implicit layer, and finally forward transmitting an input signal in a mode of outputting a result from an output layer;
(2) And calculating the error between the real output result and the expected output result of the output layer, if the difference is too large and exceeds an acceptable range, entering an error back propagation link, and reversely propagating and distributing error signals layer by layer to neurons of each layer in a certain form, thereby updating and correcting the weights and threshold matrixes of the neurons.
(3) The objective of network learning is to find a weight matrix that minimizes the objective function, at which time error correction learning translates into a typical optimization problem, which often uses gradient descent-based learning algorithms.
In the execution process of the fitting error analysis method, an activation function is required to be designated for an Elman neural network hidden layer, a continuous nonlinear function is selected, and can be solved by an optimization method because the continuous nonlinear activation function is conductive, and the specific function selected here is a hyperbolic tangent S-shaped function (hyperbolic tangent sigmoid transfer function, tansig), and the Tansig function expression is as follows:
the input data of the Elman neural network is also the status data of the communication network. The data clustering mainly completes fault diagnosis, namely communication faults exist, and the faults are specifically positioned through clustering; the prediction is that no fault is present, but the state of the communication network in the next step is predicted according to the trend of the current state data, and the fault is predicted.
As shown in fig. 5, the whole experimental environment is composed of a server, a client, a switching unit, a network control device, a wireless communication device, a software radio platform, a load generator and a main control unit. The server, the client, the switching unit, the network control device and the wireless communication device construct an equipment link, the software radio platform and the main control unit simulate communication interference and communication faults of various wireless communication links, and the load generator is used for simulating various data services. The invention collects the status data of the communication network of the electronic equipment to carry out fault diagnosis and prediction, and carries out big data analysis by taking three indexes of the utilization rate, the number of data packets sent and the number of data packets received in the communication network of the electronic equipment as judgment bases.
Through density-based partitional clustering algorithm calculation, iteration is carried out on the k value, and the minimum error square sum J corresponding to the k value from 2 to 7 is obtained through calculation c The k-SSE line graph is obtained as follows in FIG. 6:
as can be seen from fig. 6, the k value is 3 through inflection points by using the image method, and k=3 is calculated by using a density-based clustering algorithm to obtain a clustering result as shown in fig. 7:
the 387 monitored data points were divided into three categories of states by clustering, wherein the pentagram represented the normal state, the square represented the high load operating state and the inverted triangle represented the fault state. Through the processing, the fault type can be conveniently and rapidly judged when the equipment generates sudden faults, the fault position can be positioned, and the faults can be solved in the shortest time.
The following results are compared with conventional partition-based cluster analysis. The clustering result obtained by calculation based on the clustering algorithm of the invention shown in (a) of fig. 8 is compared with the clustering result obtained by the conventional clustering algorithm based on division shown in (b) of fig. 8, and it is obvious that the clustering algorithm of the invention based on the improved algorithm is more uniform in classification and dispersion of sample points, and the result is more reasonable and reliable. Because the initial point of the traditional K mean value is selected randomly, each operation may cause deviation of operation results, the result is trapped in a local optimal solution, and the result obtained by the density-based clustering algorithm is stable.
Training an Elman neural network algorithm according to sample points with known monitoring data by using the Elman neural network, and then predicting sample indexes of the next time point.
After training, the predicted result of the operation is shown in fig. 9, wherein the dotted line is the predicted result, and the solid line is the actual result. A comparison graph of prediction errors caused by setting different numbers of hidden layer neurons is shown in fig. 10.
Through an Elman neural network algorithm, equipment parameters at future time points can be predicted, and the influence of burst faults of battlefield equipment on equipment operation can be reduced effectively. Meanwhile, the more the monitoring data are input, the more the neural network is trained, and the more accurate the equipment parameter prediction is performed at the future time point.
The invention focuses on researching a K-means clustering algorithm and a neural network algorithm. The K-means clustering algorithm is one of the earliest and most widely applied clustering analysis algorithms, and has the advantages of simple structure, easy realization, strong local searching capability, suitability for processing large data sets and the like. The work of the invention mainly comprises the following three aspects:
(1) An improved K-means algorithm, namely a density-based partitional clustering algorithm, is proposed. The algorithm effectively solves the problem of initial points of different k values in the clustering algorithm based on division. A new measurement formula, namely a density threshold formula, is also provided.
(2) The method completes the calculation of the criterion function Jc under different K values by using a computer, selects proper K values by an image method, and solves the problem of uncertainty in the traditional K-means K value selection.
(3) And by utilizing an Elman neural network, the data condition of future points is predicted through the study of the monitored data, so that the pre-discovery of faults is realized.
The invention also provides a communication network fault diagnosis system, which comprises:
and the acquisition module is used for acquiring the historical electronic equipment communication network state data.
The clustering module is used for clustering the historical electronic equipment communication network state data by utilizing an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K-value-SSE polyline image algorithm to determine the number of clusters.
And the training module is used for determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model.
And the prediction module is used for predicting the communication network state data of the real-time monitoring electronic equipment as a test set by using the fault prediction model to obtain a fault diagnosis model.
In practical application, the initial cluster center determining sub-module in the cluster module specifically includes:
and the density threshold determining unit is used for calculating the density threshold under different cluster number values by using the density threshold coefficient and the Euclidean distance.
And the initial cluster center determining unit is used for determining an initial cluster center according to the historical electronic equipment communication network state data and the density threshold value.
In practical application, the clustering number determining word module in the clustering module specifically comprises:
and the error square sum determining unit is used for calculating error square sums under different clustering numbers according to the historical electronic equipment communication network state data.
And the k value-SSE line diagram determining unit is used for determining a k value-SSE line diagram according to the cluster number and the error square sum.
And the cluster number determining unit is used for determining the cluster number according to the inflection point of the k value-SSE line graph.
In practical application, the activation function of the Elman neural network is a hyperbolic tangent S-shaped function, and the expression of the hyperbolic tangent S-shaped function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
The present invention also provides an electronic device including:
one or more processors.
A storage device having one or more programs stored thereon.
The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the methods as described above.
The invention also provides a computer storage medium having stored thereon a computer program which, when executed by a processor, implements a method as described above.
Aiming at the problems of difficult positioning and difficult prediction of the communication network fault of the electronic equipment, the invention provides a method based on big data analysis to diagnose and predict the communication network fault of the electronic equipment, a clustering algorithm is used to analyze the communication network state data, aiming at the problems that the clustering result falls into a local optimal solution and the clustering result is greatly influenced by an initial clustering center due to the random selection of the initial clustering center in the traditional K-means algorithm, a division improvement algorithm based on density is provided, the Elman neural network algorithm is used to predict the fault of the communication network of the electronic equipment, the problem of difficult selection of the Elman neural network structure is solved, the structure of an Elman neural network hidden layer is determined by a fitting error analysis method, and the experimental result shows that the method provided by the invention can accurately position the communication network fault of the electronic equipment and can better predict the fault.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (6)
1. A method for diagnosing a communication network failure, comprising:
acquiring historical electronic equipment communication network state data; the method comprises the steps of determining a basic structure of an electronic equipment communication network through researching an operation mechanism of the electronic equipment communication network, and collecting state data of a network control subsystem and a communication network data link layer in the communication subsystem;
clustering the historical electronic equipment communication network state data by using an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the clustering number;
when data set x= { X 1 ，x 2 ，…，x n There are two data objects x in } i And x j The specific description is as follows: x is x i ＝{x i1 ，x i2 ，…，x ir Sum x j ＝{x j1 ，x j2 ，…，x jr The density threshold formula is:
wherein alpha is defined as a density threshold coefficient, a standard deviation range is used as a core area of data according to normal distribution, the distribution of the corresponding core area is selected according to different clustering numbers, and the density threshold coefficient alpha and k are calculated to be the number of clusters; density threshold y z Is defined as the maximum value of Euclidean distance between any two points of a sample point set multiplied by a coefficientIs a value of (2);
calculating the distance between any two points of the data sample, and taking the maximum distance as the diameter to obtain a circular area, wherein the circular area can basically comprise all sample points;
when k=1, a circle a is made at the maximum distance as the diameter, and this circle is used as the range of all data distribution; according to the normal distribution, a numerical range of a standard deviation of 68.27% is taken as a core area of experimental data under the condition of k=1, a circle B is taken as a range of the core area, and the radius ratio of the circle A to the circle B is R: r=1: 0.6827;
the geometric meaning of the density threshold yz is the radius of the circle B, and the density threshold formula is adopted The value of the density threshold coefficient alpha is alpha= 2.9295 under the condition that k=1 can be obtained by the geometric relationship;
when k=2, the distribution ranges of the two types of data in the data distribution range a are represented by circles B and C, and the numerical range of one standard deviation is 68.27% as the core area of the classified data, and the core area ranges of the circles B and C are circles D and E; the value of the density threshold coefficient alpha is alpha= 2.9295 under the condition that k=2 is obtained by the geometric relationship according to the density threshold formula;
similarly, when k=3 is calculated, the density threshold coefficient α= 2.1041;
when k=4 is calculated, the density threshold coefficient α= 1.7682;
when k=5 is calculated, the density threshold coefficient α=1.582;
determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model;
predicting the communication network state data of the real-time monitoring electronic equipment as a test set by using the fault prediction model to obtain a fault diagnosis model;
the improved K-means clustering algorithm utilizes a density division algorithm to determine an initial clustering center, and specifically comprises the following steps:
calculating density threshold values under different clustering number values by using the density threshold value coefficient and Euclidean distance;
determining an initial cluster center according to the historical electronic equipment communication network state data and the density threshold;
the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the number of clusters, and specifically comprises the following steps:
calculating error square sums under different clustering numbers according to the historical electronic equipment communication network state data;
determining a k value-SSE line graph according to the cluster number and the error square sum;
and determining the number of clusters according to the inflection point of the k value-SSE line graph.
2. The communication network fault diagnosis method according to claim 1, wherein the activation function of the Elman neural network is a hyperbolic tangent S-shaped function, and the expression of the hyperbolic tangent S-shaped function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
3. A communication network fault diagnosis system, comprising:
the acquisition module is used for acquiring historical electronic equipment communication network state data; the method comprises the steps of determining a basic structure of an electronic equipment communication network through researching an operation mechanism of the electronic equipment communication network, and collecting state data of a network control subsystem and a communication network data link layer in the communication subsystem;
the clustering module is used for clustering the historical electronic equipment communication network state data by utilizing an improved K-means clustering algorithm to obtain a clustering result; the clustering criterion function of the improved K-means clustering algorithm is an error square sum criterion function; the improved K-means clustering algorithm utilizes a density dividing algorithm to determine an initial clustering center; the improved K-means clustering algorithm utilizes a K value-SSE broken line image algorithm to determine the clustering number;
when data set x= { X 1 ，x 2 ，…，x n There are two data objects x in } i And x j The specific description is as follows: x is x i ＝{x i1 ，x i2 ，…，x ir Sum X j ＝{x j1 ，x j2 ，…，x jr The density threshold formula is:
wherein alpha is defined as a density threshold coefficient, a standard deviation range is used as a core area of data according to normal distribution, the distribution of the corresponding core area is selected according to different clustering numbers, and the density threshold coefficient alpha and k are calculated to be the number of clusters; density threshold y z Is defined as the maximum value of Euclidean distance between any two points of a sample point set multiplied by a coefficientIs a value of (2);
calculating the distance between any two points of the data sample, and taking the maximum distance as the diameter to obtain a circular area, wherein the circular area can basically comprise all sample points;
when k=1, a circle a is made at the maximum distance as the diameter, and this circle is used as the range of all data distribution; according to normal distribution, taking 68.27% of a numerical range of standard deviation as a core area of experimental data under the condition of k=1, taking a circle B as a range of the core area, and the radius ratio of the circle A to the circle B is R:r=1: 0.6827;
density threshold y z Is the radius of circle B, according to the density threshold formula From the geometrical relationship, the density threshold coefficient α can be obtained for k=1The value is alpha= 2.9295;
when k=2, the distribution ranges of the two types of data in the data distribution range a are represented by circles B and C, and the numerical range of one standard deviation is 68.27% as the core area of the classified data, and the core area ranges of the circles B and C are circles D and E; the value of the density threshold coefficient alpha is alpha= 2.9295 under the condition that k=2 is obtained by the geometric relationship according to the density threshold formula;
similarly, when k=3 is calculated, the density threshold coefficient α= 2.1041;
when k=4 is calculated, the density threshold coefficient α= 1.7682;
when k=5 is calculated, the density threshold coefficient α=1.582;
the training module is used for determining the structure of the Elman neural network by using a fitting error method, and training the Elman neural network by taking the clustering result as a training set to obtain a fault prediction model;
the prediction module is used for predicting the communication network state data of the real-time monitoring electronic equipment as a test set by using the fault prediction model to obtain a fault diagnosis model;
an initial cluster center determination sub-module in the cluster module specifically comprises:
the density threshold determining unit is used for calculating density thresholds under different cluster number values by using the density threshold coefficient and the Euclidean distance;
an initial cluster center determining unit, configured to determine an initial cluster center according to the historical electronic equipment communication network state data and the density threshold;
the clustering number determination submodule in the clustering module specifically comprises:
the error square sum determining unit is used for calculating error square sums under different clustering numbers according to the historical electronic equipment communication network state data;
a k-value-SSE line graph determining unit configured to determine a k-value-SSE line graph from the number of clusters and the sum of squares of errors;
and the cluster number determining unit is used for determining the cluster number according to the inflection point of the k value-SSE line graph.
4. A communication network fault diagnosis system according to claim 3, wherein the activation function of the Elman neural network is a hyperbolic tangent S-shaped function, and the expression of the hyperbolic tangent S-shaped function is:
wherein f (sigma) is a hyperbolic tangent S-shaped function, and sigma is an input layer output of the Elman neural network.
5. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-2.
6. A computer storage medium, characterized in that a computer program is stored thereon, wherein the computer program, when executed by a processor, implements the method according to any of claims 1 to 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211209903.1A CN115580526B (en) | 2022-09-30 | 2022-09-30 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211209903.1A CN115580526B (en) | 2022-09-30 | 2022-09-30 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115580526A CN115580526A (en) | 2023-01-06 |
| CN115580526B true CN115580526B (en) | 2024-03-22 |
Family
ID=84582838
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211209903.1A Active CN115580526B (en) | 2022-09-30 | 2022-09-30 | Communication network fault diagnosis method, system, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115580526B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102705303A (en) * | 2012-05-16 | 2012-10-03 | 北京航空航天大学 | Fault location method based on residual and double-stage Elman neural network for hydraulic servo system |
| CN105487009A (en) * | 2015-11-19 | 2016-04-13 | 上海电机学院 | Motor fault diagnosis method based on k-means RBF neural network algorithm |
| CN108460486A (en) * | 2018-03-05 | 2018-08-28 | 国网上海市电力公司 | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network |
| CN110263846A (en) * | 2019-06-18 | 2019-09-20 | 华北电力大学 | The method for diagnosing faults for being excavated and being learnt based on fault data depth |
| CN111198098A (en) * | 2020-01-14 | 2020-05-26 | 重庆邮电大学 | Wind power generator bearing fault prediction method based on artificial neural network |
| CN114297963A (en) * | 2021-12-21 | 2022-04-08 | 北京中匠云智能科技有限公司 | Soft circuit fault diagnosis method based on K-Means wavelet neural network |
| CN114692758A (en) * | 2022-03-31 | 2022-07-01 | 广东电网有限责任公司 | Power communication fault analysis method and device, terminal equipment and medium |
-
2022
- 2022-09-30 CN CN202211209903.1A patent/CN115580526B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102705303A (en) * | 2012-05-16 | 2012-10-03 | 北京航空航天大学 | Fault location method based on residual and double-stage Elman neural network for hydraulic servo system |
| CN105487009A (en) * | 2015-11-19 | 2016-04-13 | 上海电机学院 | Motor fault diagnosis method based on k-means RBF neural network algorithm |
| CN108460486A (en) * | 2018-03-05 | 2018-08-28 | 国网上海市电力公司 | A kind of voltage deviation prediction technique based on improvement clustering algorithm and neural network |
| CN110263846A (en) * | 2019-06-18 | 2019-09-20 | 华北电力大学 | The method for diagnosing faults for being excavated and being learnt based on fault data depth |
| CN111198098A (en) * | 2020-01-14 | 2020-05-26 | 重庆邮电大学 | Wind power generator bearing fault prediction method based on artificial neural network |
| CN114297963A (en) * | 2021-12-21 | 2022-04-08 | 北京中匠云智能科技有限公司 | Soft circuit fault diagnosis method based on K-Means wavelet neural network |
| CN114692758A (en) * | 2022-03-31 | 2022-07-01 | 广东电网有限责任公司 | Power communication fault analysis method and device, terminal equipment and medium |
Non-Patent Citations (2)
| Title |
|---|
| 李 波 管彦允 龚维印 韦旭勤 薛 端.基于密度的 K-means 初始聚类中心点选取算法.绥 化 学 院 学 报.2022,第42卷(第6期),第149-150页. * |
| 陶永辉 王 勇.基于初始聚类中心选取的改进 K - means 算法.国外电子测量技术.2022,第41卷(第9期),第55-56页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115580526A (en) | 2023-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112036084B (en) | Similar product life migration screening method and system | |
| CN111860982A (en) | Wind power plant short-term wind power prediction method based on VMD-FCM-GRU | |
| CN111144552B (en) | Multi-index grain quality prediction method and device | |
| CN108805193B (en) | Electric power missing data filling method based on hybrid strategy | |
| CN108052387B (en) | Resource allocation prediction method and system in mobile cloud computing | |
| CN113938488A (en) | Load balancing method based on dynamic and static weighted polling | |
| CN106656357B (en) | Power frequency communication channel state evaluation system and method | |
| CN111985845B (en) | Node priority optimization method of heterogeneous Spark cluster | |
| CN112363896A (en) | Log anomaly detection system | |
| CN112149883A (en) | Photovoltaic power prediction method based on FWA-BP neural network | |
| CN110968512B (en) | Software quality evaluation method, device, equipment and computer readable storage medium | |
| CN112085143A (en) | Assimilation mechanism-based chaotic ant colony algorithm and Web service combination optimization method | |
| CN113255873A (en) | Clustering longicorn herd optimization method, system, computer equipment and storage medium | |
| CN115099133A (en) | TLMPA-BP-based cluster system reliability evaluation method | |
| CN116668327A (en) | Small sample malicious flow classification increment learning method and system based on dynamic retraining | |
| CN111667189A (en) | Construction engineering project risk prediction method based on one-dimensional convolutional neural network | |
| CN114818487A (en) | Natural gas and wet gas pipeline liquid holdup prediction model method based on PSO-BP neural network | |
| CN113033898A (en) | Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network | |
| CN115580526B (en) | Communication network fault diagnosis method, system, electronic equipment and storage medium | |
| Li et al. | Predicting Software Quality by Optimized BP Network Based on PSO. | |
| CN111711530A (en) | Link prediction algorithm based on community topological structure information | |
| CN116522747A (en) | Two-stage optimized extrusion casting process parameter optimization design method | |
| CN112990603B (en) | Air conditioner cold load prediction method and system considering frequency domain decomposed data characteristics | |
| CN114662658A (en) | On-chip optical network hot spot prediction method based on LSTM neural network | |
| Zhang et al. | Evaluation of borrower's credit of P2P loan based on adaptive particle swarm optimisation BP neural 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 |