CN117633612A - Electric energy metering method and system based on big data self-diagnosis - Google Patents
Electric energy metering method and system based on big data self-diagnosis Download PDFInfo
- Publication number
- CN117633612A CN117633612A CN202311448205.1A CN202311448205A CN117633612A CN 117633612 A CN117633612 A CN 117633612A CN 202311448205 A CN202311448205 A CN 202311448205A CN 117633612 A CN117633612 A CN 117633612A
- Authority
- CN
- China
- Prior art keywords
- data
- electric energy
- energy metering
- diagnosis
- module
- 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
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000004092 self-diagnosis Methods 0.000 title claims abstract description 26
- 238000013135 deep learning Methods 0.000 claims abstract description 10
- 238000013480 data collection Methods 0.000 claims abstract description 4
- 238000013500 data storage Methods 0.000 claims abstract description 4
- 238000003745 diagnosis Methods 0.000 claims description 23
- 230000002159 abnormal effect Effects 0.000 claims description 15
- 238000012937 correction Methods 0.000 claims description 13
- 238000007405 data analysis Methods 0.000 claims description 12
- 238000004140 cleaning Methods 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 9
- 238000005259 measurement Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000013499 data model Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 abstract description 10
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention relates to the technical field of electric energy metering, and discloses an electric energy metering method and system based on big data self-diagnosis, wherein the method comprises the following steps: step 1, data collection and storage: a large amount of electrical energy metering data is collected and stored. According to the electric energy metering method and system based on big data self-diagnosis, the LSTM model for multivariate time series prediction is built through the big data deep learning Tensorflow and Keras deep learning library, future values are predicted based on the predicted values of past values and past values instead of only using discrete uncorrelated features such as seasons, the model can be more stable through the predicted values of the past values, meanwhile, the whole test data set can be predicted after the model is fitted through the Walk-Forward Split method and the Side-by-Side Split method, the prediction is combined with the test data set, the scale of the test data set is adjusted, the error score of the model can be calculated through the initial predicted values and the actual values, and the accuracy and reliability of electric energy metering are effectively improved.
Description
Technical Field
The invention relates to the technical field of electric energy metering, in particular to an electric energy metering method and system based on big data self-diagnosis.
Background
Along with the development of electric power technology, intelligent ammeter has begun to popularize, and intelligent ammeter brings very big facility for measurement data acquisition, but receives environment, artificial, design factor's influence, and intelligent ammeter also can break down, if through manual investigation, the work load is too big, and can not in time discover the trouble, brings inconvenience for the user, brings the loss for the power sector, if can in time discover the trouble through the analysis processing to measurement data, will bring big facility.
However, the collection and processing of a large amount of metering data face the requirements of large data volume and high processing speed, so that high requirements are put forward for the existing electric energy metering, and the current detection method is biased to the chip level and aims at solving the faults of a hardware sampling loop, but all error influencing factors of the electric energy metering are not considered, for example, the metering accuracy of an electric energy metering system is influenced due to the fact that metering module devices fail, meter calibrating parameters change, electric energy metering clock faults are caused by software faults or artificial damage, and the like.
Disclosure of Invention
(one) solving the technical problems
Aiming at the defects of the prior art, the invention provides an electric energy metering method and system based on big data self-diagnosis, which have the advantages of realizing automatic detection and alarm of related software faults and the like by adding the big data self-diagnosis function, and solve the problems that the current detection method is biased to a chip level and is focused on solving the faults of a hardware sampling loop, but all error influencing factors of electric energy metering are not considered, such as meter correction parameter change, electric energy metering clock faults and the like caused by failure of metering module devices, software faults or artificial damage, so that the metering accuracy of an electric energy metering system is influenced.
(II) technical scheme
In order to achieve the above purpose, the present invention provides the following technical solutions: an electric energy metering method based on big data self-diagnosis comprises the following steps:
step 1, data collection and storage: collecting and storing electric energy metering data, and ensuring the reliability and the integrity of the data in the collecting process;
step 2, data cleaning and abnormality detection: cleaning and preprocessing the acquired data, removing abnormal values and noise, and detecting possible abnormal conditions;
step 3, data analysis and model establishment: analyzing and modeling the electric energy metering data by utilizing a big data analysis technology, and establishing a model suitable for electric energy metering;
step 4, fault diagnosis and correction: performing fault diagnosis on the electric energy metering device based on the established model, detecting potential faults, and providing corresponding correction measures;
step 5, metering calibration and precision evaluation: the accuracy and the reliability of the metering device are ensured by calibrating the metering device, and the accuracy degree of metering is evaluated;
step 6, automatic decision and alarm: based on fault diagnosis and calibration results, the system automatically makes decisions and alarms, and timely alerts the metering device of abnormal conditions.
Preferably, the electric energy metering data in the step 1 comprises power grid data and personal user data, wherein the power grid data is subjected to data acquisition through an SCADA data acquisition system, and the personal user data is subjected to data acquisition through an AMI advanced measurement system.
Preferably, the abnormal value of the data cleaning in the step 2 includes a repeated value of the data and an illegal value of the data, and the abnormal value in the data batch is identified by the data cleaning in the step 2 through a box diagram.
Preferably, the information source-dependent information source of the model establishment in the step 3 comprises local features and global features, the model establishment in the step 3 builds an LSTM model for multivariable time series prediction through big data deep learning, and the specific steps of the model establishment are as follows:
s1, converting an original data set into data suitable for time sequence prediction;
s2, processing the data and adapting the data to an LSTM model for the multivariate time series prediction problem;
s3, making predictions and analyzing results.
Preferably, the dividing method of the training set and the verification set of the time sequence comprises a Walk-Forward Split method and a Side-by-Side Split method, wherein the Side-by-Side Split method is used for dividing the data set into independent different subsets, one part is completely used for training, the other part is completely used for verification, and the verification set divided by the Walk-Forward Split method is used for adjusting parameters.
Preferably, the fault of the electric energy metering device in the step 4 includes that a current loop breaks down, an electric energy meter breaks down, a clock recording result is inaccurate, and a supply voltage of the battery becomes small.
The electric energy metering system based on the big data self-diagnosis is suitable for the electric energy metering method based on the big data self-diagnosis, and comprises a data acquisition module, a data preprocessing module, a feature extraction module, a data analysis and modeling module, a fault diagnosis module, a metering correction module, an automatic decision and alarm module and a control computer.
Preferably, the data acquisition module is used for acquiring original data generated by the electric energy metering device, the data preprocessing module is used for preprocessing the acquired original data, the feature extraction module is used for extracting features from the preprocessed data, the features comprise power factors, harmonic content and waveform quality indexes so as to describe the performance of the electric energy metering device, the data analysis and modeling module is used for analyzing and modeling the extracted features by using big data to establish a correlation model between the performance of the electric energy metering device and the data features, the fault diagnosis module is used for extracting self-diagnosis signals and the model through a sampling loop band-pass filter to perform fault diagnosis on the electric energy metering device, the metering correction module is used for calibrating and correcting the electric energy metering device, the automatic decision and alarm module is used for automatically making decisions and alarms, sending alarm information to related personnel and providing corresponding processing suggestions, and the control computer is used for displaying analysis and diagnosis results in a visual mode.
(III) beneficial effects
Compared with the prior art, the invention provides an electric energy metering method and system based on big data self-diagnosis, which have the following beneficial effects:
according to the electric energy metering method and system based on big data self-diagnosis, the LSTM model for multivariable time series prediction is built through the big data deep learning Tensorflow and Keras deep learning library, future values are predicted based on the predicted values of past values and past values instead of only using discrete uncorrelated features such as seasons, the model can be more stable through the predicted values of the past values, meanwhile, the whole test data set can be predicted after fitting of the model, the prediction is combined with the test data set, the scale of the test data set is adjusted, the error score of the model can be calculated through the initial predicted values and actual values, the accuracy and reliability of electric energy metering are effectively improved, automatic fault diagnosis and calibration of the electric energy metering device are achieved, the supervision capability of the electric power industry on the electric energy quality is improved, accurate electric energy metering service is provided for users, and new challenges of an electric energy metering system can be met under new situation.
Drawings
FIG. 1 is a flow chart of an electric energy metering method based on big data self-diagnosis according to the invention;
FIG. 2 is a schematic diagram of the predicted and actual values of the LSTM model according to the present invention;
FIG. 3 is a schematic view of LSTM model scatter.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-3, an electric energy metering method based on big data self-diagnosis includes the following steps:
step 1, data collection and storage: collecting and storing electric energy metering data, and ensuring the reliability and the integrity of the data in the collecting process;
step 2, data cleaning and abnormality detection: cleaning and preprocessing the acquired data, removing abnormal values and noise, and detecting possible abnormal conditions;
step 3, data analysis and model establishment: analyzing and modeling the electric energy metering data by utilizing a big data analysis technology, and establishing a model suitable for electric energy metering;
step 4, fault diagnosis and correction: performing fault diagnosis on the electric energy metering device based on the established model, detecting potential faults, and providing corresponding correction measures;
step 5, metering calibration and precision evaluation: the accuracy and the reliability of the metering device are ensured by calibrating the metering device, and the accuracy degree of metering is evaluated;
step 6, automatic decision and alarm: based on fault diagnosis and calibration results, the system automatically makes decisions and alarms, and timely alerts the metering device of abnormal conditions.
In an embodiment, the electric energy metering data in step 1 includes power grid data and personal user data, the power grid data is subjected to data acquisition through an SCADA data acquisition system, and the personal user data is subjected to data acquisition through an AMI advanced measurement system.
In the embodiment, the abnormal value of the data cleaning in the step 2 includes a repeated value of the data and an illegal value of the data, the abnormal value in the data batch is identified by the data cleaning in the step 2 through a box diagram, the data is observed, the time-sharing current voltage precision of the sub-table is found to be different from that of the total table, a large number of missing values appear in the sub-table, the missing values are replaced by the whole-point current voltage, and the missing values are filled up to complete the data cleaning through deleting the error values.
In an embodiment, the information source of the model establishment in the step 3 includes local features and global features, the model establishment in the step 3 builds an LSTM model for multivariable time sequence prediction through a big data deep learning Tensorflow and Keras deep learning library, and the specific steps of the model establishment are as follows:
s1, converting an original data set into data suitable for time sequence prediction;
s2, processing the data and adapting the data to an LSTM model for the multivariate time series prediction problem;
s3, making predictions and analyzing results.
Wherein the LSTM model predicts future values based on past values and predicted values of past values instead of using only discrete uncorrelated features such as seasons, the use of predicted values of past values can make the model more stable.
In an embodiment, the dividing method of the training set and the verification set of the time sequence includes a Walk-Forward Split method and a Side-by-Side Split method, the Side-by-Side Split method is used for dividing the data set into independent different subsets, one part is completely used for training, the other part is completely used for verification, the verification set divided by the Walk-Forward Split method is used for tuning parameters, wherein the Walk-Forward Split method is selected, the first 100 days are used as training data, the next 100 days are used as prediction data in the training process, and the next 100 days are used as verification sets, so that 1/3 data points are actually used in training, and a 200-day interval exists between the last training data point and the first prediction data point.
In an embodiment, the faults of the electric energy metering device in step 4 include faults of a current loop, faults of an electric energy meter, inaccurate clock recording result and small supply voltage of a battery, and furthermore, all features (including one-hot coded features including x and y) are regularized into data with a mean value of zero and a unit variance, each feature sequence is independently regularized to the present column, and the maximum value is regularized so that the feature distribution is in a range of a given minimum value and maximum value, and is generally between [0,1 ].
The utility model provides an electric energy metering system based on big data self-diagnosis, including data acquisition module, data preprocessing module, the feature draws the module, data analysis and modeling module, fault diagnosis module, measurement correction module, automated decision and alarm module and control computer, wherein, data acquisition module is used for gathering the primitive data that electric energy metering device produced, data preprocessing module is used for carrying out preliminary treatment to the primitive data who gathers, feature draws the module and is used for extracting the characteristic from the data after the preliminary treatment, the characteristic includes power factor, harmonic content and waveform quality index, in order to describe electric energy metering device's performance, data analysis and modeling module is used for utilizing big data to carry out analysis and modeling to the characteristic that draws, establish the correlation model between electric energy metering device performance and the data characteristic, fault diagnosis module is used for drawing out the self-diagnosis signal and the model carries out fault diagnosis to electric energy metering device through sampling return circuit band-pass filter, measurement correction module is used for calibrating and correcting electric energy metering device, automated decision and alarm module are used for automatic decision and alarm, send alarm information and provide corresponding processing suggestion, control computer is used for showing with the visual mode with analysis and diagnosis result.
In summary, the large data self-diagnosis-based electric energy metering method and system build an LSTM model for multivariable time series prediction through a large data deep learning test flow and Keras deep learning library, predict future values based on predicted values of past values and past values instead of only using discrete uncorrelated features such as seasons, the model can be more stable by using the predicted values of the past values, meanwhile, the model can predict the whole test data set after fitting by using a Walk-Forward Split method and a Side-by-Side Split method, the prediction and the test data set are combined, the scale of the test data set is adjusted, the error score of the model can be calculated through the initial predicted values and actual values, the accuracy and the reliability of electric energy metering are effectively improved, automatic fault diagnosis and calibration of an electric energy metering device are realized, the power industry can provide accurate electric energy metering service for users, new challenges faced by a novel power metering system are solved, the current detection method is biased to a chip level, the error is focused on the fact that a metering system is influenced by a fault meter, and the fault is influenced by a hardware metering system, and the problem that the fault is influenced by the fault meter is influenced by the fact that the fault meter is not influenced by a hardware or a fault meter, and the error is influenced by a fault meter.
The related modules involved in the system are all hardware system modules or functional modules in the prior art combining computer software programs or protocols with hardware, and the computer software programs or protocols involved in the functional modules are all known technologies for those skilled in the art and are not improvements of the system; the system is improved in interaction relation or connection relation among the modules, namely, the overall structure of the system is improved, so that the corresponding technical problems to be solved by the system are solved.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (8)
1. The electric energy metering method based on big data self-diagnosis is characterized by comprising the following steps of:
step 1, data collection and storage: collecting and storing electric energy metering data, and ensuring the reliability and the integrity of the data in the collecting process;
step 2, data cleaning and abnormality detection: cleaning and preprocessing the acquired data, removing abnormal values and noise, and detecting possible abnormal conditions;
step 3, data analysis and model establishment: analyzing and modeling the electric energy metering data by utilizing a big data analysis technology, and establishing a model suitable for electric energy metering;
step 4, fault diagnosis and correction: performing fault diagnosis on the electric energy metering device based on the established model, detecting potential faults, and providing corresponding correction measures;
step 5, metering calibration and precision evaluation: the accuracy and the reliability of the metering device are ensured by calibrating the metering device, and the accuracy degree of metering is evaluated;
step 6, automatic decision and alarm: based on fault diagnosis and calibration results, the system automatically makes decisions and alarms, and timely alerts the metering device of abnormal conditions.
2. The electric energy metering method based on big data self-diagnosis according to claim 1, wherein the electric energy metering data in the step 1 comprises power grid data and personal user data, the power grid data is subjected to data acquisition through a SCADA data acquisition system, and the personal user data is subjected to data acquisition through an AMI advanced measurement system.
3. The method according to claim 1, wherein the abnormal values of the data cleansing in the step 2 include repeated values of the data and illegal values of the data, and the data cleansing in the step 2 uses a bin diagram to identify the abnormal values in the data batch.
4. The electric energy metering method based on big data self-diagnosis according to claim 1, wherein the information sources of the model establishment in the step 3 comprise local features and global features, the model establishment in the step 3 builds an LSTM model for multivariate time series prediction through big data deep learning, and the specific steps of the model establishment are as follows:
s1, converting an original data set into data suitable for time sequence prediction;
s2, processing the data and adapting the data to an LSTM model for the multivariate time series prediction problem;
s3, making predictions and analyzing results.
5. The method for measuring electric energy based on big data self-diagnosis according to claim 4, wherein the dividing method of the training set and the verification set of the time sequence comprises a Walk-Forward Split method and a Side-by-Side Split method, wherein the Side-by-Side Split method is used for dividing the data set into independent different subsets, one part is completely used for training, the other part is completely used for verification, and the verification set divided by the Walk-Forward Split method is used for adjusting parameters.
6. The method according to claim 1, wherein the faults of the power metering device in the step 4 include faults of a current loop, faults of a power meter, inaccurate clock recording results and small supply voltage of a battery.
7. An electric energy metering system based on big data self-diagnosis is suitable for the electric energy metering method based on big data self-diagnosis according to any one of claims 1 to 6, and is characterized by comprising a data acquisition module, a data preprocessing module, a feature extraction module, a data analysis and modeling module, a fault diagnosis module, a metering correction module, an automatic decision and alarm module and a control computer.
8. The self-diagnosis electric energy metering system of claim 7, wherein the data acquisition module is used for acquiring original data generated by the electric energy metering device, the data preprocessing module is used for preprocessing the acquired original data, the feature extraction module is used for extracting features from the preprocessed data, the features comprise power factors, harmonic content and waveform quality indexes so as to describe the performance of the electric energy metering device, the data analysis and modeling module is used for analyzing and modeling the extracted features by using the big data, a correlation model between the performance of the electric energy metering device and the data features is established, the fault diagnosis module is used for extracting self-diagnosis signals and the model through a sampling loop band-pass filter so as to perform fault diagnosis on the electric energy metering device, the measurement correction module is used for calibrating and correcting the electric energy metering device, the automatic decision and alarm module is used for automatically making decisions and alarming, sending alarm information to related personnel and providing corresponding processing suggestions, and the control computer is used for displaying the analysis and diagnosis results in a visual mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311448205.1A CN117633612A (en) | 2023-11-02 | 2023-11-02 | Electric energy metering method and system based on big data self-diagnosis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311448205.1A CN117633612A (en) | 2023-11-02 | 2023-11-02 | Electric energy metering method and system based on big data self-diagnosis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117633612A true CN117633612A (en) | 2024-03-01 |
Family
ID=90029470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311448205.1A Pending CN117633612A (en) | 2023-11-02 | 2023-11-02 | Electric energy metering method and system based on big data self-diagnosis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117633612A (en) |
-
2023
- 2023-11-02 CN CN202311448205.1A patent/CN117633612A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108375715B (en) | Power distribution network line fault risk day prediction method and system | |
| CN107346899B (en) | Photovoltaic power station system stability assessment method and system | |
| EP2992340B1 (en) | System and method for ups battery monitoring and data analysis | |
| CN109886475B (en) | Information security situation perception system of measurement automation system based on AI | |
| CN106154209A (en) | Electrical energy meter fault Forecasting Methodology based on decision Tree algorithms | |
| CN109284933B (en) | Electronic transformer state evaluation system and method based on mathematical statistics | |
| CN108802535B (en) | Screening method, main interference source identification method and device, server and storage medium | |
| CN108847682B (en) | Photovoltaic power station grid-connected performance evaluation method and system based on operation data | |
| CN115343623B (en) | Online detection method and device for faults of electrochemical energy storage battery | |
| CN114781476A (en) | Fault analysis system and method for measuring equipment | |
| CN115455358A (en) | Electrical parameter trend early warning and fault diagnosis method based on nonlinear regression model | |
| CN114265001B (en) | Smart electric meter metering error evaluation method | |
| CN113739837A (en) | State prediction processing system of standard instrument | |
| CN116449235A (en) | Method and system for processing test data of energy storage battery | |
| CN105954695A (en) | Synchronization-based homogeneous-sensor mutation parameter recognizing method and device | |
| CN117318310B (en) | Remote unmanned intelligent monitoring system for explosion-proof power distribution cabinet | |
| CN109299201A (en) | Power plant's production subsystem method for monitoring abnormality and device based on two-phase analyzing method | |
| CN112926656A (en) | Method, system and equipment for predicting state of circulating water pump of nuclear power plant | |
| CN113092927A (en) | Alternating current filter state diagnosis method based on wave recording waveform image and artificial intelligence | |
| CN112345972A (en) | Power failure event-based power distribution network line transformation relation abnormity diagnosis method, device and system | |
| CN111999695A (en) | State evaluation and abnormity diagnosis method for metering device of transformer substation | |
| CN117633612A (en) | Electric energy metering method and system based on big data self-diagnosis | |
| CN106682383A (en) | Accurate statistical processing method for collected meter code values in metering system | |
| CN115343579A (en) | Power grid fault analysis method and device and electronic equipment | |
| CN114139408A (en) | Power transformer health state assessment method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |