CN115580545A - Internet of things communication method for improving data transmission efficiency - Google Patents

Internet of things communication method for improving data transmission efficiency Download PDF

Info

Publication number
CN115580545A
CN115580545A CN202211576831.4A CN202211576831A CN115580545A CN 115580545 A CN115580545 A CN 115580545A CN 202211576831 A CN202211576831 A CN 202211576831A CN 115580545 A CN115580545 A CN 115580545A
Authority
CN
China
Prior art keywords
data
real
time
monitoring terminal
clustering
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.)
Granted
Application number
CN202211576831.4A
Other languages
Chinese (zh)
Other versions
CN115580545B (en
Inventor
胡增
钟生
杨坤龙
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.)
China Applied Technology Co Ltd
Original Assignee
China Applied Technology Co Ltd
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 China Applied Technology Co Ltd filed Critical China Applied Technology Co Ltd
Priority to CN202211576831.4A priority Critical patent/CN115580545B/en
Publication of CN115580545A publication Critical patent/CN115580545A/en
Application granted granted Critical
Publication of CN115580545B publication Critical patent/CN115580545B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/02—Capturing of monitoring data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/04—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
    • H04L63/0407—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the identity of one or more communicating identities is hidden
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/08—Network architectures or network communication protocols for network security for authentication of entities
    • H04L63/0823—Network architectures or network communication protocols for network security for authentication of entities using certificates
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/08—Network architectures or network communication protocols for network security for authentication of entities
    • H04L63/083—Network architectures or network communication protocols for network security for authentication of entities using passwords
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses an Internet of things communication method for improving data transmission efficiency, which comprises the following steps: s1, identity authentication between a monitoring terminal and a control terminal is realized; s2, identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm; s3, processing the acquired real-time initial data by using the improved radial basis function neural network model to obtain corrected real-time data; s4, analyzing the variation of the corrected real-time data by using a threshold comparison method, and screening the real-time data; and S5, transmitting the screened real-time data between the monitoring terminal and the control terminal through a communication network. The method and the device can correct abnormal data in the real-time initial data, effectively reduce the pressure of data processing, and improve the efficiency of data transmission, thereby better meeting the use requirements of the power industry.

Description

Internet of things communication method for improving data transmission efficiency
Technical Field
The invention relates to the technical field of communication of the Internet of things, in particular to a communication method of the Internet of things for improving data transmission efficiency.
Background
The concept of the internet of things is developed on the basis of the internet, and the concept of the internet of things extends and expands a user side to any article to article for information exchange and communication. The Internet of things digitalizes the real world, has a wide application range and mainly comprises the following aspects: the system has the advantages of wide market and application prospect in the fields of transportation and logistics, health and medical treatment, intelligent environment (home, office and factory), individuals and society and the like.
At present, information sensing modules such as inductors are mainly embedded or installed in field devices such as power grids, railways, bridges, buildings, water supply systems, oil and gas pipelines and industrial devices, and are in communication connection with the cloud end through the network, so that the internet of things is formed. The intelligent sensor has the main functions of collecting data of various field sensors, transmitting the data to a cloud end through various wireless and wired communication means, performing intelligent calculation, and controlling various field execution devices according to a calculation result, so that the object-to-object connection and the intelligent function are achieved.
Particularly for the power industry, in the actual project, the types of equipment needing data acquisition are various and the quantity is large, and the equipment acquisition period is short according to the real-time requirement, so that the data acquired by the data acquisition terminal and the data gathered by the gateway of the internet of things are massive. However, in the existing internet-of-things communication process, only the collected data is packed and processed, and uploaded according to the existing standard communication protocol or the enterprise private protocol, and algorithm processing is not performed on the collected mass data, so that not only is the high concurrency of a server (data receiving end) and the pressure of data processing huge, but also resources of the server are easily exhausted to cause 'false death' or downtime, and because the collected data cannot be uploaded in time, a large amount of accumulation can be generated in a cache, and finally equipment downtime is caused, and the application requirements in the power industry cannot be well met.
An effective solution to the problems in the related art has not been proposed yet.
Disclosure of Invention
Aiming at the problems in the related art, the invention provides an Internet of things communication method for improving data transmission efficiency, so as to overcome the technical problems in the prior related art.
Therefore, the invention adopts the following specific technical scheme:
an Internet of things communication method for improving data transmission efficiency comprises the following steps:
s1, identity authentication between a monitoring terminal and a control terminal is realized by using an identity authentication method;
s2, acquiring real-time initial data of the monitoring terminal, and identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm;
s3, processing the acquired real-time initial data by using an improved radial basis function neural network model based on the identification result of the abnormal data to obtain corrected real-time data;
s4, analyzing the variation of the corrected real-time data by using a threshold comparison method, and realizing the screening of the real-time data;
and S5, transmitting the screened real-time data between the monitoring terminal and the control terminal through a communication network.
Further, the method for realizing the identity authentication between the monitoring terminal and the control terminal by using the identity authentication method comprises the following steps:
s11, the monitoring terminal initiates a connection request to the control terminal, and the control terminal receives the connection request and sends an encrypted digital certificate to the monitoring terminal;
s12, the monitoring terminal receives the encrypted digital certificate and carries out decryption verification, and after verification is passed, the identity of the control terminal is confirmed to be reliable;
s13, the monitoring terminal generates a new random number, takes out a public key in the digital certificate of the control terminal, and sends the new random number encrypted by the public key to the control terminal;
s14, the control terminal receives the encrypted new random number, decrypts the new random number by using a private key of the control terminal to obtain the new random number, and uses the new random number as a session key with the monitoring terminal;
and S15, establishing communication connection between the monitoring terminal and the control terminal by using the new random number as a session key between the monitoring terminal and the control terminal.
Further, the steps of acquiring real-time initial data of the monitoring terminal and identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm comprise the following steps:
s21, acquiring real-time initial data of the monitoring terminal, and judging existence of bad data in the real-time initial data;
s22, calculating elbow break angles by utilizing an elbow criterion and determining the most appropriate clustering number;
and S23, detecting and identifying bad data in the real-time initial data according to the optimal cluster number.
Further, the step of acquiring the real-time initial data of the monitoring terminal and judging the existence of bad data in the real-time initial data comprises the following steps:
s211, acquiring real-time initial data of the monitoring terminal, and clustering the real-time initial data and the reference data respectively to obtain clustering dispersion of the clustering data;
s212, calculating intermittent values g by using intermittent calculation formulas ap (1) And g ap (2) Wherein, the intermittent calculation formula is as follows:
Figure 940605DEST_PATH_IMAGE001
（1）；
wherein k represents the number of clusters, F represents the number of groups of the reference distribution data set, lnW (k) represents the cluster dispersion, r represents the reference data, and j represents the lower bound in the summation symbol;
s213, judging whether the formula (1) is satisfied
Figure 594440DEST_PATH_IMAGE002
Figure 101776DEST_PATH_IMAGE003
；
If yes, the optimal number of clusters is 1, that is, all the real-time initial data are good data, and if not, S22 is executed.
Further, the step of calculating the elbow angle by using the elbow criterion and determining the most suitable clustering number comprises the following steps:
s221, setting an initial value of k = k +1,k to 1, respectively calculating a clustering dispersion lnW (k) of the real-time initial data, and calculating an elbow angle of a clustering dispersion curve at each clustering point, wherein a calculation formula of the elbow angle is as follows:
Figure 815654DEST_PATH_IMAGE004
；
s222, if k makes it satisfy
Figure 657708DEST_PATH_IMAGE005
And k is the optimal cluster number, and if the k is not satisfied, the step returns to S221.
Further, the detection and identification of the bad data in the real-time initial data according to the optimal clustering number comprises the following steps:
s231, judging whether the optimal clustering number is 1, if so, indicating that all real-time initial data are good data, and if not, calculating the average value of the data in each cluster;
s232, setting the cluster with the minimum average value to belong to the cluster of normal data, and determining the rest clusters to be clusters formed by bad data, so that the detection and identification of the bad data in the real-time initial data are realized.
Further, the processing the acquired real-time initial data by using the improved radial basis function neural network model based on the identification result of the abnormal data to obtain the corrected real-time data includes the following steps:
s31, optimizing a pre-constructed radial basis function neural network model by using the improved ant colony algorithm to obtain an improved radial basis function neural network model and training;
s32, acquiring normal data of bad data in the real-time initial data at the previous moment, and outputting prediction data of the current moment through an improved radial basis function neural network model;
and S33, taking the predicted data of the current time as a correction result of the bad data to obtain corrected real-time data.
Further, the optimizing the radial basis function neural network model by using the improved ant colony algorithm to obtain the improved radial basis function neural network model includes the following steps:
s311, acquiring normal data of a period of time before bad data appear in the real-time initial data as an input sample;
s312, clustering the input samples by using the improved ant colony clustering algorithm to obtain a clustering center serving as a center value of a hidden layer unit of the radial basis function neural network;
s313, adjusting the weight from the hidden layer to the output layer by using a pseudo-inverse algorithm;
s314, calculating the output value of each hidden layer unit, normalizing the output value, and judging the contribution of each hidden layer unit to network output;
s315, under the condition of meeting the network output error, removing the hidden layer unit of which the output contribution of the whole network is smaller than a preset threshold value by using a cutting method, and realizing the optimization of the network structure;
and S316, obtaining an improved radial basis function neural network model based on the optimized network parameters and structure, and training.
Further, the calculation formula of the pseudo-inverse algorithm is as follows:
Figure 93064DEST_PATH_IMAGE006
；
in the formula (I), the compound is shown in the specification,
Figure 969753DEST_PATH_IMAGE007
the weight value is represented by a weight value,da matrix of desired outputs is represented that,orepresenting an output matrixT represents transposition;
the calculation formula of the output value of the hidden layer unit is as follows:
Figure 339685DEST_PATH_IMAGE008
；
in the formula (I), the compound is shown in the specification,
Figure 403456DEST_PATH_IMAGE009
the output of the hidden layer element is represented,
Figure 399094DEST_PATH_IMAGE010
the output of the hidden layer cell representing the largest absolute value,x n the input is represented by a representation of the input,
Figure 848399DEST_PATH_IMAGE011
the allowed error values, k =1,2, …, n,
Figure 904080DEST_PATH_IMAGE012
representing the cluster centers or center vectors of the hidden layer node basis functions, exp representing the headers of the exponential function expressions.
Further, the analyzing the variation of the corrected real-time data by using a threshold comparison method to realize the screening of the real-time data comprises the following steps:
s41, collecting the corrected real-time data and acquiring preset reference data;
s42, calculating the variable quantity of the corrected real-time data and preset reference data;
s43, comparing the variable quantity with a preset variable quantity threshold value, and keeping the real-time data with the variable quantity larger than the preset variable quantity threshold value as screened real-time data.
The invention has the beneficial effects that:
1) The method analyzes the variation of the corrected real-time data by using a threshold comparison method, and screens the real-time data, so that the transmission quantity of the data is effectively reduced, the problem of mass accumulation of the data is solved, the pressure of data processing is reduced, the efficiency of data transmission is effectively improved, the real-time property of data transmission is ensured, and the use requirement of the power industry can be better met.
2) The invention can realize the detection and identification of abnormal data by utilizing an elbow criterion method based on an intermittent statistical algorithm, and can correct the abnormal data in the real-time initial data by utilizing an improved radial basis function neural network model, thereby effectively avoiding the influence on the safe operation of equipment caused by abnormal monitoring data.
3) According to the method, the network parameters and the network structure of the pre-constructed radial basis function neural network model are optimized by using the improved ant colony algorithm, so that the defects of the traditional neural network can be effectively overcome, and the abnormal data correction is performed on the real-time initial data by using the improved radial basis function neural network model, so that the correction accuracy rate of the abnormal data in the real-time initial data can be effectively improved, and the requirement for correcting the abnormal data can be better met.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a communication method of the internet of things for improving data transmission efficiency according to an embodiment of the present invention.
Detailed Description
For further explanation of the various embodiments, the drawings which form a part of the disclosure and which are incorporated in and constitute a part of this specification, illustrate embodiments and, together with the description, serve to explain the principles of operation of the embodiments, and to enable others of ordinary skill in the art to understand the various embodiments and advantages of the invention, and, by reference to these figures, reference is made to the accompanying drawings, which are not to scale and wherein like reference numerals generally refer to like elements.
According to the embodiment of the invention, the communication method of the Internet of things for improving the data transmission efficiency is provided.
Referring to the drawings and the detailed description, the invention will be further described, as shown in fig. 1, according to the communication method of the internet of things for improving data transmission efficiency, the communication method of the internet of things includes the following steps:
s1, identity authentication between a monitoring terminal and a control terminal is realized by using an identity authentication method;
the method for realizing the identity authentication between the monitoring terminal and the control terminal by using the identity authentication method comprises the following steps:
s11, the monitoring terminal initiates a connection request to the control terminal, and the control terminal receives the connection request and sends an encrypted digital certificate to the monitoring terminal;
s12, the monitoring terminal receives the encrypted digital certificate, carries out decryption verification and confirms that the identity of the control terminal is reliable after the verification is passed;
s13, the monitoring terminal generates a new random number, takes out a public key in the digital certificate of the control terminal, and sends the new random number encrypted by the public key to the control terminal;
s14, the control terminal receives the encrypted new random number, decrypts the new random number by using a private key of the control terminal to obtain the new random number, and uses the new random number as a session key with the monitoring terminal;
and S15, establishing communication connection between the monitoring terminal and the control terminal by using the new random number as a session key between the monitoring terminal and the control terminal.
S2, acquiring real-time initial data of the monitoring terminal, and identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm;
a data mining algorithm based on a batch statistics algorithm (GSA) is a strengthened clustering effect, can estimate the optimal clustering number of a data set, and can accurately distinguish clusters of good data and bad data in the identification of bad data of an electric power system to detect and identify the bad data. However, when the number of clusters is large, the calculated amount is large, and in a modern power system with the larger and larger data amount, the speed of the algorithm cannot well meet the use requirements of people, so that in order to improve the calculation speed and reduce misjudgment, the implementation provides an elbow criterion for estimating the number of clusters, and the criterion is combined with a GSA method to form the elbow criterion based on the GSA for identifying bad data of the power system.
The method for acquiring real-time initial data of the monitoring terminal and identifying abnormal data by using the elbow criterion method based on the intermittent statistical algorithm comprises the following steps of:
s21, acquiring real-time initial data of the monitoring terminal, and judging existence of bad data in the real-time initial data;
in the communication of the internet of things in the embodiment, the real-time initial data in the power grid system includes real-time data such as current, voltage and temperature, the reference data is the initial data of the acquisition device in the first acquisition period, and in order to ensure the reliability of the reference data, the initial data can be subjected to denoising processing to obtain denoising data, then the denoising data is set as the reference data, and the reference data is stored in the memory database. For the parameters considered as noise in the denoising process, the parameters need to be replenished to the corresponding positions of the in-memory database in the subsequent acquisition cycle.
Specifically, the acquiring of the real-time initial data of the monitoring terminal and the judging of the existence of the bad data in the real-time initial data include the following steps:
s211, acquiring real-time initial data of the monitoring terminal, and clustering the real-time initial data and the reference data respectively to obtain clustering dispersion of the clustering data;
in the embodiment, two principles are followed during clustering division, namely, the similarity of data in the same group is maximized; the similarity of data between different groups is minimized. The measure of similarity or dissimilarity is determined based on the value of the data object description attributes.
Specifically, the calculation formula of the clustering dispersion is as follows:
Figure 189567DEST_PATH_IMAGE013
；
in the formula (I), the compound is shown in the specification,h i representing the real-time initial data of the data,c i representing clustersx i The center of (a);
s212, calculating intermittent values g by using intermittent calculation formulas ap (1) And g ap (2) Wherein, the intermittent calculation formula is as follows:
Figure 208470DEST_PATH_IMAGE001
（1）；
wherein k represents the number of clusters, F represents the number of groups of the reference distribution data set, lnW (k) represents the cluster dispersion, r represents the reference data, and j represents the lower bound in the summation symbol;
s213, judging whether the formula (1) is satisfied
Figure 528593DEST_PATH_IMAGE002
Figure 270022DEST_PATH_IMAGE003
；
If yes, the optimal number of clusters is 1, that is, all the real-time initial data are good data, and if not, S22 is executed.
S22, calculating elbow break angles by utilizing an elbow criterion and determining the most appropriate clustering number;
specifically, the step of calculating the elbow fold angle by using the elbow criterion and determining the most appropriate clustering number comprises the following steps:
s221, setting an initial value of k = k +1,k to 1, respectively calculating a clustering dispersion lnW (k) of the real-time initial data, and calculating an elbow angle of a clustering dispersion curve at each clustering point, wherein a calculation formula of the elbow angle is as follows:
Figure 777227DEST_PATH_IMAGE014
；
s222, if k makes it satisfy
Figure 114667DEST_PATH_IMAGE015
And k is the optimal cluster number, and if the k is not satisfied, the step returns to S221.
And S23, detecting and identifying bad data in the real-time initial data according to the optimal cluster number.
Specifically, the detection and identification of the bad data in the real-time initial data according to the optimal clustering number comprises the following steps:
s231, judging whether the optimal clustering number is 1, if so, indicating that all real-time initial data are good data, and if not, calculating the average value of the data in each cluster;
s232, setting the cluster with the minimum average value as the cluster of normal data, and determining the rest as the cluster formed by bad data, so as to realize the detection and identification of the bad data in the real-time initial data.
S3, processing the acquired real-time initial data by using an improved radial basis function neural network model based on the identification result of the abnormal data to obtain corrected real-time data;
the method comprises the following steps of obtaining an identification result based on abnormal data, processing the obtained real-time initial data by using an improved radial basis function neural network model to obtain corrected real-time data, wherein the identification result based on the abnormal data comprises the following steps:
s31, optimizing a pre-constructed Radial Basis Function (RBF) neural network model by using the improved ant colony algorithm to obtain an improved radial basis function neural network model and training;
the Ant Colony Algorithm (Ant Colony Algorithm) is a heuristic bionic evolution Algorithm based on population from the beginning, which is provided by simulating the behavior of collective Ant routing in nature. The method is a general heuristic algorithm and can be used for solving various different combination optimization problems. By using the method to solve the TSP problem, the distribution problem and the job-shop scheduling problem, a better test result is obtained. Although the strict theoretical basis of the ant colony algorithm is not established, the emerging intelligent evolution bionic algorithm shows vigorous vigor and is a method with great development prospect. The idea of the ant colony algorithm is as follows: the feasible solution of the problem to be solved is represented by the walking route of the ants, each ant independently searches the feasible solution in the solution space, the higher the solution quality is, the more hormones are left on the walking route, along with the advancing of the algorithm, the more information hormones on the route representing the better solution are, the ants learn to gradually increase, and finally the whole ant colony is concentrated on the route representing the optimal solution under the action of positive feedback. The optimal solution is also found.
In this embodiment, although the ant colony algorithm has a strong capability of finding a better solution, the search time is long, and a stagnation phenomenon easily occurs. When the ant colony algorithm is used for clustering, if the clustering scale is large, the information quantity of solutions which are never searched can be reduced and even close to 0 due to the existence of the volatility coefficient P, so that the global searching capability of the algorithm is greatly reduced. When P is too large and the information amount of the solution increases, the possibility that the previously searched solution is selected is too large, and the global searching capability of the algorithm is also affected, so that the convergence speed of the algorithm is not reduced by only reducing P. Accordingly, in the present embodiment, the value of P is changed by an adaptive method, and its initial value is set to P =0.9. When the optimum value found by the algorithm does not improve significantly over N cycles, P varies according to the following equation:
Figure 289296DEST_PATH_IMAGE016
；
in the formula (I), the compound is shown in the specification,P low to representPSo as to preventPToo small results in a decrease in the convergence rate of the algorithm, and t represents time.
Specifically, the optimizing the radial basis function neural network model by using the improved ant colony algorithm to obtain the improved radial basis function neural network model includes the following steps:
s311, acquiring normal data of a period of time before bad data appear in the real-time initial data as an input sample;
s312, clustering the input samples by using the improved ant colony clustering algorithm to obtain a clustering center as a central value of a radial basis function neural network hidden layer unit;
clustering operation is carried out by adopting an improved ant colony clustering algorithm, and the specific clustering process is as follows:
(1) Initializing basic parameters, wherein the basic parameters comprise the number N of samples, the attributes m of the samples, the clustering radius r and the allowable error value delta 0 And a reference probability p 0 ；
(2) Calculating the distance between any two samples by adopting a distance calculation formula, wherein the calculation formula is as follows:
Figure 906354DEST_PATH_IMAGE017
；
(3) Initializing the information amount on each path:
Figure 432013DEST_PATH_IMAGE018
；
(4) Computingx i Cluster tox j The probability of clustering is as follows:
Figure 41986DEST_PATH_IMAGE019
；
if it isp ij （t）≥p 0 Then, thenx i Belong tox j Class, otherwise, two classes are divided;
(5) Calculation andx j clustering centers to one class:
Figure 585968DEST_PATH_IMAGE020
；
wherein J represents and x j Number of samples classified as a class;
calculating the total error:
Figure 623195DEST_PATH_IMAGE021
；
if it isε>ε 0 Then proceed, otherwise stop the algorithm if the error requirement is met,ε 0 is the standard error;
(6) Calculating the distance d between each sample and the new cluster center ij Modifying the volatilization coefficient P of the pheromone to modify the pheromone, wherein the calculation formula of the modifying pheromone is as follows:
Figure 636150DEST_PATH_IMAGE022
；
in the formula (I), the compound is shown in the specification,
Figure 800546DEST_PATH_IMAGE023
which represents the original pheromone of the message,
Figure 949768DEST_PATH_IMAGE024
representing a new pheromone, wherein Q represents absolute real-time initial data monitored by a monitoring terminal;
(7) Repeatedly executing the steps (4) to (6) until the error is met, and terminating the algorithm;
s313, adjusting the weight from the hidden layer to the output layer by using a pseudo-inverse algorithm;
specifically, the weight is obtained by using a pseudo-inverse algorithm, and a calculation formula of the pseudo-inverse algorithm is as follows:
Figure 423474DEST_PATH_IMAGE025
；
in the formula (I), the compound is shown in the specification,
Figure 392567DEST_PATH_IMAGE026
the weight value is represented by a weight value,da matrix of expected outputs is represented that is,orepresents the output matrix, T represents transpose;
s314, calculating the output value of each hidden layer unit, normalizing the output value and judging the contribution of each hidden layer unit to network output;
in particular, for each input and output: (x n ，y n ) And calculating the hidden layer unit output each time:
Figure 865049DEST_PATH_IMAGE027
；
finding out hidden layer unit with maximum absolute value of hidden layer output value
Figure 868777DEST_PATH_IMAGE028
And calculating the output value of each hidden layer unit:
Figure 247806DEST_PATH_IMAGE029
；
in the formula (I), the compound is shown in the specification,
Figure 720507DEST_PATH_IMAGE030
the output of the hidden layer element is represented,
Figure 475973DEST_PATH_IMAGE031
the output of the hidden layer cell representing the largest absolute value,x n the input is represented by a representation of the input,
Figure 334208DEST_PATH_IMAGE032
the allowed error values representing the k-th hidden layer node, k =1,2, …, n,
Figure 664564DEST_PATH_IMAGE033
representing the cluster centers or center vectors of the hidden layer node basis functions, exp representing the headers of the exponential function expressions.
S315, under the condition of meeting the network output error, removing the hidden layer unit of which the output contribution of the whole network is smaller than a preset threshold value by using a cutting method, so as to realize the optimization of the network structure;
specifically, for M consecutive inputs and outputs, if
Figure 608249DEST_PATH_IMAGE034
Then the hidden layer sheet is deletedThe primitive and decrements the number of hidden layer elements by one.
And S316, obtaining an improved radial basis function neural network model based on the optimized network parameters and structure, and training.
S32, acquiring normal data of bad data in the real-time initial data at the previous moment, and outputting prediction data of the current moment through an improved radial basis function neural network model;
and S33, taking the predicted data of the current time as a correction result of the bad data to obtain corrected real-time data.
S4, analyzing the variation of the corrected real-time data by using a threshold comparison method, and realizing the screening of the real-time data;
the method for analyzing the variation of the corrected real-time data by using the threshold comparison method to realize the screening of the real-time data comprises the following steps of:
s41, collecting the corrected real-time data and acquiring preset reference data;
s42, calculating the variation of the corrected real-time data and preset reference data;
specifically, the variation in this embodiment may be calculated by a least square method, and the calculation formula is:
Figure 901827DEST_PATH_IMAGE035
；
wherein Δ is the amount of change, U a ' is real-time data, U a Is the reference data.
S43, comparing the variable quantity with a preset variable quantity threshold value, and keeping the real-time data with the variable quantity larger than the preset variable quantity threshold value as screened real-time data.
Specifically, in actual operation, data of some parameters generally do not change much in adjacent acquisition cycles, and by setting a threshold, a variation calculated by comparing the data of the parameters in the current acquisition cycle with reference data is compared with a preset threshold, and a parameter of which the variation is larger than the variation threshold in the acquisition cycle is focused.
And S5, transmitting the screened real-time data between the monitoring terminal and the control terminal through a communication network.
In summary, according to the technical scheme of the invention, the threshold comparison method is used for analyzing the variation of the corrected real-time data, so as to realize the screening of the real-time data, thereby effectively reducing the transmission quantity of the data, alleviating the problem of mass accumulation of the data, reducing the pressure of data processing, further effectively improving the efficiency of data transmission, ensuring the real-time performance of the data transmission, and better meeting the use requirements of the power industry.
In addition, the detection and identification of abnormal data can be realized by using an elbow criterion method based on an intermittent statistical algorithm, and the abnormal data in the real-time initial data can be corrected by using an improved radial basis function neural network model, so that the influence on the safe operation of equipment caused by abnormal monitoring data can be effectively avoided.
In addition, the invention optimizes the network parameters and the network structure of the pre-constructed radial basis function neural network model by utilizing the improved ant colony algorithm, can effectively solve the defects of the traditional neural network, and corrects the abnormal data of the real-time initial data by utilizing the improved radial basis function neural network model, thereby effectively improving the correction accuracy rate of the abnormal data in the real-time initial data and better meeting the requirement of abnormal data correction.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (8)

1. An Internet of things communication method for improving data transmission efficiency is characterized by comprising the following steps:
s1, identity authentication between a monitoring terminal and a control terminal is realized by using an identity authentication method;
s2, acquiring real-time initial data of the monitoring terminal, and identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm;
s3, processing the acquired real-time initial data by using an improved radial basis function neural network model based on the identification result of the abnormal data to obtain corrected real-time data;
s4, analyzing the variation of the corrected real-time data by using a threshold comparison method, and screening the real-time data;
s5, transmission of screened real-time data is achieved between the monitoring terminal and the control terminal through a communication network;
the method for processing the acquired real-time initial data by using the improved radial basis function neural network model based on the identification result of the abnormal data to obtain the corrected real-time data comprises the following steps:
s31, optimizing a pre-constructed radial basis function neural network model by using the improved ant colony algorithm to obtain an improved radial basis function neural network model and training the improved radial basis function neural network model; which comprises the following steps:
s311, acquiring normal data of a period of time before bad data appear in the real-time initial data as an input sample;
s312, clustering the input samples by using the improved ant colony clustering algorithm to obtain a clustering center serving as a center value of a hidden layer unit of the radial basis function neural network;
s313, adjusting the weight from the hidden layer to the output layer by using a pseudo-inverse algorithm;
s314, calculating the output value of each hidden layer unit, normalizing the output value, and judging the contribution of each hidden layer unit to network output;
s315, under the condition of meeting the network output error, removing the hidden layer unit of which the output contribution of the whole network is smaller than a preset threshold value by using a cutting method, and realizing the optimization of the network structure;
s316, obtaining an improved radial basis function neural network model based on the optimized network parameters and structure and training;
s32, acquiring normal data of bad data in the real-time initial data at the previous moment, and outputting prediction data of the current moment through an improved radial basis function neural network model;
and S33, taking the predicted data of the current time as a correction result of the bad data to obtain corrected real-time data.
2. The communication method of the internet of things for improving the data transmission efficiency according to claim 1, wherein the identity authentication between the monitoring terminal and the control terminal by using the identity authentication method comprises the following steps:
s11, the monitoring terminal initiates a connection request to the control terminal, and the control terminal receives the connection request and sends an encrypted digital certificate to the monitoring terminal;
s12, the monitoring terminal receives the encrypted digital certificate and carries out decryption verification, and after verification is passed, the identity of the control terminal is confirmed to be reliable;
s13, the monitoring terminal generates a new random number, takes out a public key in the digital certificate of the control terminal, and sends the new random number encrypted by the public key to the control terminal;
s14, the control terminal receives the encrypted new random number, decrypts the new random number by using a private key of the control terminal to obtain the new random number, and uses the new random number as a session key with the monitoring terminal;
and S15, establishing communication connection between the monitoring terminal and the control terminal by using the new random number as a session key between the monitoring terminal and the control terminal.
3. The communication method of the internet of things for improving the data transmission efficiency according to claim 1, wherein the step of acquiring real-time initial data of the monitoring terminal and identifying abnormal data by using an elbow criterion method based on an intermittent statistical algorithm comprises the following steps:
s21, acquiring real-time initial data of the monitoring terminal, and judging existence of bad data in the real-time initial data;
s22, calculating elbow folding angles by utilizing an elbow criterion and determining the most appropriate clustering number;
and S23, detecting and identifying bad data in the real-time initial data according to the optimal clustering number.
4. The internet of things communication method for improving data transmission efficiency according to claim 3, wherein the step of obtaining real-time initial data of the monitoring terminal and judging existence of bad data in the real-time initial data comprises the following steps:
s211, acquiring real-time initial data of the monitoring terminal, and clustering the real-time initial data and the reference data respectively to obtain clustering dispersion of the clustering data;
s212, calculating intermittent values g by using intermittent calculation formulas ap (1) And g ap (2) Wherein, the intermittent calculation formula is as follows:
Figure 971768DEST_PATH_IMAGE001
（1）；
wherein k represents the number of clusters, F represents the number of groups of the reference distribution data set, lnW (k) represents the cluster dispersion, r represents the reference data, and j represents the lower bound in the summation symbol;
s213, judging whether the formula (1) is satisfied
Figure 214530DEST_PATH_IMAGE002
Figure 125723DEST_PATH_IMAGE003
；
If yes, the optimal number of clusters is 1, that is, all the real-time initial data are good data, and if not, S22 is executed.
5. The internet of things communication method for improving data transmission efficiency according to claim 4, wherein the step of calculating the elbow angle by using an elbow criterion and determining the most suitable cluster number comprises the following steps:
s221, setting an initial value of k = k +1,k to 1, respectively calculating a clustering dispersion lnW (k) of the real-time initial data, and calculating an elbow angle of a clustering dispersion curve at each clustering point, wherein a calculation formula of the elbow angle is as follows:
Figure 530160DEST_PATH_IMAGE004
；
s222, if k makes it satisfy
Figure 910326DEST_PATH_IMAGE005
And k is the optimal cluster number, and if the k is not satisfied, the step returns to S221.
6. The communication method of the internet of things for improving the data transmission efficiency according to claim 5, wherein the detecting and identifying bad data in the real-time initial data according to the optimal clustering number comprises the following steps:
s231, judging whether the optimal clustering number is 1, if so, indicating that all real-time initial data are good data, and if not, calculating the average value of the data in each cluster;
s232, setting the cluster with the minimum average value to belong to the cluster of normal data, and determining the rest clusters to be clusters formed by bad data, so that the detection and identification of the bad data in the real-time initial data are realized.
7. The communication method of the internet of things for improving data transmission efficiency according to claim 1, wherein the calculation formula of the pseudo-inverse algorithm is as follows:
Figure 441932DEST_PATH_IMAGE006
；
in the formula (I), the compound is shown in the specification,
Figure 692785DEST_PATH_IMAGE007
the weight value is represented by a weight value,da matrix of expected outputs is represented that is,orepresents the output matrix, T represents transpose;
the calculation formula of the output value of the hidden layer unit is as follows:
Figure 799281DEST_PATH_IMAGE008
；
in the formula (I), the compound is shown in the specification,
Figure 401164DEST_PATH_IMAGE009
the output of the hidden layer element is represented,
Figure 234997DEST_PATH_IMAGE010
the output of the hidden layer cell representing the largest absolute value,x n the input is represented by a representation of the input,
Figure 605935DEST_PATH_IMAGE011
the allowed error values, k =1,2, …, n,
Figure 617753DEST_PATH_IMAGE012
representing the cluster centers or center vectors of the hidden layer node basis functions, exp representing the headers of the exponential function expressions.
8. The internet of things communication method for improving data transmission efficiency according to claim 1, wherein the step of analyzing the variation of the corrected real-time data by using a threshold comparison method to realize the screening of the real-time data comprises the following steps:
s41, collecting the corrected real-time data and acquiring preset reference data;
s42, calculating the variable quantity of the corrected real-time data and preset reference data;
s43, comparing the variable quantity with a preset variable quantity threshold value, and keeping the real-time data with the variable quantity larger than the preset variable quantity threshold value as screened real-time data.
CN202211576831.4A 2022-12-09 2022-12-09 Internet of things communication method for improving data transmission efficiency Active CN115580545B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211576831.4A CN115580545B (en) 2022-12-09 2022-12-09 Internet of things communication method for improving data transmission efficiency

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211576831.4A CN115580545B (en) 2022-12-09 2022-12-09 Internet of things communication method for improving data transmission efficiency

Publications (2)

Publication Number Publication Date
CN115580545A true CN115580545A (en) 2023-01-06
CN115580545B CN115580545B (en) 2023-04-07

Family

ID=84590324

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211576831.4A Active CN115580545B (en) 2022-12-09 2022-12-09 Internet of things communication method for improving data transmission efficiency

Country Status (1)

Country Link
CN (1) CN115580545B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017120968A1 (en) * 2016-01-17 2017-07-20 衣佳鑫 Internet of things-based data acquisition method and system
CN113612645A (en) * 2021-08-13 2021-11-05 纪琳 Internet of things data processing method and system
CN113918411A (en) * 2021-09-18 2022-01-11 中标慧安信息技术股份有限公司 Terminal equipment management and control method and system based on edge calculation
CN113934720A (en) * 2021-10-18 2022-01-14 北京八分量信息科技有限公司 Data cleaning method and equipment and computer storage medium
CN114444033A (en) * 2021-12-07 2022-05-06 国网山东省电力公司电力科学研究院 Data security protection system and method based on Internet of things
CN114998792A (en) * 2022-05-30 2022-09-02 中用科技有限公司 Safety monitoring method with AI network camera
EP4064078A1 (en) * 2021-03-23 2022-09-28 Accenture Global Solutions Limited Utilizing a neural network model to generate a reference image based on a combination of images

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017120968A1 (en) * 2016-01-17 2017-07-20 衣佳鑫 Internet of things-based data acquisition method and system
EP4064078A1 (en) * 2021-03-23 2022-09-28 Accenture Global Solutions Limited Utilizing a neural network model to generate a reference image based on a combination of images
CN113612645A (en) * 2021-08-13 2021-11-05 纪琳 Internet of things data processing method and system
CN113918411A (en) * 2021-09-18 2022-01-11 中标慧安信息技术股份有限公司 Terminal equipment management and control method and system based on edge calculation
CN113934720A (en) * 2021-10-18 2022-01-14 北京八分量信息科技有限公司 Data cleaning method and equipment and computer storage medium
CN114444033A (en) * 2021-12-07 2022-05-06 国网山东省电力公司电力科学研究院 Data security protection system and method based on Internet of things
CN114998792A (en) * 2022-05-30 2022-09-02 中用科技有限公司 Safety monitoring method with AI network camera

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王晶: "电力系统异常数据检测辨识方法综述", 《电力与能源》 *

Also Published As

Publication number Publication date
CN115580545B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
US11418526B2 (en) Detecting anomalous network activity
Ortet Lopes et al. Towards effective detection of recent DDoS attacks: A deep learning approach
CN109413710B (en) Clustering method and device of wireless sensor network based on genetic algorithm optimization
CN113378899B (en) Abnormal account identification method, device, equipment and storage medium
CN111783845B (en) Hidden false data injection attack detection method based on local linear embedding and extreme learning machine
CN117041019B (en) Log analysis method, device and storage medium of content delivery network CDN
CN113765880B (en) Power system network attack detection method based on space-time correlation
CN116545954B (en) Communication gateway data transmission method and system based on internet of things
Ghalehgolabi et al. Intrusion detection system using genetic algorithm and data mining techniques based on the reduction
CN110224852A (en) Network security monitoring method and device based on HTM algorithm
CN113886225A (en) Unknown industrial control protocol-oriented fuzzy test system and method
CN112738109A (en) Web attack detection method and device
CN114785548A (en) Virtual flow anomaly detection method and system based on weighted adaptive ensemble learning and intelligent flow monitoring platform
CN115580545B (en) Internet of things communication method for improving data transmission efficiency
CN116738354B (en) Method and system for detecting abnormal behavior of electric power Internet of things terminal
CN111291078B (en) Domain name matching detection method and device
CN112491627A (en) Network quality real-time analysis method and device
CN112149121A (en) Malicious file identification method, device, equipment and storage medium
CN112929369B (en) Distributed real-time DDoS attack detection method
CN115766081A (en) Abnormal flow detection method and device for power industrial control cloud platform
CN114362972B (en) Botnet hybrid detection method and system based on flow abstract and graph sampling
CN113254485A (en) Real-time data flow abnormity detection method and system
Ku et al. Intrusion detection based on self-adaptive differential evolutionary extreme learning machine
CN112052449A (en) Malicious file identification method, device, equipment and storage medium
Benova et al. Web server load prediction and anomaly detection from hypertext transfer protocol logs

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