CN111488711A - Network robustness assessment method and system - Google Patents

Network robustness assessment method and system Download PDF

Info

Publication number
CN111488711A
CN111488711A CN202010282838.XA CN202010282838A CN111488711A CN 111488711 A CN111488711 A CN 111488711A CN 202010282838 A CN202010282838 A CN 202010282838A CN 111488711 A CN111488711 A CN 111488711A
Authority
CN
China
Prior art keywords
network
initial
model
robustness
load
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010282838.XA
Other languages
Chinese (zh)
Inventor
郑桦
邓原
梁倬骞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jinan University
Original Assignee
Jinan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jinan University filed Critical Jinan University
Priority to CN202010282838.XA priority Critical patent/CN111488711A/en
Publication of CN111488711A publication Critical patent/CN111488711A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • 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
    • 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)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Medical Informatics (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Image Analysis (AREA)

Abstract

The invention belongs to the technical field of network detection, and particularly relates to a network robustness assessment method and system. The method comprises the following steps: obtaining initial parameters of a network, and obtaining an initial load model of the network according to the initial parameters, wherein the initial parameters comprise: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network; the convolutional neural network model is used for extracting the features, the features are learned through the features, network faults based on random walk are simulated in the network, and the robust graph is characterized and learned according to the initial parameters and the initial load model, so that the accuracy of obtaining the network robust graph is improved. Furthermore, the robustness of the network is evaluated according to the corresponding target network huge component when the network is stabilized again after the network fails, the connection condition of all nodes in the network is not needed, the data processing amount is small, and the speed and the precision of the evaluation of the network robustness are improved.

Description

Network robustness assessment method and system
Technical Field
The invention belongs to the technical field of network detection, and particularly relates to a network robustness assessment method and system.
Background
The network is composed of nodes and links, and represents a plurality of objects and their mutual connection. There are many complex networks in life such as: the power network is a network formed by power stations and cables connected with the power stations; the social relationship network is a network formed by social people and the relationship among people; the traffic network is a network formed by crossroads and roads among the crossroads; neural networks, computer networks, internet of things, and the like are all similar networks. These networks are often open, that is, nodes and edges in the network interact with the external environment, and when the external environment generates a small disturbance and causes several nodes in the network to fail, the interaction between the nodes may cause successive failures of the nodes in the system, and even cause the entire network to be dysfunctional, resulting in system crash, which is called as a cascading failure of the network.
In recent years, several catastrophic events have occurred in various places resulting from cascading failures. In response to this problem, although many scholars have made a lot of effort and research, with the progress of science and technology and the demand of people for improving living standards, the functions and structures of these networks are more and more complicated, which also results in the reduction of security and reliability of the networks. When the network has a cascading failure, the network is more prone to breakdown, so that a method for effectively preventing or stopping the cascading failure needs to be found urgently to improve the robustness of the network, wherein the robustness of the network refers to the characteristic that the network maintains certain performance under certain (structure and size) parameter perturbation. In the prior art, in order to prevent cascading failures and improve network robustness, a load-capacity model is proposed in numerous researches. This model defines the load mainly by the betweenness of the edges (nodes), which is the sum of the minimum paths through the nodes (edges) in the network. However, it is necessary to know the connection situation of all nodes in the network, which is unlikely in practice, and the required data volume is large, the data processing process is complicated, and the speed and accuracy of the network robustness evaluation are affected.
Therefore, how to provide a scheme can improve the speed and the accuracy of network robustness evaluation, which becomes a problem to be solved urgently.
Disclosure of Invention
The invention aims to provide a network robustness assessment method and a network robustness assessment system, and solves the problem that a rotary fastener for supporting an existing template cannot control a diagonal brace to participate in stress by adjusting clamping force.
The technical scheme adopted by the invention is as follows:
the invention provides a network robustness assessment method, which comprises the following steps:
obtaining initial parameters of a network, and obtaining an initial load model of the network according to the initial parameters, wherein the initial parameters comprise: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
s1 preprocessing stage: the relation between the basic statistics of the graph data access degree and r and s robustness values is researched at this stage, meanwhile, a k-means algorithm is used for clustering the feature vectors for feature expansion, and a mode is sought from a spectrum space to distinguish the matrixes;
s2 training stage: at this stage, respective loss functions, optimization functions and full connection layers of the sensors are customized according to r and s values, and an adam optimization algorithm is mainly adopted to train the multilayer sensor model in a batch mode in the training process;
and an S3 testing stage, wherein the M L P model hyperparameters obtained in the training stage are used for carrying out class prediction on the features obtained in the preprocessing stage in the process to obtain a final classification result.
Simulating a network fault based on random walk in the network, extracting features based on a convolutional neural network model according to the initial parameters and the initial load model, and performing feature learning through the features, wherein a network mathematical expression comprising L hidden layers is as follows:
f(x)＝σ(WL…σ(W2σ(W1x+b1)+b2)…+bL)
wherein x is the input feature vector, W is the weight matrix, b is the bias vector, σ is the activation function, after passing through some neurons, the input feature vector will be passed into the classifier, and the loss function of the classifier is expressed as follows:
C(f(x),y)＝l(f(x),y)
wherein, y corresponds to a real label and represents r and s robust performance values of a network, and l represents a classifier function;
and finally, inputting the output of the multi-layer perceptron into a classifier to realize the evaluation of the multi-agent network robustness.
In the S2 training stage, respective loss functions, optimization functions and hyper-parameters are configured for r and S values in the convolutional neural network structure design process, the adam optimization algorithm and the softmax cross entropy loss function are adopted in the training process, a cross-folding cross verification method is adopted, and the finally obtained CNN model is used as a test model.
And analyzing the complexity of the preprocessed algorithm, firstly acquiring all basic statistics of the robust network, then decomposing the matrix, and directly performing characterization learning on the robust graph by using a convolutional neural network model.
The invention also provides a network robustness evaluation system, which comprises:
an initial model obtaining unit, configured to obtain initial parameters of a network, and obtain an initial load model of the network according to the initial parameters, where the initial parameters include: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
the multilayer perceptron unit is used for sending all the extracted features to the neural network model for feature learning, and training the multilayer perceptron model in a batch mode by adopting an adam optimization algorithm;
the convolutional neural network unit is used for extracting the characteristics and learning the characteristics through the characteristics, network faults based on random walk are simulated in the network, and the robust graph is characterized and learned according to the initial parameters and the initial load model;
and the robustness evaluation unit is used for inputting the output of the multilayer perceptron into the classifier so as to realize the evaluation of the robustness of the multi-intelligent-body network.
The invention has the beneficial effects that: according to the network robustness assessment method and system, through the multilayer perceptron unit, the convolutional neural network unit and the robustness assessment unit, network faults based on random walk are simulated in a network, the convolutional neural network model is used for extracting features and learning the features, the network faults based on random walk are simulated in the network, and the robust graph is subjected to characterization learning according to the initial parameters and the initial load model, so that the accuracy of obtaining the network robust graph is improved. Furthermore, the robustness of the network is evaluated according to the corresponding target network huge component when the network is stabilized again after the network fails, the connection condition of all nodes in the network is not needed, the data processing amount is small, and the speed and the precision of the network robustness evaluation are improved.
Detailed Description
The specific implementation mode adopts the following technical scheme:
the invention provides a network robustness assessment method, which comprises the following steps:
obtaining initial parameters of a network, and obtaining an initial load model of the network according to the initial parameters, wherein the initial parameters comprise: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
s1 preprocessing stage: the relation between the basic statistics of the graph data access degree and r and s robustness values is researched at this stage, meanwhile, a k-means algorithm is used for clustering the feature vectors for feature expansion, and a mode is sought from a spectrum space to distinguish the matrixes;
s2 training stage: at this stage, respective loss functions, optimization functions and full connection layers of the sensors are customized according to r and s values, and an adam optimization algorithm is mainly adopted to train the multilayer sensor model in a batch mode in the training process;
and an S3 testing stage, wherein the M L P model hyperparameters obtained in the training stage are used for carrying out class prediction on the features obtained in the preprocessing stage in the process to obtain a final classification result.
Simulating a network fault based on random walk in the network, extracting features based on a convolutional neural network model according to the initial parameters and the initial load model, and performing feature learning through the features, wherein a network mathematical expression comprising L hidden layers is as follows:
f(x)＝σ(WL…σ(W2σ(W1x+b1)+b2)…+bL)
wherein x is the input feature vector, W is the weight matrix, b is the bias vector, σ is the activation function, after passing through some neurons, the input feature vector will be passed into the classifier, and the loss function of the classifier is expressed as follows:
C(f(x),y)＝l(f(x),y)
wherein, y corresponds to a real label and represents r and s robust performance values of a network, and l represents a classifier function;
and finally, inputting the output of the multi-layer perceptron into a classifier to realize the evaluation of the multi-agent network robustness.
In the S2 training stage, respective loss functions, optimization functions and hyper-parameters are configured for r and S values in the convolutional neural network structure design process, the adam optimization algorithm and the softmax cross entropy loss function are adopted in the training process, a cross-folding cross verification method is adopted, and the finally obtained CNN model is used as a test model.
And analyzing the complexity of the preprocessed algorithm, firstly acquiring all basic statistics of the robust network, then decomposing the matrix, and directly performing characterization learning on the robust graph by using a convolutional neural network model.
The invention also provides a network robustness evaluation system, which comprises:
an initial model obtaining unit, configured to obtain initial parameters of a network, and obtain an initial load model of the network according to the initial parameters, where the initial parameters include: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
the multilayer perceptron unit is used for sending all the extracted features to the neural network model for feature learning, and training the multilayer perceptron model in a batch mode by adopting an adam optimization algorithm;
the convolutional neural network unit is used for extracting the characteristics and learning the characteristics through the characteristics, network faults based on random walk are simulated in the network, and the robust graph is characterized and learned according to the initial parameters and the initial load model;
and the robustness evaluation unit is used for inputting the output of the multilayer perceptron into the classifier so as to realize the evaluation of the robustness of the multi-intelligent-body network.
While the invention has been described above with reference to an embodiment, various modifications may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In particular, the various features of the embodiments disclosed herein may be used in any combination, provided that there is no structural conflict, and the combinations are not exhaustively described in this specification merely for the sake of brevity and conservation of resources. Therefore, it is intended that the invention not be limited to the particular embodiments disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.

Claims (4)

1. A method for network robustness assessment, the method comprising:
obtaining initial parameters of a network, and obtaining an initial load model of the network according to the initial parameters, wherein the initial parameters comprise: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
s1 preprocessing stage: the relation between the basic statistics of the graph data access degree and r and s robustness values is researched at this stage, meanwhile, a k-means algorithm is used for clustering the feature vectors for feature expansion, and a mode is sought from a spectrum space to distinguish the matrixes;
s2 training stage: at this stage, respective loss functions, optimization functions and full connection layers of the sensor are customized according to r and s values, and an adam optimization algorithm is mainly adopted to train the multi-layer sensor model in a batch mode in the training process;
and an S3 testing stage, wherein the M L P model hyperparameters obtained in the training stage are used for carrying out class prediction on the features obtained in the preprocessing stage in the process to obtain a final classification result.
Simulating a network fault based on random walk in the network, extracting features based on a convolutional neural network model according to the initial parameters and the initial load model, and performing feature learning through the features, wherein a network mathematical expression comprising L hidden layers is as follows:
f(x)＝σ(WL…σ(W2σ(W1x+b1)+b2)…+bL)
wherein x is the input feature vector, W is the weight matrix, b is the bias vector, σ is the activation function, after passing through some neurons, the input feature vector will be passed into the classifier, and the loss function of the classifier is expressed as follows:
C(f(x),y)＝l(f(x),y)
wherein, y corresponds to a real label and represents r and s robust performance values of a network, and l represents a classifier function;
and finally, inputting the output of the multi-layer perceptron into a classifier to realize the evaluation of the multi-agent network robustness.
2. The method of claim 1, wherein: in the S2 training stage, respective loss functions, optimization functions and hyper-parameters are configured for r and S values in the convolutional neural network structure design process, the adam optimization algorithm and the softmax cross entropy loss function are adopted in the training process, a cross-folding cross verification method is adopted, and the finally obtained CNN model is used as a test model.
3. The method of claim 1, wherein: and analyzing the complexity of the preprocessed algorithm, firstly acquiring all basic statistics of the robust network, then decomposing the matrix, and directly performing characterization learning on the robust graph by using a convolutional neural network model.
4. A network robustness assessment system, comprising:
an initial model obtaining unit, configured to obtain initial parameters of a network, and obtain an initial load model of the network according to the initial parameters, where the initial parameters include: an initial node number, an initial edge number, an initial random walker number, a load tolerance parameter, and a load shedding parameter of the network;
the multilayer perceptron unit is used for sending all the extracted features to the neural network model for feature learning, and training the multilayer perceptron model in a batch mode by adopting an adam optimization algorithm;
the convolutional neural network unit is used for extracting the characteristics and learning the characteristics through the characteristics, network faults based on random walk are simulated in the network, and the robust graph is characterized and learned according to the initial parameters and the initial load model;
and the robustness evaluation unit is used for inputting the output of the multi-layer perceptron into the classifier so as to realize the evaluation of the robustness of the multi-agent network.
CN202010282838.XA 2020-04-08 2020-04-08 Network robustness assessment method and system Pending CN111488711A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010282838.XA CN111488711A (en) 2020-04-08 2020-04-08 Network robustness assessment method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010282838.XA CN111488711A (en) 2020-04-08 2020-04-08 Network robustness assessment method and system

Publications (1)

Publication Number Publication Date
CN111488711A true CN111488711A (en) 2020-08-04

Family

ID=71798161

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010282838.XA Pending CN111488711A (en) 2020-04-08 2020-04-08 Network robustness assessment method and system

Country Status (1)

Country Link
CN (1) CN111488711A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022062649A1 (en) * 2020-09-25 2022-03-31 苏州浪潮智能科技有限公司 Neural network robustness measurement method, and apparatus
CN114329867A (en) * 2021-12-30 2022-04-12 太原理工大学 Scale-free network robustness measuring method based on motif

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107426000A (en) * 2017-04-24 2017-12-01 北京航空航天大学 A kind of network robustness appraisal procedure and system
CN109583092A (en) * 2018-11-30 2019-04-05 中南大学 A kind of intelligent machine diagnosis method for system fault of multi-level multi-mode feature extraction
US10402692B1 (en) * 2019-01-22 2019-09-03 StradVision, Inc. Learning method and learning device for fluctuation-robust object detector based on CNN using target object estimating network adaptable to customers' requirements such as key performance index, and testing device using the same
CN110224862A (en) * 2019-05-20 2019-09-10 杭州电子科技大学 Multi-agent system network appearance based on multilayer perceptron invades capability assessment method
CN110289987A (en) * 2019-05-27 2019-09-27 杭州电子科技大学 Multi-agent system network resilience appraisal procedure based on representative learning
CN110348399A (en) * 2019-07-15 2019-10-18 中国人民解放军国防科技大学 EO-1 hyperion intelligent method for classifying based on prototype study mechanism and multidimensional residual error network
US20200065664A1 (en) * 2018-08-22 2020-02-27 Fujitsu Limited System and method of measuring the robustness of a deep neural network
CN110852417A (en) * 2019-10-10 2020-02-28 华东师范大学 Single-depth neural network model robustness improving method for application of Internet of things

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107426000A (en) * 2017-04-24 2017-12-01 北京航空航天大学 A kind of network robustness appraisal procedure and system
US20200065664A1 (en) * 2018-08-22 2020-02-27 Fujitsu Limited System and method of measuring the robustness of a deep neural network
CN109583092A (en) * 2018-11-30 2019-04-05 中南大学 A kind of intelligent machine diagnosis method for system fault of multi-level multi-mode feature extraction
US10402692B1 (en) * 2019-01-22 2019-09-03 StradVision, Inc. Learning method and learning device for fluctuation-robust object detector based on CNN using target object estimating network adaptable to customers' requirements such as key performance index, and testing device using the same
CN110224862A (en) * 2019-05-20 2019-09-10 杭州电子科技大学 Multi-agent system network appearance based on multilayer perceptron invades capability assessment method
CN110289987A (en) * 2019-05-27 2019-09-27 杭州电子科技大学 Multi-agent system network resilience appraisal procedure based on representative learning
CN110348399A (en) * 2019-07-15 2019-10-18 中国人民解放军国防科技大学 EO-1 hyperion intelligent method for classifying based on prototype study mechanism and multidimensional residual error network
CN110852417A (en) * 2019-10-10 2020-02-28 华东师范大学 Single-depth neural network model robustness improving method for application of Internet of things

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
付丹龙等: "基于不同子网络级联机制的相依网络鲁棒性研究", 《计算机工程与科学》 *
周漩等: "利用节点效率评估复杂网络功能鲁棒性", 《物理学报》 *
梁倬骞等: "级联过程中基于动态信息的负载重分配策略", 《计算机工程与科学》 *
王广: "恶意攻击下多智能体网络鲁棒性评估方法", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 *
郝羽成等: "考虑节点过载的复杂网络级联失效模型", 《系统工程与电子技术》 *
韩丽等: "加权无标度网络的级联失效模型", 《软件学报》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022062649A1 (en) * 2020-09-25 2022-03-31 苏州浪潮智能科技有限公司 Neural network robustness measurement method, and apparatus
CN114329867A (en) * 2021-12-30 2022-04-12 太原理工大学 Scale-free network robustness measuring method based on motif
CN114329867B (en) * 2021-12-30 2023-07-21 太原理工大学 Scale-free network robustness measuring method based on die body

Similar Documents

Publication Publication Date Title
Gou et al. Aeroengine control system sensor fault diagnosis based on CWT and CNN
Jana et al. CNN and Convolutional Autoencoder (CAE) based real-time sensor fault detection, localization, and correction
Fan et al. A hybrid FDD strategy for local system of AHU based on artificial neural network and wavelet analysis
CN105973594B (en) A kind of rolling bearing fault Forecasting Methodology based on continuous depth confidence network
Smarsly et al. Machine learning techniques for structural health monitoring
Saxena et al. Evolving an artificial neural network classifier for condition monitoring of rotating mechanical systems
CN108921230A (en) Method for diagnosing faults based on class mean value core pivot element analysis and BP neural network
CN109766992A (en) Industry control abnormality detection and attack classification based on deep learning
CN115204302A (en) Unmanned aerial vehicle small sample fault diagnosis system and method
CN105425150A (en) Motor fault diagnosis method based on RBF and PCA-SVDD
Marugán et al. SCADA and artificial neural networks for maintenance management
Kabore et al. Review of anomaly detection systems in industrial control systems using deep feature learning approach
CN109472097A (en) A kind of transmission line of electricity on-line monitoring equipment method for diagnosing faults
Entezami et al. Condition assessment of civil structures for structural health monitoring using supervised learning classification methods
CN111488711A (en) Network robustness assessment method and system
Vera-Tudela et al. On the selection of input variables for a wind turbine load monitoring system
Li et al. Intelligent fault diagnosis of aeroengine sensors using improved pattern gradient spectrum entropy
Tang et al. Prediction of bearing performance degradation with bottleneck feature based on LSTM network
Yang et al. A pairwise graph regularized constraint based on deep belief network for fault diagnosis
Li et al. Structural health monitoring data anomaly detection by transformer enhanced densely connected neural networks
Zhou et al. Degradation State Recognition of Rolling Bearing Based on K‐Means and CNN Algorithm
Ghate et al. Induction machine fault detection using support vector machine based classifier
Gao et al. Fault detection of electric vehicle charging piles based on extreme learning machine algorithm
He et al. Abnormal data detection and recovery of sensors network based on spatiotemporal deep learning methodology
Segovia et al. Wind turbine alarm management with artificial neural networks

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20200804