CN107730117A - A kind of cable maintenance method for early warning and system based on heterogeneous data comprehensive analysis - Google Patents
A kind of cable maintenance method for early warning and system based on heterogeneous data comprehensive analysis Download PDFInfo
- Publication number
- CN107730117A CN107730117A CN201710966945.2A CN201710966945A CN107730117A CN 107730117 A CN107730117 A CN 107730117A CN 201710966945 A CN201710966945 A CN 201710966945A CN 107730117 A CN107730117 A CN 107730117A
- Authority
- CN
- China
- Prior art keywords
- power cable
- data
- cable data
- neural networks
- convolutional neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012423 maintenance Methods 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000004458 analytical method Methods 0.000 title claims abstract description 20
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 57
- 238000012545 processing Methods 0.000 claims description 27
- 230000004913 activation Effects 0.000 claims description 13
- 238000011156 evaluation Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 8
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 7
- 241000208340 Araliaceae Species 0.000 claims description 7
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 7
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 7
- 235000008434 ginseng Nutrition 0.000 claims description 7
- 238000005457 optimization Methods 0.000 claims description 6
- 230000005611 electricity Effects 0.000 claims description 4
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 238000007689 inspection Methods 0.000 abstract description 4
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 238000003908 quality control method Methods 0.000 abstract description 3
- 241001269238 Data Species 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004870 electrical engineering Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of cable based on heterogeneous data comprehensive analysis to overhaul method for early warning, including:Power cable data comprising heterogeneous data are handled, obtain the power cable data of standardization and tape identification;Exercised supervision study using convolutional neural networks using the power cable data of the standardization and tape identification, be the parameter in convolutional network by maintenance decision Knowledge Internalization, obtain the convolutional neural networks model trained;The power cable data gathered in real time are analyzed and excavated using the convolutional neural networks model trained, the running status of power cable are evaluated, and push maintenance warning information.The present invention is exercised supervision using convolutional neural networks framework to be learnt to excavate, and cable running status is evaluated automatically using the convolutional neural networks model trained and pushes maintenance warning information, the flexible service work that inspection unit can be transported for power cable provides technical support, and certain reference can be also provided for the Quality Control of cable manufacturing enterprise.
Description
Technical field
Heterogeneous number is based on the present invention relates to power cable operation and maintenance technical field, and more particularly, to one kind
Method for early warning and system are overhauled according to the cable of comprehensive analysis.
Background technology
In recent years, importance of the power cable in urban distribution network is increasingly highlighted, and science maintenance is carried out to it and is related to electricity
The reliability of network operation.The repair method used in the industry at present depends primarily on the classical model such as office in electrical engineering subject
Portion's discharge test, aging analysis etc., the formulation of repair schedule is using the time as Main Basiss.Because model is easily by noise effect, no
Certainty factor is more, and this strategy generally used is highly dependent on the professional qualities of maintainer, limited reliability.In addition,
Such as shelf depreciation experiment, a kind of Strategies of Maintenance needs to power off and carried out offline, influences normal production and living.Recently what is occurred is more
Secondary cable fault also illustrates that traditional maintenance model has problem in terms of reliability and overhaul efficiency, therefore, is badly in need of having more
The maintenance method for early warning of intelligent level.
The content of the invention
The invention provides a kind of cable maintenance method for early warning and system based on heterogeneous data comprehensive analysis, to solve such as
The problem of what carries out maintenance early warning to power cable.
In order to solve the above problems, according to an aspect of the invention, there is provided a kind of be based on heterogeneous data comprehensive analysis
Cable maintenance method for early warning, it is characterised in that methods described includes:
Power cable data comprising heterogeneous data are handled, obtain the power cable number of standardization and tape identification
According to, wherein, the power cable data include:Cable attribute and status data, channel environment data and cable status evaluation number
According to;
Exercised supervision study, will be examined using convolutional neural networks using the power cable data of the standardization and tape identification
The parameter turned in DECISION KNOWLEDGE in convolutional network is repaiied, obtains the convolutional neural networks model trained;
The power cable data gathered in real time are analyzed and dug using the convolutional neural networks model trained
Pick, obtains the probabilistic information of output layer, the running status of power cable is evaluated according to the probabilistic information, and according to commenting
Valency result pushes maintenance warning information to administrative staff.
Preferably, wherein the described pair of power cable data comprising heterogeneous data are handled, standardization is obtained and with mark
The power cable data of knowledge, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable of discrete type
Data are converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable number after normalized
According to;
Power cable data after the normalized are carried out, from mark processing, the overhaul data of power cable to be believed
Breath is associated with the power cable data after normalized, obtains the power cable data of standardization and tape identification.
Preferably, wherein the power cable data to the continuous type are normalized, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;x_max
For the maximum of current attribute;X is the value of current attribute.
Preferably, wherein the power cable data using the standardization and tape identification are entered using convolutional neural networks
Row supervised learning, it is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolutional neural networks model trained,
Including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable of tape identification
Data, to it is described standardization and tape identification power cable data carry out the first preset times threshold value convolution, PReLU activation and
Pondization processing, the power cable data by convolution, PReLU activation and pondization processing are subjected to the complete of the second preset times threshold value
Connection is handled, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
Preferably, wherein using Adam algorithm optimizations using convolutional neural networks exercise supervision study in variety of processes it
Between link weight parameter.
Preferably, wherein the push mode of the maintenance warning information includes:System internal message, short message and phone.
According to another aspect of the present invention, there is provided a kind of cable maintenance early warning system based on heterogeneous data comprehensive analysis
System, it is characterised in that the system includes:
Data processing unit, for handling the power cable data comprising heterogeneous data, obtain standardization and band
The power cable data of mark, wherein, the power cable data include:Cable attribute and status data, channel environment data
With cable status evaluating data;
Model acquiring unit, for using convolutional neural networks using the power cable data of the standardization and tape identification
Exercise supervision study, is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolutional neural networks mould trained
Type;
Prewarning unit is overhauled, for the convolutional neural networks model that is trained described in utilization to the power cable that gathers in real time
Data are analyzed and excavated, and obtain the probabilistic information of output layer, the running status according to the probabilistic information to power cable
Evaluated, and maintenance warning information is pushed to administrative staff according to evaluation result.
Preferably, wherein the power cable data comprising heterogeneous data are handled by the data processing unit, obtain
Standardization and the power cable data of tape identification, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable of discrete type
Data are converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable number after normalized
According to;
Power cable data after the normalized are carried out, from mark processing, the overhaul data of power cable to be believed
Breath is associated with the power cable data after normalized, obtains the power cable data of standardization and tape identification.
Preferably, wherein the power cable data to the continuous type are normalized, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;x_max
For the maximum of current attribute;X is the value of current attribute.
Preferably, wherein the model acquiring unit, is used using the power cable data of the standardization and tape identification
Convolutional neural networks exercise supervision study, are the parameter in convolutional network by maintenance decision Knowledge Internalization, obtain the volume trained
Product neural network model, including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable of tape identification
Data, to it is described standardization and tape identification power cable data carry out the first preset times threshold value convolution, PReLU activation and
Pondization processing, the power cable data by convolution, PReLU activation and pondization processing are subjected to the complete of the second preset times threshold value
Connection is handled, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
Preferably, wherein using Adam algorithm optimizations using convolutional neural networks exercise supervision study in variety of processes it
Between link weight parameter.
Preferably, wherein the push mode of the maintenance warning information includes:System internal message, short message and phone.
The invention provides a kind of cable maintenance method for early warning and system based on heterogeneous data comprehensive analysis, by electricity
Cable attribute is handled with three kinds of status data, channel environment data and cable status evaluating data heterogeneous datas, is then used
The convolutional neural networks framework designed for these three categorical data features, which exercises supervision, to be learnt to excavate, by maintenance decision knowledge
The parameter in convolutional network is turned to, it is determined that the convolutional neural networks model trained, and utilize the convolutional neural networks trained
Model carries out continual quantitatively evaluating and the warning information of push maintenance in time to cable running status automatically.The method of the present invention
The flexible service work that inspection unit can be transported for power cable provides technical support, also can be that the Quality Control of cable manufacturing enterprise carries
For certain reference.
Brief description of the drawings
By reference to the following drawings, the illustrative embodiments of the present invention can be more fully understood by:
Fig. 1 is to overhaul method for early warning 100 according to the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention
Flow chart;And
Fig. 2 is to overhaul early warning system 200 according to the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention
Structural representation.
Embodiment
The illustrative embodiments of the present invention are introduced with reference now to accompanying drawing, however, the present invention can use many different shapes
Formula is implemented, and is not limited to embodiment described herein, there is provided these embodiments are to disclose at large and fully
The present invention, and fully pass on the scope of the present invention to person of ordinary skill in the field.Show for what is be illustrated in the accompanying drawings
Term in example property embodiment is not limitation of the invention.In the accompanying drawings, identical cells/elements are attached using identical
Map logo.
Unless otherwise indicated, term (including scientific and technical terminology) used herein has to person of ordinary skill in the field
It is common to understand implication.Further it will be understood that the term limited with usually used dictionary, be appreciated that and its
The linguistic context of association area has consistent implication, and is not construed as Utopian or overly formal meaning.
Fig. 1 is to overhaul method for early warning 100 according to the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention
Flow chart.As shown in figure 1, the maintenance method for early warning of the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention, first
Power cable data comprising heterogeneous data are handled, obtain the power cable data of standardization and tape identification;Then it is sharp
Exercised supervision study using convolutional neural networks with the power cable data of the standardization and tape identification, by maintenance decision knowledge
The parameter in convolutional network is inside turned to, obtains the convolutional neural networks model trained;Finally, the convolution trained is utilized
Neural network model, which is analyzed and excavated to the power cable data gathered in real time to administrative staff, pushes maintenance warning information,
The flexible service work that inspection unit can be transported for power cable provides technical support, also can be that the Quality Control of cable manufacturing enterprise carries
For certain reference.The cable based on heterogeneous data comprehensive analysis of invention embodiment overhauls method for early warning 100 from step 101
Place starts, and the power cable data comprising heterogeneous data are handled in step 101, obtains the electric power of standardization and tape identification
Cable data, wherein, the power cable data include:Cable attribute and status data, channel environment data and cable status
Evaluating data.
Preferably, wherein the described pair of power cable data comprising heterogeneous data are handled, standardization is obtained and with mark
The power cable data of knowledge, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable of discrete type
Data are converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable number after normalized
According to;
Power cable data after the normalized are carried out, from mark processing, the overhaul data of power cable to be believed
Breath is associated with the power cable data after normalized, obtains the power cable data of standardization and tape identification.
Preferably, wherein the power cable data to the continuous type are normalized, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;x_max
For the maximum of current attribute;X is the value of current attribute.
In embodiments of the present invention, need to gather cable attribute and status data, channel environment data, cable shape first
Three kinds of source datas of state evaluating data, wherein cable attribute and status data include Operation Condition for Power Cable online monitoring data with
Cable account data, channel environment packet is containing geological conditions, meteorological condition, surrounding enviroment data, cable status evaluating data
Include cable fault relative recording information.These three data sources can be by power cable transport inspection department door from Electric Power Enterprise Information system
Directly exported in system.Because raw data format is different, scope differs, such as:Voltage is continuous type numerical value 110kV, and electric current is
Continuous type numerical value 1000A, the severity of injuries recorded in cable record of examination for discrete type numerical value it is high, neutralize it is low.Analyzing
It is preceding that missing values are filled according to equalization mode first, and the numerical value of discrete type is subjected to continuous treatment, then successively
Pretreatment is normalized in three kinds of data and is handled from mark.
Continuous treatment is that discrete type numerical value is converted into continuous type numerical value using integer Shift Method, such as can will be high, medium and low
Three Estate is converted into numerical value 3,2,1.
Normalized is that different number ranges is normalized into [0,1] section.Specific method for normalizing is using such as
Lower formula x_new=(x-x_min)/(x_max-x_min) is calculated, i.e., subtracts current attribute minimum using current property value
Value and then divided by current attribute maximum difference.Such as:If some property value is respectively:3rd, 2,1, then current value 2 is entered
Row normalized, computational methods are:X_new=(2-1)/(3-1)=0.5.
Finally also need to carry out the numerical value after normalization from marking, be by cable status evaluation information from annotation process
Data after cable record of examination and above-mentioned standardization are matched according to recording mechanism during data acquisition, are realized to characteristic
Automatic marking, i.e., according to information such as the cable number sections recorded in record of examination by record of examination and corresponding cable attribute
Associated with status data, channel environment data, associative search can simply use the SQL based on keyword to retrieve sentence.In advance
Data after processing only include the numeric data of serialization, and numerical value is in [0,1] interval range, is easy to follow-up study to dig
Pick.
Preferably, convolutional neural networks are used using the power cable data of the standardization and tape identification in step 102
Exercise supervision study, is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolutional neural networks mould trained
Type.
Preferably, wherein the power cable data using the standardization and tape identification are entered using convolutional neural networks
Row supervised learning, it is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolutional neural networks model trained,
Including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable of tape identification
Data, to it is described standardization and tape identification power cable data carry out the first preset times threshold value convolution, PReLU activation and
Pondization processing, the power cable data by convolution, PReLU activation and pondization processing are subjected to the complete of the second preset times threshold value
Connection is handled, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
Preferably, wherein using Adam algorithm optimizations using convolutional neural networks exercise supervision study in variety of processes it
Between link weight parameter.
In embodiments of the present invention, it is contemplated that the characteristics of cable data species is more, potential relation complexity, in this method
The convolutional network overall structure of design is:[input] → [convolution → PReLU → pond] * 5 → [full connection] * 4 → Softmax,
Input layer directly receives the good data of above-mentioned pretreatment.3 groups of wave filters are designed in first convolutional layer to correspond to 3 introduces a collection numbers
According to the feature of type, every group of wave filter is only connected with a kind of data type feature, and every group of wave filter includes 20 wave filters, so
Not only enhance specific aim filtering but also reduce number of parameters.The activation primitive of neuron is used without saturated characteristic and easily had in network
There are the PReLU functions of more preferable convergence property, output unit uses Softmax, and cost function uses corresponding Softmax graders
Cost.In view of the magnanimity of cable monitoring data, parameter training algorithm uses efficient Adam algorithms, optimizes network with this
In link weight parameter between each layer, complete supervised learning process, be the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
Preferably, the convolutional neural networks model that trains described in being utilized in step 103 is to the power cable that gathers in real time
Data are analyzed and excavated, and obtain the probabilistic information of output layer, the running status according to the probabilistic information to power cable
Evaluated, and maintenance warning information is pushed to administrative staff according to evaluation result.
Preferably, wherein the push mode of the maintenance warning information includes:System internal message, short message and phone.
In embodiments of the present invention, using the above-mentioned convolutional neural networks model trained, to for certain section of cable
The measured data of collection is analyzed and excavated, and cable running status is commented according to the probabilistic information of Softmax output layers
Valency, evaluate and be identified with five danger classes of ABCDE.A identifies degree of danger highest, may be broken down in 1 hour, B
For 24 hours, C was 72 hours, and D is one week, and E is one month.Based on this danger classes, to the fortune belonging to different cut cables
Tie up unit and responsible person concerned and push maintenance early warning letter with various ways such as management system internal information, SMS, phones
Breath.
Fig. 2 is to overhaul early warning system 200 according to the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention
Structural representation.As shown in Fig. 2 the maintenance early warning system of the cable based on heterogeneous data comprehensive analysis of embodiment of the present invention
200 include:Data processing unit 201, model acquiring unit 202 and maintenance prewarning unit 203.Preferably, in data processing list
Power cable data comprising heterogeneous data are handled by member 201, obtain the power cable data of standardization and tape identification,
Wherein, the power cable data include:Cable attribute and status data, channel environment data and cable status evaluating data.
Preferably, wherein the power cable data comprising heterogeneous data are handled by the data processing unit 201,
The power cable data of standardization and tape identification are obtained, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable of discrete type
Data are converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable number after normalized
According to;
Power cable data after the normalized are carried out, from mark processing, the overhaul data of power cable to be believed
Breath is associated with the power cable data after normalized, obtains the power cable data of standardization and tape identification.
Preferably, wherein the power cable data to the continuous type are normalized, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;x_max
For the maximum of current attribute;X is the value of current attribute.
Preferably, in model acquiring unit 202, convolution is used using the power cable data of the standardization and tape identification
Neutral net exercises supervision study, is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolution god trained
Through network model.
Preferably, wherein the model acquiring unit 202, is adopted using the power cable data of the standardization and tape identification
Exercised supervision study with convolutional neural networks, be the parameter in convolutional network by maintenance decision Knowledge Internalization, obtain what is trained
Convolutional neural networks model, including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable of tape identification
Data, to it is described standardization and tape identification power cable data carry out the first preset times threshold value convolution, PReLU activation and
Pondization processing, the power cable data by convolution, PReLU activation and pondization processing are subjected to the complete of the second preset times threshold value
Connection is handled, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
Preferably, wherein using Adam algorithm optimizations using convolutional neural networks exercise supervision study in variety of processes it
Between link weight parameter.
Preferably, in maintenance prewarning unit 203, using the convolutional neural networks model trained to gathering in real time
Power cable data are analyzed and excavated, and obtain the probabilistic information of output layer, according to the probabilistic information to power cable
Running status is evaluated, and pushes maintenance warning information to administrative staff according to evaluation result.
Preferably, wherein the push mode of the maintenance warning information includes:System internal message, short message and phone.
The maintenance early warning system 200 of the cable based on heterogeneous data comprehensive analysis of embodiments of the invention is another with the present invention's
Cable based on the heterogeneous data comprehensive analysis maintenance method for early warning 100 of one embodiment is corresponding, will not be repeated here.
The present invention is described by reference to a small amount of embodiment.However, it is known in those skilled in the art, as
What subsidiary Patent right requirement was limited, except the present invention other embodiments disclosed above equally fall the present invention's
In the range of.
Normally, all terms used in the claims are all solved according to them in the usual implication of technical field
Release, unless clearly being defined in addition wherein.All references " one/described/be somebody's turn to do [device, component etc.] " are all opened ground
At least one example being construed in described device, component etc., unless otherwise expressly specified.Any method disclosed herein
Step need not all be run with disclosed accurately order, unless explicitly stated otherwise.
Claims (12)
1. a kind of cable maintenance method for early warning based on heterogeneous data comprehensive analysis, it is characterised in that methods described includes:
Power cable data comprising heterogeneous data are handled, obtain the power cable data of standardization and tape identification, its
In, the power cable data include:Cable attribute and status data, channel environment data and cable status evaluating data;
Exercised supervision study using convolutional neural networks using the power cable data of the standardization and tape identification, maintenance is determined
Plan Knowledge Internalization is the parameter in convolutional network, obtains the convolutional neural networks model trained;
The power cable data gathered in real time are analyzed and excavated using the convolutional neural networks model trained, are obtained
The probabilistic information of output layer is taken, the running status of power cable is evaluated according to the probabilistic information, and is tied according to evaluation
Fruit pushes maintenance warning information to administrative staff.
2. according to the method for claim 1, it is characterised in that the described pair of power cable data comprising heterogeneous data are carried out
Processing, the power cable data of standardization and tape identification are obtained, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable data of discrete type
Be converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable data after normalized;
To after the normalized power cable data carry out from mark processing, by the overhaul data information of power cable with
Power cable data after normalized are associated, and obtain the power cable data of standardization and tape identification.
3. according to the method for claim 2, it is characterised in that the power cable data to the continuous type are returned
One change is handled, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;X_max is to work as
The maximum of preceding attribute;X is the value of current attribute.
4. according to the method for claim 1, it is characterised in that it is described using it is described standardization and tape identification power cable
Data are exercised supervision study using convolutional neural networks, are the parameter in convolutional network by maintenance decision Knowledge Internalization, are obtained instruction
The convolutional neural networks model perfected, including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable number of tape identification
According to the standardization and convolution, PReLU activation and the pond of power cable data the first preset times threshold value of progress of tape identification
Change is handled, and the power cable data by convolution, PReLU activation and pondization processing are carried out into the complete of the second preset times threshold value connects
Processing is connect, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
5. according to the method for claim 4, it is characterised in that carried out using Adam algorithm optimizations using convolutional neural networks
Link weight parameter in supervised learning between variety of processes.
6. according to the method for claim 1, it is characterised in that the push mode of the maintenance warning information includes:System
Inside story, short message and phone.
7. a kind of cable maintenance early warning system based on heterogeneous data comprehensive analysis, it is characterised in that the system includes:
Data processing unit, for handling the power cable data comprising heterogeneous data, obtain standardization and tape identification
Power cable data, wherein, the power cable data include:Cable attribute and status data, channel environment data and electricity
Cable state evaluation data;
Model acquiring unit, for being carried out using the power cable data of the standardization and tape identification using convolutional neural networks
Supervised learning, it is the parameter in convolutional network by maintenance decision Knowledge Internalization, obtains the convolutional neural networks model trained;
Prewarning unit is overhauled, for the convolutional neural networks model that is trained described in utilization to the power cable data that gather in real time
Analyzed and excavated, obtain the probabilistic information of output layer, the running status of power cable is carried out according to the probabilistic information
Evaluation, and maintenance warning information is pushed to administrative staff according to evaluation result.
8. system according to claim 7, it is characterised in that the data processing unit, to the electricity comprising heterogeneous data
Power cable data is handled, and obtains the power cable data of standardization and tape identification, including:
The power cable data comprising heterogeneous data are filled using average completion method;
Continuous treatment is carried out to the power cable data after filling, using integer Shift Method by the power cable data of discrete type
Be converted to the power cable data of continuous type;
The power cable data of the continuous type are normalized, obtain the power cable data after normalized;
To after the normalized power cable data carry out from mark processing, by the overhaul data information of power cable with
Power cable data after normalized are associated, and obtain the power cable data of standardization and tape identification.
9. system according to claim 8, it is characterised in that the power cable data to the continuous type are returned
One change is handled, including:
X_new=(x-x_min)/(x_max-x_min),
Wherein, x_new be normalized after current attribute value;X_min is the minimum value of current attribute;X_max is to work as
The maximum of preceding attribute;X is the value of current attribute.
10. system according to claim 7, it is characterised in that the model acquiring unit, utilize the standardization and band
The power cable data of mark are exercised supervision study using convolutional neural networks, are in convolutional network by maintenance decision Knowledge Internalization
Parameter, obtain the convolutional neural networks model trained, including:
Exercised supervision study using convolutional neural networks, in the input layer input standardization and the power cable number of tape identification
According to the standardization and convolution, PReLU activation and the pond of power cable data the first preset times threshold value of progress of tape identification
Change is handled, and the power cable data by convolution, PReLU activation and pondization processing are carried out into the complete of the second preset times threshold value connects
Processing is connect, and result is exported using Softmax graders, is the ginseng in convolutional network by maintenance decision Knowledge Internalization
Number, obtains the convolutional neural networks model trained.
11. system according to claim 10, it is characterised in that entered using Adam algorithm optimizations using convolutional neural networks
Link weight parameter in row supervised learning between variety of processes.
12. system according to claim 7, it is characterised in that the push mode of the maintenance warning information includes:System
Inside story, short message and phone.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710966945.2A CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710966945.2A CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107730117A true CN107730117A (en) | 2018-02-23 |
CN107730117B CN107730117B (en) | 2021-12-21 |
Family
ID=61211622
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710966945.2A Active CN107730117B (en) | 2017-10-17 | 2017-10-17 | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107730117B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108427986A (en) * | 2018-02-26 | 2018-08-21 | 中车青岛四方机车车辆股份有限公司 | A kind of production line electrical fault prediction technique and device |
CN108537415A (en) * | 2018-03-20 | 2018-09-14 | 深圳市泰和安科技有限公司 | A kind of distribution method, the apparatus and system of online safety utilization of electric power |
CN109325537A (en) * | 2018-09-26 | 2019-02-12 | 深圳供电局有限公司 | Power utilization management method and device, computer equipment and storage medium |
CN109359866A (en) * | 2018-10-17 | 2019-02-19 | 平安国际融资租赁有限公司 | Risk hidden danger monitoring method, device and computer equipment based on leased equipment |
CN109921515A (en) * | 2019-03-12 | 2019-06-21 | 上海荷福人工智能科技(集团)有限公司 | A kind of overall power distribution management system |
CN111445103A (en) * | 2020-02-25 | 2020-07-24 | 国网河南省电力公司电力科学研究院 | Power transmission cable production quality management feedback system based on industrial internet |
CN113032458A (en) * | 2021-03-23 | 2021-06-25 | 中国人民解放军63920部队 | Method and device for determining abnormality of spacecraft |
CN113657437A (en) * | 2021-07-08 | 2021-11-16 | 中国南方电网有限责任公司 | Power grid maintenance alarm confirmation method and system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150213302A1 (en) * | 2014-01-30 | 2015-07-30 | Case Western Reserve University | Automatic Detection Of Mitosis Using Handcrafted And Convolutional Neural Network Features |
CN105426908A (en) * | 2015-11-09 | 2016-03-23 | 国网冀北电力有限公司信息通信分公司 | Convolutional neural network based substation attribute classification method |
CN106203741A (en) * | 2016-08-10 | 2016-12-07 | 国家电网公司 | Multi-element heterogeneous Data Cleaning Method for network load prediction |
CN106251059A (en) * | 2016-07-27 | 2016-12-21 | 中国电力科学研究院 | A kind of cable status appraisal procedure based on probabilistic neural network algorithm |
CN106651188A (en) * | 2016-12-27 | 2017-05-10 | 贵州电网有限责任公司贵阳供电局 | Electric transmission and transformation device multi-source state assessment data processing method and application thereof |
CN106844425A (en) * | 2016-12-09 | 2017-06-13 | 国网北京市电力公司 | The data handling system of pipeline and cable |
CN107145675A (en) * | 2017-05-17 | 2017-09-08 | 国网天津市电力公司 | Diagnosing fault of power transformer device and method based on BP neural network algorithm |
-
2017
- 2017-10-17 CN CN201710966945.2A patent/CN107730117B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150213302A1 (en) * | 2014-01-30 | 2015-07-30 | Case Western Reserve University | Automatic Detection Of Mitosis Using Handcrafted And Convolutional Neural Network Features |
CN105426908A (en) * | 2015-11-09 | 2016-03-23 | 国网冀北电力有限公司信息通信分公司 | Convolutional neural network based substation attribute classification method |
CN106251059A (en) * | 2016-07-27 | 2016-12-21 | 中国电力科学研究院 | A kind of cable status appraisal procedure based on probabilistic neural network algorithm |
CN106203741A (en) * | 2016-08-10 | 2016-12-07 | 国家电网公司 | Multi-element heterogeneous Data Cleaning Method for network load prediction |
CN106844425A (en) * | 2016-12-09 | 2017-06-13 | 国网北京市电力公司 | The data handling system of pipeline and cable |
CN106651188A (en) * | 2016-12-27 | 2017-05-10 | 贵州电网有限责任公司贵阳供电局 | Electric transmission and transformation device multi-source state assessment data processing method and application thereof |
CN107145675A (en) * | 2017-05-17 | 2017-09-08 | 国网天津市电力公司 | Diagnosing fault of power transformer device and method based on BP neural network algorithm |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108427986A (en) * | 2018-02-26 | 2018-08-21 | 中车青岛四方机车车辆股份有限公司 | A kind of production line electrical fault prediction technique and device |
CN108537415A (en) * | 2018-03-20 | 2018-09-14 | 深圳市泰和安科技有限公司 | A kind of distribution method, the apparatus and system of online safety utilization of electric power |
CN109325537A (en) * | 2018-09-26 | 2019-02-12 | 深圳供电局有限公司 | Power utilization management method and device, computer equipment and storage medium |
CN109359866A (en) * | 2018-10-17 | 2019-02-19 | 平安国际融资租赁有限公司 | Risk hidden danger monitoring method, device and computer equipment based on leased equipment |
CN109359866B (en) * | 2018-10-17 | 2024-05-03 | 塔比星信息技术(深圳)有限公司 | Risk hidden danger monitoring method and device based on leasing equipment and computer equipment |
CN109921515A (en) * | 2019-03-12 | 2019-06-21 | 上海荷福人工智能科技(集团)有限公司 | A kind of overall power distribution management system |
CN111445103A (en) * | 2020-02-25 | 2020-07-24 | 国网河南省电力公司电力科学研究院 | Power transmission cable production quality management feedback system based on industrial internet |
CN111445103B (en) * | 2020-02-25 | 2023-01-31 | 国网河南省电力公司电力科学研究院 | Power transmission cable production quality management feedback system based on industrial internet |
CN113032458A (en) * | 2021-03-23 | 2021-06-25 | 中国人民解放军63920部队 | Method and device for determining abnormality of spacecraft |
CN113657437A (en) * | 2021-07-08 | 2021-11-16 | 中国南方电网有限责任公司 | Power grid maintenance alarm confirmation method and system |
CN113657437B (en) * | 2021-07-08 | 2024-04-19 | 中国南方电网有限责任公司 | Power grid overhaul alarm confirmation method and system |
Also Published As
Publication number | Publication date |
---|---|
CN107730117B (en) | 2021-12-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107730117A (en) | A kind of cable maintenance method for early warning and system based on heterogeneous data comprehensive analysis | |
CN107179503B (en) | Wind turbine generator fault intelligent diagnosis and early warning method based on random forest | |
CN110929918B (en) | 10kV feeder fault prediction method based on CNN and LightGBM | |
CN109086889B (en) | Terminal fault diagnosis method, device and system based on neural network | |
CN110674189B (en) | Method for monitoring secondary state and positioning fault of intelligent substation | |
CN111047082A (en) | Early warning method and device for equipment, storage medium and electronic device | |
CN111680820B (en) | Distributed photovoltaic power station fault diagnosis method and device | |
CN108733966A (en) | A kind of multidimensional electric energy meter field thermodynamic state verification method based on decision woodlot | |
CN110245745B (en) | Equipment residual service life prediction method based on integrated bidirectional cyclic neural network | |
Liu et al. | Fault diagnosis of water quality monitoring devices based on multiclass support vector machines and rule-based decision trees | |
CN106932670A (en) | A kind of distribution power automation terminal method for diagnosing status based on D S evidence theories | |
CN111380686A (en) | XGboost algorithm model-based fan main bearing fault monitoring and diagnosis method | |
CN107145675A (en) | Diagnosing fault of power transformer device and method based on BP neural network algorithm | |
CN107784395A (en) | A kind of electric power line ice-covering thickness Forecasting Methodology based on LSTM artificial neural networks | |
CN111126489A (en) | Power transmission equipment state evaluation method based on ensemble learning | |
CN110766313A (en) | Cable tunnel comprehensive state evaluation method based on operation and maintenance system | |
CN110378586B (en) | Power transformation equipment defect early warning method and system based on dynamic closed-loop knowledge management | |
CN110569278A (en) | transformer defect assessment method based on big data analysis | |
CN116862081A (en) | Operation and maintenance method and system for pollution treatment equipment | |
CN117041312A (en) | Enterprise-level information technology monitoring system based on Internet of things | |
CN112305388A (en) | On-line monitoring and diagnosing method for partial discharge fault of generator stator winding insulation | |
CN116842337A (en) | Transformer fault diagnosis method based on LightGBM (gallium nitride based) optimal characteristics and COA-CNN (chip on board) model | |
CN110096723B (en) | High-voltage switch cabinet insulation state analysis method based on operation and maintenance detection big data | |
CN110489852A (en) | Improve the method and device of the wind power system quality of data | |
CN112527572A (en) | Disk failure prediction method and device, computer readable storage medium and server |
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 |