CN109639485A - The monitoring method and device of electricity consumption acquisition communication link - Google Patents
The monitoring method and device of electricity consumption acquisition communication link Download PDFInfo
- Publication number
- CN109639485A CN109639485A CN201811522073.1A CN201811522073A CN109639485A CN 109639485 A CN109639485 A CN 109639485A CN 201811522073 A CN201811522073 A CN 201811522073A CN 109639485 A CN109639485 A CN 109639485A
- Authority
- CN
- China
- Prior art keywords
- data
- communication link
- time
- upload
- predicted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/40—Monitoring; Testing of relay systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/1851—Systems using a satellite or space-based relay
- H04B7/18513—Transmission in a satellite or space-based system
-
- 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
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
Abstract
The invention discloses the monitoring methods and device of a kind of electricity consumption acquisition communication link.Wherein, this method comprises: obtaining Time series forecasting model, wherein the data volume for the upload data that Time series forecasting model is used to receive communication link in preset time period is predicted；By input time, the data sequence of section is input to Time series forecasting model, obtains the predicted data amount for the upload data that communication link within a preset period of time receives；When the time reaching preset time period, the actual amount of data for the upload data that communication link is an actually-received within a preset period of time is obtained；Based on predicted data amount and actual amount of data, the state of communication link is monitored.The present invention, which solves, to be needed to acquire communication link by increasing satellite communication transmission load to monitor electricity consumption in the prior art, and satellite communication transmission is caused to load larger technical problem.
Description
Technical field
The present invention relates to the communications fields, in particular to the monitoring method and device of a kind of electricity consumption acquisition communication link.
Background technique
The short message bi-directional communication function of Beidou satellite navigation system is applied successfully in electric system, especially
It is to apply not only to provide safe data biography in wireless public network coverage hole, Shao Renqu and the electricity consumption acquisition system in depopulated zone
Transmission link, while also providing convenience for monitoring, the inspection etc. of transmission line of electricity.Electricity consumption acquisition system based on Beidou communication is general
It is made of Database Systems, main station system, Beidou communication terminal, concentrator, electricity consumption acquisition terminal.Electric power data is packaged into
376.1 protocol massages analytical decompositions are sent one at Big Dipper short message by Beidou communication terminal by 376.1 protocol massages per minute
Secondary message, and heartbeat message is regularly sent to main website；Main station system receives assembling message information, and stores to Database Systems and supply
Background process.
The communication link of electricity consumption acquisition system, the link including acquisition terminal to concentrator, concentrator-big-dipper satellite-master
Communication link between standing.Big Dipper short message communication is a kind of unreliable transport protocol, and data packet transmission success rate is usually
95.5%, increasing with Beidou terminal, transmission channel conflict increases, and transmission success rate will continue to decline, while natural
The many factors such as disaster and weather conditions, equipment fault, the communication that all can lead to electricity consumption acquisition system occur time-out error, influence
Data transmission efficiency.It needs suitable system to be monitored communication link, find link problem in time and takes corresponding
Treatment measures.
Currently, usually used monitoring method is, main station system travels frequently over heartbeat message monitoring communication link state, this
It will aggravate the load of transmission channel.If arranging communication monitoring equipment in outfield, though it can accurately acquire electricity consumption acquisition terminal and collection
Middle device working condition, but still suffer from Shao Renqu, depopulated zone deployment, difficult in maintenance, there is still a need for by Beidou transmission
Monitoring data.
For needing to acquire communication link by increasing satellite communication transmission load to monitor electricity consumption in the prior art, cause
Satellite communication transmission loads larger problem, and currently no effective solution has been proposed.
Summary of the invention
The embodiment of the invention provides the monitoring methods and device of a kind of electricity consumption acquisition communication link, to solve the prior art
It is middle to need to acquire communication link by increasing satellite communication transmission load to monitor electricity consumption, cause satellite communication transmission load larger
The technical issues of.
According to an aspect of an embodiment of the present invention, a kind of monitoring method of communication link is provided, comprising: obtain timing
Prediction model, wherein the data volume for the upload data that Time series forecasting model is used to receive communication link in preset time period
It is predicted；By input time, the data sequence of section is input to Time series forecasting model, obtains communication link within a preset period of time
The predicted data amount of the upload data received；When the time reaching preset time period, communication link is obtained in preset time period
The actual amount of data for the upload data being inside an actually-received；Based on predicted data amount and actual amount of data, to the shape of communication link
State is monitored.
Further, data area is determined based on predicted data amount；According to data area and actual amount of data and prediction
Deviation between data volume, is monitored communication link.
Further, before obtaining Time series forecasting model, Time series forecasting model is constructed, wherein building time series forecasting mould
Type includes: acquisition historical data, wherein historical data includes the number for the upload data that communication link is received in historical time section
According to amount；Sample data is constructed based on historical data；Preset model is trained by sample data, obtains time series forecasting mould
Type, wherein Time series forecasting model is for predicting the data volume that subsequent time period communication link receives.
Further, the electricity consumption that the heartbeat record and concentrator that concentrator is obtained from preset main website data road upload is adopted
Collect data；Acquiring data by the electricity consumption that the heartbeat record and concentrator of concentrator upload determines communication link in historical time section
The data volume of the upload data received.
Further, the status data of communication link is determined by following formula:
Wherein, the data for the upload data that communication link is received in historical time section before data (t) is used to indicate time t-1 to t
Amount, the electricity consumption acquisition data that concentrator uploads before frame is used to indicate time t-1 to t, heart_beat is for indicating the time
The heartbeat of concentrator before t-1 to t.
It further, is more than the target data of preset time from the time that historical data extracts failure-free operation；Extract mesh
The data characteristics of data is marked, and data characteristics is divided according to preset time point, obtains the corresponding number of each preset time point
According to vector, sample data is obtained；Sample data is divided into training dataset and test data set according to preset ratio.
Further, sequence is established to the deep layer Network Prediction Model of sequence, and the training data of continuous time period is concentrated
Including deep layer Network Prediction Model of the data vector sequence inputting to sequence to sequence that constitutes of data, and by GRU encoder
The input of data sequence vector is encoded；Coding result is input to GRU decoder, is got under the output of GRU decoder
The prediction result of one period；Using stochastic gradient descent method optimization to the parameter of sequence deep layer Network Prediction Model, obtain
To training result；Training result is tested using test data set, and in the case where test passes through, determines training result
For Time series forecasting model.
Further, data area is determined based on predicted data amount, comprising: using the minimum value of current slot as data
The lower limit value of range, and using the maximum value of current slot as the upper limit value of data area；It is true according to upper limit value and lower limit value
Determine data area；Communication link is carried out according to actual amount of data and predicted data amount, comprising: be divided into data area multiple
Region determines the region that data volume belongs to more；The region according to belonging to actual amount of data, it is determined whether issue warning information.
According to an aspect of an embodiment of the present invention, a kind of monitoring device of communication link state is provided, comprising: data
Flux prediction model training module, for obtaining Time series forecasting model, wherein Time series forecasting model is used for in preset time period
The data volume for the upload data that communication link receives is predicted；Data traffic prediction module, for by input time section
Data sequence is input to Time series forecasting model, obtains the prediction number for the upload data that communication link within a preset period of time receives
According to amount；Module is obtained, is an actually-received within a preset period of time for when the time reaching preset time period, obtaining communication link
Upload data actual amount of data；Link evaluating and alarm module, for being based on predicted data amount and actual amount of data, to logical
The state of letter link is monitored.
According to an aspect of an embodiment of the present invention, a kind of storage medium is provided, storage medium includes the program of storage,
Wherein, equipment where controlling storage medium when program is run executes the monitoring method of above-mentioned communication link.
According to an aspect of an embodiment of the present invention, a kind of processor is provided, processor is for running program, wherein
Program executes the monitoring method of above-mentioned communication link when running.
In embodiments of the present invention, Time series forecasting model is obtained, wherein Time series forecasting model is used for in preset time period
The data volume for the upload data that communication link receives is predicted；By input time, the data sequence of section is input to time series forecasting
Model obtains the predicted data amount for the upload data that communication link within a preset period of time receives；When the time reaches it is default when
Between section when, obtain the actual amount of data of upload data that communication link is an actually-received within a preset period of time；Based on prediction number
According to amount and actual amount of data, the state of communication link is monitored.Above scheme passes through the upload data to predicted time section
Amount is predicted, and the upload data volume of the actual transmitting data amount of predicted time section and prediction is compared, to logical
Letter link is monitored, without in outfield deployment-specific monitoring device；Without active transmission detection data, therefore it not will increase volume
Outer satellite communication flow；Compared with conventional Time series forecasting model, the prediction model based on deep learning is more accurate.Pass through
Prediction model actively judges communication link state, solve need in the prior art by increase satellite communication transmission load come
It monitors electricity consumption and acquires communication link, satellite communication transmission is caused to load larger technical problem.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of the monitoring method of communication link state according to an embodiment of the present invention；
Fig. 2 is a kind of schematic diagram of the monitoring method of optional communication link according to an embodiment of the present invention；
Fig. 3 is a kind of schematic diagram of trained Time series forecasting model according to an embodiment of the present invention；
Fig. 4 is a kind of schematic diagram for dividing training dataset and test data set according to an embodiment of the present invention；And
Fig. 5 is the schematic diagram of the monitoring device of communication link state according to an embodiment of the present invention.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
The model that the present invention protects all should belong in member's every other embodiment obtained without making creative work
It encloses.
It should be noted that description and claims of this specification and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to the embodiment of the present invention described herein can in addition to illustrating herein or
Sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, for example, the process, method, system, product or equipment for containing a series of steps or units are not necessarily limited to
Step or unit those of is clearly listed, but may include be not clearly listed or for these process, methods, product
Or other step or units that equipment is intrinsic.
Embodiment 1
According to embodiments of the present invention, a kind of embodiment of the monitoring method of communication link state is provided, needs to illustrate
It is that step shown in the flowchart of the accompanying drawings can execute in a computer system such as a set of computer executable instructions,
Also, although logical order is shown in flow charts, and it in some cases, can be to be different from sequence execution herein
Shown or described step.
Fig. 1 is the flow chart of the monitoring method of communication link state according to an embodiment of the present invention, as shown in Figure 1, the party
Method includes the following steps:
Step S102 obtains Time series forecasting model, wherein Time series forecasting model is used for communication link in preset time period
The data volume of the upload data received is predicted.
Specifically, above-mentioned Time series forecasting model can be data traffic prediction model, for predicting the number of future time section
According to flow.
In an alternative embodiment, above-mentioned data traffic prediction model can be obtained by training, and training process is such as
Under: historical data is utilized, deep learning RNN technology is based on, establishes and the network sequence of one sequence of training to sequence predicts mould
Type predicts the upload data volume that subsequent time system receives according to time and the time series data belonged to；It is pre- after saving optimization
Survey model；Incremental training and optimization are carried out to the prediction model of preservation；Wherein, the Network Prediction Model of sequence to sequence can be with base
It encodes and realizes in tensorflow.
Step S104, by input time, the data sequence of section is input to Time series forecasting model, obtains within a preset period of time
The predicted data amount for the upload data that communication link receives.
In above-mentioned steps, trained Time series forecasting model in advance is loaded, according to the current data sequence of input, prediction
The upload total amount of data of subsequent time.
Step S106 obtains communication link and is an actually-received within a preset period of time when the time reaching preset time period
Upload data actual amount of data.
Step S108 is based on predicted data amount and actual amount of data, is monitored to the state of communication link.
Preferably, the communication link in the present embodiment is that the electricity consumption of Beidou acquires communication link.
In the above scheme, it according to actual amount of data, is compared with the data volume of prediction, to determine whether issuing alarm
Information.In an alternative embodiment, if actual amount of data is substantially less than predicted data amount, or it is significantly higher than prediction data
Amount, then issue warning information.
From the foregoing, it will be observed that the above embodiments of the present application obtain Time series forecasting model, wherein Time series forecasting model is used for default
The data volume for the upload data that communication link receives in period is predicted；By input time, the data sequence of section is input to
Time series forecasting model obtains the predicted data amount for the upload data that communication link within a preset period of time receives；It is reached when the time
When to preset time period, the actual amount of data for the upload data that communication link is an actually-received within a preset period of time is obtained；Base
In predicted data amount and actual amount of data, the state of communication link is monitored.Above scheme passes through to predicted time section
It uploads data volume to be predicted, and the upload data volume of the actual transmitting data amount of predicted time section and prediction is compared
It is right, to be monitored to communication link, without in outfield deployment-specific monitoring device；Without active transmission detection data, therefore
Additional satellite communication flow is not will increase；Compared with conventional Time series forecasting model, the prediction model based on deep learning is more
It is accurate to add.By prediction model, communication link state is actively judged, solve and needed in the prior art by increasing satellite communication
Traffic load acquires communication link to monitor electricity consumption, and satellite communication transmission is caused to load larger technical problem.
As a kind of optional embodiment, it is based on predicted data amount and actual amount of data, the state of communication link is carried out
Monitoring, comprising: according to the deviation between data area and actual amount of data and predicted data amount, communication link is supervised
It surveys.
In an alternative embodiment, can be using predicted data amount as median, and determined on the basis of median
Upper limit value and lower limit value, the range determined using upper limit value and lower limit value is data area.If actual data value is in data area
It is interior, it is determined that actual data value is normal, if actual data value except data area, issues warning information.
As a kind of optional embodiment, before obtaining Time series forecasting model, the above method further include: building timing is pre-
Survey model, wherein building Time series forecasting model includes: acquisition historical data, wherein historical data includes communication link in history
The data volume for the upload data that period receives；Sample data is constructed based on historical data；By sample data to default mould
Type is trained, and obtains Time series forecasting model, wherein Time series forecasting model is for predicting that subsequent time period communication link receives
Data volume.
Fig. 2 is a kind of schematic diagram of the monitoring method of optional communication link according to an embodiment of the present invention, one kind can
In the embodiment of choosing, as shown in connection with fig. 2, data preparation is carried out first, for obtaining training the training data of Time series forecasting model,
Then model is established according to training data and be trained, further according to trained Time series forecasting model to the number of subsequent time period
It is predicted according to amount, finally link is evaluated and alarmed according to the result of prediction and subsequent time period actual data volume.
As a kind of optional embodiment, historical data is obtained, comprising: obtain concentrator from preset main website data road
Heartbeat record and concentrator upload electricity consumption acquire data；It is adopted by the electricity consumption that the heartbeat record and concentrator of concentrator upload
Collection data determine the data volume for the upload data that communication link is received in historical time section.
As a kind of optional embodiment, it is true that data are acquired by the electricity consumption that the heartbeat record and concentrator of concentrator upload
Determine the data volume for the upload data that communication link is received in historical time section, comprising: communication link is determined by following formula
Status data:Wherein, data (t) is for logical before indicating time t-1 to t
The data volume for the upload data that letter link is received in historical time section, frame is for indicating concentrator before time t-1 to t
The electricity consumption of upload acquires data, and heart_beat is used to indicate the heartbeat of concentrator before time t-1 to t.
For above-mentioned steps for carrying out data preparation, Fig. 3 is a kind of trained Time series forecasting model according to an embodiment of the present invention
Schematic diagram, in an alternative embodiment, as shown in connection with fig. 3, the step of data preparation includes: mobile phone historical data, spy
Value indicative is extracted and data prediction, specific as follows: extraction time marks from main website database, terminal seat area numbers,
Zhou Shangchuan total amount of data, each terminal upload data amount that the moon uploads data count amount, current slot receives, terminal heartbeat note
Record, each terminal beams power；Collect statistics current slot uploads total amount of data, average beam power；Each characteristic is done
Standardization processing zooms to [0,1] section；And according to Fixed Time Interval align data.It is special for the statistics being unrelated with the time
Sign then stretches and is filled into each time point, such as Zhou Shangchuan total amount of data, the moon upload data count amount etc..T moment receives upper
Pass data volume are as follows:
Above formula indicates that the upload data volume of t period is equal to time t-1
The summation of the electricity consumption acquisition data and heartbeat data that are uploaded to concentrator each between t, unit byte (Byte)；
Wherein, each feature field of required extraction is described as follows:
(1) time marks: statistical data or the specific time for predicting numerical value.
(2) working day weekly: recording current working day, is formed with the time series of Zhou Weiyi circulation, makes model
Study uploads data volume and the relationship between working day.
(3) terminal upload data total amount: it can reflect the transmission state of Beidou terminal indirectly.
(4) terminal seat area numbers: the statistical nature of different regions may be different.
(5) Zhou Shangchuan total amount of data: the historical data feature of long period.
(6) moon uploads data count amount: the historical data feature of long period.
(7) beam power: current slot, the average value of each Beidou terminal 1#~6# beam power intensity, reaction transmission
Whether signal condition is good.
As a kind of optional embodiment, sample data is constructed based on historical data, comprising: extract without reason from historical data
The time of barrier operation is more than the target data of preset time；The data characteristics of target data is extracted, and by data characteristics according to pre-
If time point divide, obtain the corresponding data vector of each preset time point, obtain sample data；By sample data according to pre-
If ratio cut partition is at training dataset and test data set.
Historical data required for model training is extracted by data preparation module in a kind of optional embodiment；From going through
The longer period data record that failure-free operation is chosen in history data, extracts each data characteristics, divides data according to time point,
Each time point corresponds to a data vector Xt；By data according to 7:3 ratio cut partition be training dataset and test data set.
The present invention also provides the modes that the following two kinds divides training dataset and test data set, and Fig. 4 is according to this hair
The schematic diagram of a kind of the division training dataset and test data set of bright embodiment as shown in connection with fig. 4 can be by Walk-
The partition strategy of forward split or the partition strategy of Side-by-side split divide sample data, strategy
Time series data is divided into training dataset and validation data set.In Walk-forward split partition strategy, every construction
One training data vector as training dataset, need to be by time point relative between one prediction of training time forward movement
Every suitable for the less situation of data set.In the partition strategy based on Side-by-side split, data set is divided into independent
Two parts, for a part for training, another part is suitable for the more situation of data for verifying.
As a kind of optional embodiment, preset model is trained by sample data, obtains Time series forecasting model,
Include: the deep layer Network Prediction Model for establishing sequence to sequence, the training data of continuous time period is concentrated to the data structure for including
At deep layer Network Prediction Model of the data vector sequence inputting to sequence to sequence, and by GRU (Gated Recurrent
Unit) encoder encodes the input of data sequence vector；Coding result is input to GRU decoder, gets GRU decoding
The prediction result of the subsequent time period of device output；Use the deep layer neural network forecast mould of stochastic gradient descent method optimization to sequence
The parameter of type, obtains training result；Training result is tested using test data set, and in the case where test passes through,
Determine that training result is Time series forecasting model.
Specifically, deep layer network (Seq2Seq) model of the sequence established in above-mentioned steps to sequence is initial model.Even
The training dataset of continuous period can be the data of preceding 24 time points (being divided between two time points one hour).
In an alternative embodiment, as shown in connection with fig. 3, the step of establishing simultaneously training pattern includes: to establish Initial R NN
Model, model training and verifying, specific as follows: the prediction model of load pre-training is (i.e. as the Seq2Seq of initial model first
Model)；If current point in time is t, according to the data at preceding 24 time points of data preparation module input, next time is predicted
The upload data volume that point t+1 is received；Seq2Seq sequential forecasting models are realized using tensorflow.The model is by two masters
It partially to constitute, respectively encoder and decoder.Encoder hidden layer is a GRU, and decoder hidden layer is also one
GRU.Input data vector sequence in continuous 24 hours, [Xt, Xt+1..., Xt+23], by GRU as encoder to data sequence into
Row coding, obtains a series of encoding state values；Then, decoder according to the historical data of input and a series of encoding state values into
Row decoding, the upload data volume sequence [Y of output subsequent time predictiont+1, Yt+2..., Yt+24].Activation primitive selects tanh, makes
With stochastic gradient descent method Optimal Parameters, prediction error is minimized.
Hardware aspect, server is minimum need to configure 2 pieces of calculating GPU, and depth model training can be rapidly completed.Model one
Denier training is completed, i.e., reusable, is not necessarily to each run all training patterns.At the appointed time or specified event occurs constantly,
Newly-increased data can be used, incremental training and optimization are carried out to prediction model.
After establishing model, can with as shown in figure 3, predicted by data volume of the model to subsequent time period,
And prediction result is compared with actual data volume.
As a kind of optional embodiment, data area is determined based on predicted data amount, comprising: most by current slot
Lower limit value of the small value as data area, and using the maximum value of current slot as the upper limit value of data area；According to the upper limit
Value and lower limit value determine data area；Communication link is carried out according to actual amount of data and predicted data amount, comprising: by data model
It encloses and is divided into multiple regions, determine the region that data volume belongs to more；The region according to belonging to actual amount of data, it is determined whether issue and accuse
Alert information.
In an alternative embodiment, above-mentioned steps can be executed by link evaluating and alarm module, link evaluating with
Alarm module is with model predication value for expected normal value, and the maximum value and minimum value obtained using statistics is value range, to chain
Line state is evaluated, and is counted in a natural day, each t moment (such as one moment of per half an hour), minimum value and maximum
The average value of value constructs daily average minimum sequence { Mint } and average maximum sequence { Maxt }.T moment is [minimum
Value Mint, predicted value Vt] and [predicted value Vt, maximum value Maxt], six sections are averagely divided into, section reference numeral is followed successively by
[1,2,3,4,5,6]；The primitive rule of link state evaluation is, if actual measured value falls in 3,4 two sections, to be positive
Often；Falling in 2nd area is general warning message, and numerical value is relatively low, and falling in 5th area is general warning message, and numerical value is higher；It falls in or lower than 1
Area triggers significant alarm, and link state is prompted to be substantially less than the same time value of history；It falls in or is higher than 6th area and trigger minor alarm, prompt
There is excessive terminal to be communicated, communication link is busy.
Embodiment 2
According to embodiments of the present invention, a kind of embodiment of the monitoring device of communication link state is provided, Fig. 5 is according to this
The schematic diagram of the monitoring device of the communication link state of inventive embodiments, as shown in figure 5, the device includes:
Data traffic prediction model training module 50, for obtaining Time series forecasting model, wherein Time series forecasting model is used for
The data volume for the upload data that communication link in preset time period receives is predicted.
Data traffic prediction module 52 is obtained for the data sequence of input time section to be input to Time series forecasting model
The predicted data amount for the upload data that communication link receives within a preset period of time.
Module 54 is obtained, for it is practical within a preset period of time to obtain communication link when the time reaching preset time period
The actual amount of data of the upload data received.
Link evaluating and alarm module 56, for being based on predicted data amount and actual amount of data, to the state of communication link
It is monitored.
As a kind of optional embodiment, link evaluating and alarm module, comprising: submodule is determined, for based on prediction
Data volume determines data area；Submodule is monitored, for according between data area and actual amount of data and predicted data amount
Deviation, communication link is monitored.
As a kind of optional embodiment, above-mentioned apparatus further include: building module, for obtain Time series forecasting model it
Before, construct Time series forecasting model, wherein building module includes: acquisition submodule, for obtaining historical data, wherein history number
According to the data volume for including the upload data that communication link is received in historical time section；Submodule is constructed, for being based on history number
According to building sample data；Training submodule obtains time series forecasting mould for being trained by sample data to preset model
Type, wherein Time series forecasting model is for predicting the data volume that subsequent time period communication link receives.
As a kind of optional embodiment, acquisition submodule includes: acquiring unit, for from preset main website data road
The electricity consumption that the heartbeat record and concentrator for obtaining concentrator upload acquires data；First determination unit, for passing through concentrator
The electricity consumption acquisition data that heartbeat record and concentrator upload determine the upload data that communication link is received in historical time section
Data volume.
As a kind of optional embodiment, the first determination unit includes: computation subunit, for being determined by following formula
The status data of communication link:Wherein, data (t) is for indicating time t-1
The data volume for the upload data that communication link is received in historical time section before to t, frame for indicate time t-1 to t it
The electricity consumption that preceding concentrator uploads acquires data, and heart_beat is used to indicate the heartbeat of concentrator before time t-1 to t.
As a kind of optional embodiment, constructing submodule includes: extraction unit, for extracting fault-free from historical data
The time of operation is more than the target data of preset time；Extraction unit, for extracting the data characteristics of target data, and by data
Feature is divided according to preset time point, is obtained the corresponding data vector of each preset time point, is obtained sample data；It divides single
Member, for sample data to be divided into training dataset and test data set according to preset ratio.
As a kind of optional embodiment, training submodule includes: to establish unit, the deep layer for establishing sequence to sequence
Network Prediction Model, input unit, the data vector for concentrating the data for including to constitute the training data of continuous time period
Deep layer Network Prediction Model of the sequence inputting to sequence to sequence, and the input of data sequence vector is compiled by GRU encoder
Code；Acquiring unit gets the pre- of the subsequent time period of GRU decoder output for coding result to be input to GRU decoder
Survey result；Optimize unit, for using stochastic gradient descent method optimization to the parameter of the deep layer Network Prediction Model of sequence,
Obtain training result；Second determination unit, for being tested using test data set training result, and pass through in test
In the case of, determine that training result is Time series forecasting model.
As a kind of optional embodiment, determine that submodule includes: third determination unit, for by current slot most
Lower limit value of the small value as data area, and using the maximum value of current slot as the upper limit value of data area；According to the upper limit
Value and lower limit value determine data area；Monitoring submodule includes: division unit, for data area to be divided into multiple regions,
Determine the region that data volume belongs to more；The region according to belonging to actual amount of data, it is determined whether issue warning information.
Embodiment 3
According to embodiments of the present invention, a kind of storage medium is provided, storage medium includes the program of storage, wherein in institute
State the monitoring that equipment where controlling the storage medium when program operation executes communication link described in any one of embodiment 1
Method.
Embodiment 4
According to embodiments of the present invention, a kind of processor is provided, processor is for running program, wherein described program fortune
The monitoring method of communication link described in any one of embodiment 1 is executed when row.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
In the above embodiment of the invention, it all emphasizes particularly on different fields to the description of each embodiment, does not have in some embodiment
The part of detailed description, reference can be made to the related descriptions of other embodiments.
In several embodiments provided herein, it should be understood that disclosed technology contents can pass through others
Mode is realized.Wherein, the apparatus embodiments described above are merely exemplary, such as the division of the unit, Ke Yiwei
A kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or components can combine or
Person is desirably integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Between coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or communication link of unit or module
It connects, can be electrical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
On unit.It can some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can for personal computer, server or network equipment etc.) execute each embodiment the method for the present invention whole or
Part steps.And storage medium above-mentioned includes: that USB flash disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic or disk etc. be various to can store program code
Medium.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (11)
1. a kind of monitoring method of electricity consumption acquisition communication link characterized by comprising
Obtain Time series forecasting model, wherein the Time series forecasting model is used to receive communication link described in preset time period
To the data volumes of upload data predicted；
By input time, the data sequence of section is input to the Time series forecasting model, obtains described logical in the preset time period
The predicted data amount for the upload data that letter link receives；
When the time reaching the preset time period, obtain what the communication link was an actually-received in the preset time period
Upload the actual amount of data of data；
Based on the predicted data amount and the actual amount of data, the state of the communication link is monitored.
2. the method according to claim 1, wherein be based on the predicted data amount and the actual amount of data,
The state of the communication link is monitored, comprising:
Data area is determined based on the predicted data amount；
According to the deviation between the data area and the actual amount of data and the predicted data amount, to the communication
Link is monitored.
3. the method according to claim 1, wherein the method is also wrapped before obtaining Time series forecasting model
It includes: constructing the Time series forecasting model, wherein constructing the Time series forecasting model includes:
Obtain historical data, wherein the historical data includes the upload data that communication link is received in historical time section
Data volume；
Sample data is constructed based on the historical data；
Preset model is trained by sample data, obtains the Time series forecasting model, wherein the Time series forecasting model
The data volume received for predicting communication link described in subsequent time period.
4. according to the method described in claim 3, it is characterized in that, obtaining historical data, comprising:
The electricity consumption that the heartbeat record and the concentrator that concentrator is obtained from preset main website data road upload acquires data；
Acquiring data by the electricity consumption that the heartbeat record of the concentrator and the concentrator upload determines communication link in history
The data volume for the upload data that period receives.
5. according to the method described in claim 4, it is characterized in that, passing through the heartbeat record of the concentrator and the concentrator
The electricity consumption acquisition data of upload determine the data volume for the upload data that communication link is received in historical time section, comprising:
The status data of the communication link is determined by following formula:
Wherein, data (t) is for indicating the upload number that time t-1 is received to t foregoing description communication link in historical time section
According to data volume, frame is used to indicate the electricity consumption acquisition data that time t-1 is uploaded to t foregoing description concentrator, heart_beat
For indicating the heartbeat of time t-1 to t foregoing description concentrator.
6. according to the method described in claim 3, it is characterized in that, constructing sample data based on the historical data, comprising:
The time for extracting failure-free operation from the historical data is more than the target data of preset time；
The data characteristics of the target data is extracted, and the data characteristics is divided according to preset time point, is obtained each
The corresponding data vector of preset time point, obtains sample data；
The sample data is divided into training dataset and test data set according to preset ratio.
7. according to the method described in claim 6, obtaining it is characterized in that, be trained by sample data to preset model
The Time series forecasting model, comprising:
Establish sequence to sequence deep layer Network Prediction Model,
The data vector sequence inputting for the training data of continuous time period being concentrated the data for including constitute is to the sequence to sequence
The deep layer Network Prediction Model of column, and the data vector sequence inputting is encoded by GRU encoder；
Coding result is input to GRU decoder, gets the prediction result of the subsequent time period of the GRU decoder output；
Using stochastic gradient descent method optimize the sequence to sequence deep layer Network Prediction Model parameter, obtain train knot
Fruit；
The training result is tested using the test data set, and in the case where test passes through, determines the instruction
Practicing result is the Time series forecasting model.
8. according to the method described in claim 2, it is characterized in that,
Data area is determined based on the predicted data amount, comprising: using the minimum value of current slot as the data area
Lower limit value, and using the maximum value of the current slot as the upper limit value of the data area；According to the upper limit value and
The lower limit value determines the data area；
The communication link is carried out according to the actual amount of data and the predicted data amount, comprising: divide data area
For multiple regions, the region that the data volume belongs to more is determined；
According to region belonging to the actual amount of data, it is determined whether issue warning information.
9. a kind of monitoring device of electricity consumption acquisition communication link state characterized by comprising
Data traffic prediction model training module, for obtaining Time series forecasting model, wherein the Time series forecasting model for pair
The data volume for the upload data that the communication link receives in preset time period is predicted；
The data traffic prediction module is obtained for the data sequence of input time section to be input to the Time series forecasting model
The predicted data amount of the upload data received to the communication link described in the preset time period；
Module is obtained, for when the time reaching the preset time period, obtaining the communication link in the preset time period
The actual amount of data for the upload data being inside an actually-received；
Link evaluating and alarm module, for being based on the predicted data amount and the actual amount of data, to the communication link
State be monitored.
10. a kind of storage medium, which is characterized in that the storage medium includes the program of storage, wherein run in described program
When control the storage medium where equipment perform claim require any one of 1 to 8 described in communication link monitoring method.
11. a kind of processor, which is characterized in that the processor is for running program, wherein right of execution when described program is run
Benefit require any one of 1 to 8 described in communication link monitoring method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811522073.1A CN109639485A (en) | 2018-12-13 | 2018-12-13 | The monitoring method and device of electricity consumption acquisition communication link |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811522073.1A CN109639485A (en) | 2018-12-13 | 2018-12-13 | The monitoring method and device of electricity consumption acquisition communication link |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109639485A true CN109639485A (en) | 2019-04-16 |
Family
ID=66073410
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811522073.1A Pending CN109639485A (en) | 2018-12-13 | 2018-12-13 | The monitoring method and device of electricity consumption acquisition communication link |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109639485A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110658905A (en) * | 2019-09-23 | 2020-01-07 | 珠海格力电器股份有限公司 | Early warning method, early warning system and early warning device for equipment running state |
| CN111767275A (en) * | 2020-06-28 | 2020-10-13 | 北京林克富华技术开发有限公司 | Data processing method and device and data processing system |
| CN114257344A (en) * | 2022-02-28 | 2022-03-29 | 国网浙江省电力有限公司金华供电公司 | Feeder terminal converged communication method and device for converged client front-end equipment |
| CN117292530A (en) * | 2023-11-27 | 2023-12-26 | 深圳龙电华鑫控股集团股份有限公司 | Carrier communication data acquisition efficiency optimization method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101263499A (en) * | 2005-07-11 | 2008-09-10 | 布鲁克斯自动化公司 | Intelligent condition monitoring and fault diagnostic system |
| CN101483547A (en) * | 2009-02-12 | 2009-07-15 | 中国人民解放军信息工程大学 | Evaluation method and system for network burst affair |
| US20170257304A1 (en) * | 2016-03-07 | 2017-09-07 | General Electric Company | Systems and methods for monitoring system performance and availability |
| CN108270608A (en) * | 2017-01-04 | 2018-07-10 | 中国科学院声学研究所 | A kind of foundation of link prediction model and link prediction method |
| CN108306756A (en) * | 2017-12-21 | 2018-07-20 | 国网北京市电力公司 | One kind being based on electric power data network holography assessment system and its Fault Locating Method |
-
2018
- 2018-12-13 CN CN201811522073.1A patent/CN109639485A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101263499A (en) * | 2005-07-11 | 2008-09-10 | 布鲁克斯自动化公司 | Intelligent condition monitoring and fault diagnostic system |
| CN101483547A (en) * | 2009-02-12 | 2009-07-15 | 中国人民解放军信息工程大学 | Evaluation method and system for network burst affair |
| US20170257304A1 (en) * | 2016-03-07 | 2017-09-07 | General Electric Company | Systems and methods for monitoring system performance and availability |
| CN108270608A (en) * | 2017-01-04 | 2018-07-10 | 中国科学院声学研究所 | A kind of foundation of link prediction model and link prediction method |
| CN108306756A (en) * | 2017-12-21 | 2018-07-20 | 国网北京市电力公司 | One kind being based on electric power data network holography assessment system and its Fault Locating Method |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110658905A (en) * | 2019-09-23 | 2020-01-07 | 珠海格力电器股份有限公司 | Early warning method, early warning system and early warning device for equipment running state |
| CN110658905B (en) * | 2019-09-23 | 2023-08-04 | 珠海格力电器股份有限公司 | Early warning method, early warning system and early warning device for equipment operation state |
| CN111767275A (en) * | 2020-06-28 | 2020-10-13 | 北京林克富华技术开发有限公司 | Data processing method and device and data processing system |
| CN111767275B (en) * | 2020-06-28 | 2024-04-19 | 北京林克富华技术开发有限公司 | Data processing method and device and data processing system |
| CN114257344A (en) * | 2022-02-28 | 2022-03-29 | 国网浙江省电力有限公司金华供电公司 | Feeder terminal converged communication method and device for converged client front-end equipment |
| CN117292530A (en) * | 2023-11-27 | 2023-12-26 | 深圳龙电华鑫控股集团股份有限公司 | Carrier communication data acquisition efficiency optimization method |
| CN117292530B (en) * | 2023-11-27 | 2024-02-13 | 深圳龙电华鑫控股集团股份有限公司 | Carrier communication data acquisition efficiency optimization method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109639485A (en) | The monitoring method and device of electricity consumption acquisition communication link | |
| CN112887145B (en) | Distributed network slice fault detection method | |
| CN106846729A (en) | A kind of fall detection method and system based on convolutional neural networks | |
| CN108171259A (en) | A kind of wisdom tip water quality monitoring system and method based on NB-IoT | |
| CN106953802A (en) | A kind of network optimal route selection method based on deep learning | |
| CN113110207A (en) | Insect pest remote monitoring method and system based on sensor of Internet of things and storage medium | |
| CN105577685A (en) | Intrusion detection independent analysis method and system in cloud calculation environment | |
| CN114299723B (en) | Traffic flow prediction method | |
| CN111726768B (en) | Edge calculation-oriented reliable data collection method based on compressed sensing | |
| CN108966239A (en) | A method of relay node is disposed based on Energy volution degree | |
| CN107622308A (en) | A kind of generating equipment parameter method for early warning based on DBN networks | |
| US7877337B2 (en) | Auto-adaptive network for sensor data processing and forecasting | |
| CN112684301A (en) | Power grid fault detection method and device | |
| Alippi et al. | On-line reconstruction of missing data in sensor/actuator networks by exploiting temporal and spatial redundancy | |
| CN115099512A (en) | Freshwater fish culture intelligent management system and management method | |
| CN104092503A (en) | Artificial neural network spectrum sensing method based on wolf pack optimization | |
| CN107886678B (en) | Indoor monitoring method, medium and electronic equipment | |
| CN101477201A (en) | System and method for forecasting location of mobile object | |
| CN112382081A (en) | Traffic flow prediction method based on multiple tasks | |
| CN109166628A (en) | The method and device of locomitivity detection, storage medium, processor | |
| CN106292612B (en) | A kind of ladle baking facility on-line fault diagnosis system | |
| CN108738219B (en) | The intelligent monitor system that electric system is diagnosed based on street lamp | |
| Sung et al. | Enhance the efficient of WSN data fusion by neural networks training process | |
| CN107454611A (en) | Immune dangerous wireless sensor network fault diagnosis method based on KNN | |
| CN105897926A (en) | Wireless sensor based anti-terrorism system and method |
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 |
Application publication date: 20190416 |
|
| RJ01 | Rejection of invention patent application after publication |