CN107831465A - A kind of intelligent electric energy meter fault judgment method based on BP neural network - Google Patents
A kind of intelligent electric energy meter fault judgment method based on BP neural network Download PDFInfo
- Publication number
- CN107831465A CN107831465A CN201711050270.3A CN201711050270A CN107831465A CN 107831465 A CN107831465 A CN 107831465A CN 201711050270 A CN201711050270 A CN 201711050270A CN 107831465 A CN107831465 A CN 107831465A
- Authority
- CN
- China
- Prior art keywords
- electric energy
- energy meter
- failure
- intelligent electric
- neural network
- 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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
- G01R35/04—Testing or calibrating of apparatus covered by the other groups of this subclass of instruments for measuring time integral of power or current
Abstract
The present invention relates to a kind of intelligent electric energy meter fault judgment method based on BP neural network, it in order to solve existing intelligent electric energy meter breakdown judge technology is difficult to be accurate to the breakdown judge to circuit components to be, the shortcomings that detection efficiency and the not high degree of accuracy and propose, including:Each group training sample data and its corresponding fault type when extraction intelligent electric energy meter breaks down;Each group training sample data are directed to respectively, are normalized, and obtain each group training sample normalization data;According to neural network input layer neuron number N to be built and neutral net output layer neuron number M to be built, the neuron number K of neutral net hidden layer to be built is obtained;Build neutral net;Judge failure component corresponding to intelligent electric energy meter failure.The present invention is applied to the fault detect of intelligent electric energy meter.
Description
Technical field
The present invention relates to intelligent electric energy meter breakdown judge technical field, more particularly to a kind of intelligence based on BP neural network
Electrical energy meter fault determination methods.
Background technology
Intelligent electric energy meter is the important component of intelligent grid, its promotion and application can effectively take precautions against electricity arrears,
The electricity charge risk of reduction company;By the standard and technical conditions of specification intelligent electric energy meter, acquisition terminal and system main website are specified
Constructive direction, the horizontal raising of electric energy metrical industry technology can be promoted, drive the development of power equipment manufacturing industry.Intelligent electric energy
The reliability of table directly affects the safe and stable and economical operation of whole acquisition system, is also directly connected to huge numbers of families residence
The power supply reliability and security of the people.Failure is occurring in practical application in intelligent electric energy meter, or even exposes batch quality and ask
Topic.
Electrical energy meter fault reason is complicated, and electric energy meter Reliability Engineering is related to substantial amounts of, long-term reliability field test data
Collect, arrange and analyze, be one long-term and hard work.The field data collection of electric energy meter is very necessary with statistics
's.It is to effectively utilize data in the life cycle for electric energy meter to collect field data, to ensure that the reliability of electric energy meter takes
Business, the fail-safe analysis and dependability parameter that according to field data, can carry out electric energy meter are assessed.The collection of field data can be with
Network analysis is carried out on electrical energy meter fault pattern and influence and failure cause.
The intelligent electric energy meter breakdown judge technology of prior art is difficult to be accurate to the breakdown judge to circuit components, detection
Efficiency and the degree of accuracy be not high, it is therefore desirable to a kind of new electrical energy meter fault judgment technology, to solve this defect.
The content of the invention
It is difficult to be accurate to circuit the invention aims to solve existing intelligent electric energy meter breakdown judge technology
The shortcomings that breakdown judge of component, detection efficiency and not high degree of accuracy, and propose a kind of intelligence based on BP neural network
Electrical energy meter fault determination methods, including:
Each group training sample data and its corresponding fault type when extraction intelligent electric energy meter breaks down;It is directed to respectively
Each group training sample data, are normalized, and obtain each group training sample normalization data;According to neutral net to be built
Input layer number N and neutral net output layer neuron number M to be built, obtain neutral net hidden layer to be built
Neuron number K;Neutral net is built, and the neutral net is trained using each group training sample normalization data,
Neutral net after being trained;Judge the component that failed corresponding to intelligent electric energy meter failure using the neutral net after training.
On the basis of above-mentioned technical proposal, the present invention can also do following improvement.
Further, the BP neural network is disposed with along input to outbound course:Input layer, hidden layer and defeated
Go out layer.
Further, the input layer includes being used to represent the error in dipping X1 of fault type, for representing fault type
Communication failure X2, the display failure X3 for representing fault type and the battery failures X4 for representing fault type.
Further, the neuron number of the hidden layer is 5.
Further, the output layer includes:Fail key componentses sampling resistor Y1, the failure pressure-sensitive electricity of key componentses
Hinder Y2, failure key componentses thermistor Y3, failure key componentses 485 chip Y4, failure key componentses optocoupler Y5, mistake
Imitate key componentses TVSY6, failure key componentses liquid crystal Y7, failure key componentses current-limiting resistance Y8, the crucial first device of failure
Part backlight Y9, failure key componentses filter capacitor Y10, failure key componentses diode Y11, failure key componentses MOS
Pipe Y12 and failure key componentses battery Y13.
Preferably, neural network input layer neuron number N=4 to be built, neutral net output layer neuron to be built
Number M=13, obtain the neuron number K=5 of neutral net hidden layer to be built.
The beneficial effects of the invention are as follows:BP neural network algorithm can be utilized to quickly determine that intelligent electric energy meter loses
The key componentses of effect, detection efficiency at least lift 20%, and Detection accuracy at least lifts 5%.
Brief description of the drawings
Fig. 1 is the step flow chart of the intelligent electric energy meter fault judgment method based on BP neural network of the present invention;
Fig. 2 is the Organization Chart of the BP neural network of the present invention.
Embodiment
Embodiment one:The intelligent electric energy meter fault judgment method based on BP neural network of present embodiment is as schemed
Shown in 1, including:
S101:Each group training sample data and its corresponding fault type when extraction intelligent electric energy meter breaks down.
S102:Each group training sample data are directed to respectively, are normalized, and obtain each group training sample normalization number
According to.
S103:According to neural network input layer neuron number N to be built and neutral net output layer neuron to be built
Number M, obtain the neuron number K of neutral net hidden layer to be built.The specific mode for determining K can be according to following any
A kind of mode determines:
(1) according to formulaTo determine, wherein k is sample number, n1For hidden layer unit number (K i.e. to be asked), i
For the constant between [0, n].
(2) according to formulaTo determine, wherein n1For hidden layer unit number (K i.e. to be asked), n is input
Unit number, m are output unit number, constants of a between [1,10].
(3) according to formula n1=log2N determines, wherein n1For hidden layer unit number (K i.e. to be asked), n is input block
Quantity.
Can all it be attempted using each way in practical problem, then the best implicit layer number of using effect carrys out structure
Build neutral net.
S104:Build neutral net, i.e., according to previous step determine input, imply, the neuron number of output layer enters
Row structure.Reuse each group training sample normalization data to be trained the neutral net, the nerve net after being trained
Network.
S105:Judge the component that failed corresponding to intelligent electric energy meter failure using the neutral net after training.This step is
Carry out the process of actual test using the model that trains, and S101 is to the process that S104 steps are training.
Embodiment two:Present embodiment is unlike embodiment one:Next, referring again to Fig. 2
Shown, it is the Organization Chart of the BP neural network of the present invention;The BP neural network of the present invention is set successively along input to outbound course
It is equipped with:Input layer, hidden layer and output layer.
Other steps and parameter are identical with embodiment one.
Embodiment three:Present embodiment is unlike embodiment one or two:The input layer includes
For representing the error in dipping X1 of fault type, the communication failure X2 for representing fault type, for representing fault type
Show the failure X3 and battery failures X4 for representing fault type.
Other steps and parameter are identical with embodiment one or two.
Embodiment four:Unlike one of present embodiment and embodiment one to three:The god of hidden layer
It is that 5 other steps and parameter are identical with one of embodiment one to three through first number.
Embodiment five:Unlike one of present embodiment and embodiment one to four:Output layer includes:
Failure key componentses sampling resistor Y1, failure key componentses piezo-resistance Y2, failure key componentses thermistor Y3, mistake
Imitate the chip Y4 of key componentses 485, failure key componentses optocoupler Y5, failure key componentses TVSY6, failure key componentses
Liquid crystal Y7, failure key componentses current-limiting resistance Y8, failure key componentses backlight Y9, failure key componentses filter capacitor
Y10, failure key componentses diode Y11, failure key componentses metal-oxide-semiconductor Y12 and failure key componentses battery Y13.
Thus, according to neural network input layer neuron number N to be built and neutral net output neuron number to be built
Mesh M, obtain the neuron number K of neutral net hidden layer to be built;Neutral net is built, judges intelligence electricity using neutral net
Can failure component corresponding to table failure.
Other steps and parameter are identical with one of embodiment one to four.
Embodiment six:Unlike one of present embodiment and embodiment one to five:
In above-mentioned practical application, neural network input layer neuron number N=4 to be built, neutral net output to be built
Layer neuron number M=13, obtain the neuron number K=5 of neutral net hidden layer to be built.
As shown in table 1 below, a fault type may correspond to multiple failure components, so a fault type is corresponding more
Individual output, corresponding failure component values output is 1, and non-failed component output is 0.
Table 1
Other steps and parameter are identical with one of embodiment one to five.
The present invention can also have other various embodiments, in the case of without departing substantially from spirit of the invention and its essence, this area
Technical staff works as can make various corresponding changes and deformation according to the present invention, but these corresponding changes and deformation should all belong to
The protection domain of appended claims of the invention.
Claims (6)
1. a kind of intelligent electric energy meter fault judgment method based on BP neural network, it is characterised in that comprise the following steps:
S101:Each group training sample data and its corresponding fault type when extraction intelligent electric energy meter breaks down;
S102:Each group training sample data are directed to respectively, are normalized, and obtain each group training sample normalization data;
S103:According to neural network input layer neuron number N to be built and neutral net output layer neuron number to be built
M, obtain the neuron number K of neutral net hidden layer to be built;
S104:Neutral net is built, and the neutral net is trained using each group training sample normalization data, is obtained
Neutral net after training;
S105:Judge the component that failed corresponding to intelligent electric energy meter failure using the neutral net after training.
2. the intelligent electric energy meter fault judgment method according to claim 1 based on BP neural network, it is characterised in that institute
BP neural network is stated to be disposed with along input to outbound course:Input layer, hidden layer and output layer.
3. the intelligent electric energy meter fault judgment method according to claim 2 based on BP neural network, it is characterised in that institute
Stating input layer includes being used for representing the error in dipping X1 of fault type, the communication failure X2 for representing fault type, for table
Show the display failure X3 of the fault type and battery failures X4 for representing fault type.
4. the intelligent electric energy meter fault judgment method according to claim 2 based on BP neural network, it is characterised in that institute
The neuron number for stating hidden layer is 5.
5. the intelligent electric energy meter fault judgment method according to claim 3 based on BP neural network, it is characterised in that institute
Stating output layer includes:Fail key componentses sampling resistor Y1, failure key componentses piezo-resistance Y2, failure key componentses
Thermistor Y3, failure key componentses 485 chip Y4, failure key componentses optocoupler Y5, failure key componentses TVSY6,
Fail key componentses liquid crystal Y7, failure key componentses current-limiting resistance Y8, failure key componentses backlight Y9, the crucial member of failure
Device filter capacitor Y10, failure key componentses diode Y11, failure key componentses metal-oxide-semiconductor Y12 and the crucial member of failure
Device battery Y13.
6. the intelligent electric energy meter fault judgment method according to claim 5 based on BP neural network, it is characterised in that treat
Neural network input layer neuron number N=4, neutral net output layer neuron number M=13 to be built are built, structure is treated in acquisition
Build the neuron number K=5 of neutral net hidden layer.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711050270.3A CN107831465A (en) | 2017-10-31 | 2017-10-31 | A kind of intelligent electric energy meter fault judgment method based on BP neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711050270.3A CN107831465A (en) | 2017-10-31 | 2017-10-31 | A kind of intelligent electric energy meter fault judgment method based on BP neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107831465A true CN107831465A (en) | 2018-03-23 |
Family
ID=61651191
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711050270.3A Pending CN107831465A (en) | 2017-10-31 | 2017-10-31 | A kind of intelligent electric energy meter fault judgment method based on BP neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107831465A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108732528A (en) * | 2018-05-28 | 2018-11-02 | 国网福建省电力有限公司电力科学研究院 | A kind of digitalized electrical energy meter method for diagnosing faults based on depth confidence network |
CN109635950A (en) * | 2018-11-30 | 2019-04-16 | 国网上海市电力公司 | The electric energy meter method for monitoring operation states clustered based on genetic algorithm and corporations |
CN110244256A (en) * | 2019-07-22 | 2019-09-17 | 广东工业大学 | A kind of intelligent electric energy meter fault recognition method, device and equipment |
CN111650548A (en) * | 2020-04-23 | 2020-09-11 | 宁夏隆基宁光仪表股份有限公司 | Intelligent electric energy meter metering data remote online monitoring system and method |
CN111814900A (en) * | 2020-07-20 | 2020-10-23 | 安徽南瑞中天电力电子有限公司 | Electric energy meter fault classification method and device based on MATLAB neural network |
CN112816934A (en) * | 2021-03-01 | 2021-05-18 | 云南电网有限责任公司电力科学研究院 | Method and system for judging error self-monitoring accuracy and timeliness of electric energy meter |
CN115936166A (en) * | 2022-09-28 | 2023-04-07 | 海南电网有限责任公司 | Electric energy meter calibration error analysis and prediction method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102707256A (en) * | 2012-06-20 | 2012-10-03 | 北京航空航天大学 | Fault diagnosis method based on BP-Ada Boost nerve network for electric energy meter |
CN105023042A (en) * | 2015-07-10 | 2015-11-04 | 国家电网公司 | User electricity stealing suspicion analyzing device and method based on big data neural network algorithm |
CN105353255A (en) * | 2015-11-27 | 2016-02-24 | 南京邮电大学 | Transformer fault diagnosis method based on neural network |
CN106779069A (en) * | 2016-12-08 | 2017-05-31 | 国家电网公司 | A kind of abnormal electricity consumption detection method based on neutral net |
-
2017
- 2017-10-31 CN CN201711050270.3A patent/CN107831465A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102707256A (en) * | 2012-06-20 | 2012-10-03 | 北京航空航天大学 | Fault diagnosis method based on BP-Ada Boost nerve network for electric energy meter |
CN105023042A (en) * | 2015-07-10 | 2015-11-04 | 国家电网公司 | User electricity stealing suspicion analyzing device and method based on big data neural network algorithm |
CN105353255A (en) * | 2015-11-27 | 2016-02-24 | 南京邮电大学 | Transformer fault diagnosis method based on neural network |
CN106779069A (en) * | 2016-12-08 | 2017-05-31 | 国家电网公司 | A kind of abnormal electricity consumption detection method based on neutral net |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108732528A (en) * | 2018-05-28 | 2018-11-02 | 国网福建省电力有限公司电力科学研究院 | A kind of digitalized electrical energy meter method for diagnosing faults based on depth confidence network |
CN109635950A (en) * | 2018-11-30 | 2019-04-16 | 国网上海市电力公司 | The electric energy meter method for monitoring operation states clustered based on genetic algorithm and corporations |
CN110244256A (en) * | 2019-07-22 | 2019-09-17 | 广东工业大学 | A kind of intelligent electric energy meter fault recognition method, device and equipment |
CN111650548A (en) * | 2020-04-23 | 2020-09-11 | 宁夏隆基宁光仪表股份有限公司 | Intelligent electric energy meter metering data remote online monitoring system and method |
CN111814900A (en) * | 2020-07-20 | 2020-10-23 | 安徽南瑞中天电力电子有限公司 | Electric energy meter fault classification method and device based on MATLAB neural network |
CN112816934A (en) * | 2021-03-01 | 2021-05-18 | 云南电网有限责任公司电力科学研究院 | Method and system for judging error self-monitoring accuracy and timeliness of electric energy meter |
CN115936166A (en) * | 2022-09-28 | 2023-04-07 | 海南电网有限责任公司 | Electric energy meter calibration error analysis and prediction method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107831465A (en) | A kind of intelligent electric energy meter fault judgment method based on BP neural network | |
CN108051709A (en) | Transformer state online evaluation analysis method based on artificial intelligence technology | |
CN106408687B (en) | A kind of automobile engine fault early warning method based on machine learning method | |
CN103218695A (en) | Secondary equipment intelligence state evaluation diagnostic system and method thereof | |
CN107422266A (en) | A kind of method for diagnosing faults and device of high capacity cell energy-storage system | |
CN106291234A (en) | A kind of transmission line of electricity internal fault external fault based on convolutional neural networks judges and fault phase-selecting method | |
CN105117839B (en) | A kind of power system weak link identification method based on cascading failure | |
CN107978807A (en) | A kind of battery detecting and maintaining method and system | |
CN109919495A (en) | Photovoltaic plant monitoring data exception and degree judgment method based on threshold calculations | |
CN106291233A (en) | A kind of fault phase-selecting method based on convolutional neural networks | |
CN114977483A (en) | Fault diagnosis system for intelligent power grid regulation and control equipment | |
CN106779269A (en) | Reliability assessment method for secondary system of intelligent substation based on VC Method | |
CN109615273A (en) | A kind of electric car electrically-charging equipment method for evaluating state and system | |
CN106802599A (en) | A kind of diagnosing fault of power transformer system based on expert database | |
CN106656040A (en) | Fault diagnosis method and apparatus for photovoltaic modules in photovoltaic power station | |
CN202421448U (en) | Relay detecting device for nuclear power stations | |
CN107222497A (en) | Network traffic anomaly monitor method and electronic equipment | |
CN117036103A (en) | Electric power system operation situation prediction method based on LSTM (least squares) circulating neural network | |
CN107807325B (en) | Railway track circuit reliability analysis system and method based on multi-state theory | |
CN109062740A (en) | A kind of auxiliary Check System and method based on direct fault location | |
CN109861214B (en) | Method and system for judging weak line with stable transient power angle of regional power grid | |
CN107122829A (en) | A kind of method that utilization virtual sample trains Neural Network Diagnosis transformer fault | |
CN107194465A (en) | A kind of method that utilization virtual sample trains Neural Network Diagnosis transformer fault | |
CN103647677B (en) | Power communication network state detection method | |
CN106709817A (en) | Microgrid equipment full life cycle management system and fault early warning 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: 20180323 |
|
RJ01 | Rejection of invention patent application after publication |