CN110503211A - Failure prediction method based on machine learning - Google Patents
Failure prediction method based on machine learning Download PDFInfo
- Publication number
- CN110503211A CN110503211A CN201910779470.5A CN201910779470A CN110503211A CN 110503211 A CN110503211 A CN 110503211A CN 201910779470 A CN201910779470 A CN 201910779470A CN 110503211 A CN110503211 A CN 110503211A
- Authority
- CN
- China
- Prior art keywords
- data
- value
- prediction
- module
- monitoring
- 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 29
- 238000010801 machine learning Methods 0.000 title claims abstract description 13
- 238000012544 monitoring process Methods 0.000 claims abstract description 40
- 238000004458 analytical method Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims abstract description 7
- 238000012423 maintenance Methods 0.000 claims description 10
- 238000013499 data model Methods 0.000 claims description 5
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 238000007726 management method Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract description 2
- 238000012806 monitoring device Methods 0.000 abstract 1
- 230000000737 periodic effect Effects 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q10/20—Administration of product repair or maintenance
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Public Health (AREA)
- Probability & Statistics with Applications (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Biology (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Alarm Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
Failure prediction method based on machine learning is related to power grid power supply and sets field, especially a kind of by data monitoring, adjustment, comparison, realizes that monitoring device voluntarily learns to adjust, the failure prediction method based on machine learning predicted line defct.The present invention includes data monitoring, comparing, failure prediction and prediction four modules of adjustment, the operation data of data monitoring module monitors supply network, it is analysed and compared by data comparing module, in comparison process, prediction data is corrected by prediction adjustment module, prediction data is analyzed, the exception that will be generated is sounded an alarm by failure prediction module.The present invention is by training supply network operation data to it, evolution algorithm is introduced simultaneously, realize analysis result being mutually authenticated with real-time running data, the error between analysis result and real time data for analyzing supply network data reaches tolerance interval, to use analysis modified result abnormal data or amended record missing data, the influence of data exception is solved.
Description
Technical field
The present invention relates to power grid power supplies to set field, especially a kind of by data monitoring, adjustment, comparison, realizes that monitoring is set
Standby voluntarily study adjustment, the failure prediction method based on machine learning that line defct is predicted.
Background technique
Currently, in a power supply network, including power supply unit and power supply line, it is all made of traditional maintenance mode, is such as carried out
Periodic maintenance, detection, it is when a failure occurs, then interim to carry out to the components of equipment or route by the way of periodic replacement
Maintenance replacement, but often there is certain equipment or route, because components are overhauled and replaced to failure, within a short period of time, again
Because periodic maintenance is overhauled, components are replaced again, so cause the waste of lot of materials, also result in the maintenance of power supply system
Cost is high, is not only unfavorable for energy conservation and environmental protection, and also cause the waste of manpower, when an unexpected situation occurs, causes not
It can repair in time, generate bigger loss.
Meanwhile in the power system, due to by conditions such as unit creep speed, power grid static security, whole society's total loads
Constraint, operation of power networks parameter can change in a certain range.When cataclysm occurs for operation of power networks parameter, illustrate that operation of power networks is deposited
It is often exactly overvoltage when the yielding of its voltage is raised and lowered if main transformer voltage is relative constant in substation in some problems
Or the generation of substation power loss event.
Summary of the invention
To be solved by this invention is exactly in existing supply network, using conventional maintenance mode, so that maintenance cost is high, no
Conducive to energy conservation and environmental protection, the problem of manpower wastes also will cause, provide one kind by data monitoring, adjustment, comparison, realize that monitoring is set
Standby voluntarily study adjustment, the failure prediction method based on machine learning that line defct is predicted.
Failure prediction method based on machine learning of the invention, it is characterised in that the prediction technique include data monitoring,
Comparing, failure prediction and prediction four modules of adjustment, the operation data of data monitoring module monitors supply network pass through
Data comparing module is analysed and compared, and in comparison process, is corrected prediction data by prediction adjustment module, is analyzed prediction data,
The exception that will be generated is sounded an alarm by failure prediction module, in which:
1) data monitoring module includes voltage monitoring and current monitoring, the real time data of monitoring power-supply networks network in the process of running,
Including current value and voltage value;
2) data comparing module records the real time data of the voltage and current of data monitoring module monitors, analyzed and is compared
Right, when recording data, using the period as parameter, when normal operating condition, the numerical value in daily each period of record takes it
Average value, as the reference point in the period
In above formula,For the sampled data in the daily same period, average value is calculated using week or the moon as the period;
Meanwhile the fluctuation numerical value in the period a cycle is recorded, which is gone out in a cycle by mean value calculation
Interior deviation and median, and maximum value and minimum value in the period are recorded, while calculating the frequency distribution of deviation;
3) by technical staff's setting electric current and the standard value of voltage, threshold value, the difference of standard value and threshold value in prediction adjustment module
Three parameters of range are modified standard value by being compared with the average value that data comparing module calculates;And by inclined
The difference of the range of difference and frequency correction threshold, revised standard value and threshold value is no more than the standard value and threshold value set
Difference;
4) failure prediction module is according to the real time data of data monitoring module, compares the modified standard value of prediction adjustment module, partially
Difference and threshold value, pre-process real time data, when real time data exceeds deviation, issue primary alarm, prompt prison
Survey personnel issue second-level alarm, monitoring personnel are notified to handle when real time data exceeds threshold value.
In the step 4), after failure prediction module issues primary alarm, monitoring personnel determines according to the actual situation should
Exceeded numerical value belongs to normal fluctuation or improper fluctuation;Judgement belong to normal fluctuation, then data comparing module by the numerical value the most
Maximum value or minimum value are recorded, and update average value, deviation level-one median;When being judged as improper fluctuation, then by
Monitoring personnel carries out emergent management, such as reports, maintenance is even shut down.
The failure prediction method based on machine learning, including training stage and detection-phase, training stage utilize
Data object constructs data model;Data exception situation is judged according to data model in detection-phase.
Failure prediction method based on machine learning of the invention, by being trained to supply network operation data to it, together
When introduce evolution algorithm, realize analysis result being mutually authenticated with real-time running data, and continue through real time data and promote prison
Measurement equipment is learnt, and the error between analysis result and real time data for analyzing supply network data reaches acceptable model
It encloses, to solve the influence that data exception runs supply network with analysis modified result abnormal data or amended record missing data.
Detailed description of the invention
Fig. 1 is module connection diagram of the present invention.
Fig. 2 is work step schematic diagram of the present invention.
Specific embodiment
Embodiment 1: a kind of failure prediction method based on machine learning, the prediction technique include data monitoring, data ratio
Data ratio is passed through to, failure prediction and prediction four modules of adjustment, the operation data of data monitoring module monitors supply network
Module analysis is compared, in comparison process, prediction data is corrected by prediction adjustment module, prediction data is analyzed, by lacking
Prediction module is fallen into sound an alarm the exception that will be generated, in which:
1) data monitoring module includes voltage monitoring and current monitoring, the real time data of monitoring power-supply networks network in the process of running,
Including current value and voltage value;
2) data comparing module records the real time data of the voltage and current of data monitoring module monitors, analyzed and is compared
Right, when recording data, using the period as parameter, when normal operating condition, the numerical value in daily each period of record takes it
Average value, as the reference point in the period
In above formula,For the sampled data in the daily same period, average value is calculated using week or the moon as the period;
Meanwhile the fluctuation numerical value in the period a cycle is recorded, which is gone out in a cycle by mean value calculation
Interior deviation and median, and maximum value and minimum value in the period are recorded, while calculating the frequency distribution of deviation;
3) by technical staff's setting electric current and the standard value of voltage, threshold value, the difference of standard value and threshold value in prediction adjustment module
Three parameters of range are modified standard value by being compared with the average value that data comparing module calculates;And by inclined
The difference of the range of difference and frequency correction threshold, revised standard value and threshold value is no more than the standard value and threshold value set
Difference;
4) failure prediction module is according to the real time data of data monitoring module, compares the modified standard value of prediction adjustment module, partially
Difference and threshold value, pre-process real time data, when real time data exceeds deviation, issue primary alarm, prompt prison
Survey personnel issue second-level alarm, monitoring personnel are notified to handle when real time data exceeds threshold value.
After failure prediction module issues primary alarm, monitoring personnel determines that the exceeded numerical value belongs to normally according to the actual situation
Fluctuation or improper fluctuation;Judgement belong to normal fluctuation, then data comparing module by the numerical value maximum value or minimum value the most into
Row record, and update average value, deviation level-one median;When being judged as improper fluctuation, then carried out by monitoring personnel urgent
Processing is such as reported, maintenance is even shut down.
Above-mentioned failure prediction method, including training stage and detection-phase, training stage utilize data object, construct number
According to model;Data exception situation is judged according to data model in detection-phase.
Claims (3)
1. a kind of failure prediction method based on machine learning, it is characterised in that the prediction technique includes data monitoring, data ratio
Data ratio is passed through to, failure prediction and prediction four modules of adjustment, the operation data of data monitoring module monitors supply network
Module analysis is compared, in comparison process, prediction data is corrected by prediction adjustment module, prediction data is analyzed, by lacking
Prediction module is fallen into sound an alarm the exception that will be generated, in which:
1) data monitoring module includes voltage monitoring and current monitoring, the real time data of monitoring power-supply networks network in the process of running,
Including current value and voltage value;
2) data comparing module records the real time data of the voltage and current of data monitoring module monitors, analyzed and is compared
Right, when recording data, using the period as parameter, when normal operating condition, the numerical value in daily each period of record takes it
Average value, as the reference point in the period
In above formula,For the sampled data in the daily same period, average value is calculated using week or the moon as the period;
Meanwhile the fluctuation numerical value in the period a cycle is recorded, which is gone out in a cycle by mean value calculation
Interior deviation and median, and maximum value and minimum value in the period are recorded, while calculating the frequency distribution of deviation;
3) by technical staff's setting electric current and the standard value of voltage, threshold value, the difference of standard value and threshold value in prediction adjustment module
Three parameters of range are modified standard value by being compared with the average value that data comparing module calculates;And by inclined
The difference of the range of difference and frequency correction threshold, revised standard value and threshold value is no more than the standard value and threshold value set
Difference;
4) failure prediction module is according to the real time data of data monitoring module, compares the modified standard value of prediction adjustment module, partially
Difference and threshold value, pre-process real time data, when real time data exceeds deviation, issue primary alarm, prompt prison
Survey personnel issue second-level alarm, monitoring personnel are notified to handle when real time data exceeds threshold value.
2. the failure prediction method based on machine learning as described in claim 1, it is characterised in that in the step 4), lack
Fall into prediction module issue primary alarm after, monitoring personnel determine according to the actual situation the exceeded numerical value belong to normal fluctuation or it is non-just
Ordinary wave is dynamic;Judgement belongs to normal fluctuation, then data comparing module records numerical value maximum value or minimum value the most, and more
New average value, deviation level-one median;When being judged as improper fluctuation, then emergent management is carried out by monitoring personnel, such as report,
Maintenance is even shut down.
3. the failure prediction method based on machine learning as described in claim 1, it is characterised in that described based on engineering
The failure prediction method of habit, including training stage and detection-phase, training stage utilize data object, construct data model;In
Detection-phase judges data exception situation according to data model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910779470.5A CN110503211A (en) | 2019-08-22 | 2019-08-22 | Failure prediction method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910779470.5A CN110503211A (en) | 2019-08-22 | 2019-08-22 | Failure prediction method based on machine learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110503211A true CN110503211A (en) | 2019-11-26 |
Family
ID=68588931
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910779470.5A Pending CN110503211A (en) | 2019-08-22 | 2019-08-22 | Failure prediction method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110503211A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111860936A (en) * | 2020-05-25 | 2020-10-30 | 北京致远互联软件股份有限公司 | Method for predicting defects of office business process |
CN113465777A (en) * | 2021-06-03 | 2021-10-01 | 广州番禺电缆集团有限公司 | Cable temperature monitoring platform and method |
CN113945823A (en) * | 2021-09-26 | 2022-01-18 | 成都嘉纳海威科技有限责任公司 | Method for detecting potential defects of chip |
CN117349778A (en) * | 2023-12-04 | 2024-01-05 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
-
2019
- 2019-08-22 CN CN201910779470.5A patent/CN110503211A/en active Pending
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111860936A (en) * | 2020-05-25 | 2020-10-30 | 北京致远互联软件股份有限公司 | Method for predicting defects of office business process |
CN113465777A (en) * | 2021-06-03 | 2021-10-01 | 广州番禺电缆集团有限公司 | Cable temperature monitoring platform and method |
CN113945823A (en) * | 2021-09-26 | 2022-01-18 | 成都嘉纳海威科技有限责任公司 | Method for detecting potential defects of chip |
CN113945823B (en) * | 2021-09-26 | 2024-04-09 | 成都嘉纳海威科技有限责任公司 | Method for detecting potential defects of chip |
CN117349778A (en) * | 2023-12-04 | 2024-01-05 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
CN117349778B (en) * | 2023-12-04 | 2024-02-20 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110503211A (en) | Failure prediction method based on machine learning | |
CN106655522B (en) | A kind of main station system suitable for electric grid secondary equipment operation management | |
CN112039075B (en) | Converter station multidimensional data analysis and monitoring system | |
US10691085B2 (en) | Defect detection in power distribution system | |
CN104821789B (en) | A kind of detection method of photovoltaic generating system reliability | |
CN111476385B (en) | Building facility maintenance supervisory systems based on BIM | |
CN115622054B (en) | Operation monitoring method and system of energy system | |
CN117689214B (en) | Dynamic safety assessment method for energy router of flexible direct-current traction power supply system | |
CN116388402B (en) | Abnormality alarm analysis method applied to power transformation equipment | |
CN117289659A (en) | Intelligent automatic monitoring system for centralized control operation of power plant | |
CN114928173A (en) | Intelligent power distribution system based on power grid business middling station and electric power data safety interaction | |
CN117110936A (en) | UPS running state prediction maintenance system based on time sequence analysis | |
CN117394535A (en) | Digital twin system of AC/DC hybrid power distribution network | |
CN116308306B (en) | New energy station intelligent management system and method based on 5G | |
CN117279348A (en) | Strong electromagnetic pulse protection system with monitoring function | |
CN116231757A (en) | Power generation energy efficiency analysis management system of water-wind-solar power station group | |
CN115934714A (en) | Information storage and classification system for micro-grid | |
CN114742146A (en) | Medium-voltage uninterruptible power device monitoring system | |
KR102291365B1 (en) | Apparatus and method for collecting history data of solar power system | |
CN107958505A (en) | A kind of stable intelligent inspection system for operating condition and its control method | |
CN117808456B (en) | Equipment fault early warning method and device based on intelligent operation management | |
CN116243072B (en) | Electric equipment systematic maintenance management system and method suitable for construction site | |
CN117638928B (en) | Intelligent power distribution network management system based on cloud computing | |
CN116957328A (en) | Active medical instrument fault adverse event risk early warning method and system | |
Pei et al. | Trend Prediction of DC Measuring System Based on LSTM |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191126 |
|
WD01 | Invention patent application deemed withdrawn after publication |