CN104636999B - A kind of abnormal energy data detection method of building - Google Patents
A kind of abnormal energy data detection method of building Download PDFInfo
- Publication number
- CN104636999B CN104636999B CN201510001544.4A CN201510001544A CN104636999B CN 104636999 B CN104636999 B CN 104636999B CN 201510001544 A CN201510001544 A CN 201510001544A CN 104636999 B CN104636999 B CN 104636999B
- Authority
- CN
- China
- Prior art keywords
- data
- building
- subsequence
- abnormal
- energy
- 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.)
- Active
Links
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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/08—Construction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention discloses a kind of abnormal energy data detection methods of building, and initial data is sorted out first, in accordance with time series;It refers again to industry index to be judged, finds out the abnormal data of special time period;Then it using the separate-blas estimation technology in data mining, filters out with the exceptional value in energy data;The abnormal point in data is finally found out using the Time Series Similarity based on Discrete Fourier Transform.The method of abnormal data precise positioning in the data provided by the invention to energy for building, it may replace the empirical value determining method generally used in current energy industry, by the application of the analysis method and Discrete Fourier Transform principle of artificial intelligence in data mining, the erroneous judgement to exception energy data is prevented, reports by mistake and fails to report, effective exception energy monitoring information is provided for energy unit and personnel.
Description
Technical field
The present invention relates to building energy consumption monitoring field, the abnormal energy data detection method of specifically a kind of building.
Background technology
In building energy consumption monitoring system, exception can detect be the safe efficient operation of guarantee system an important ring, by
The function, operation maintenance personnel can know in real time with can extremely, be accurately positioned abort situation, in time exclude or repair system failure.When
Before, judge exception can data substantially dependent on expert or user experience, with reference to industry index, set fixed limit or
Threshold value, once monitoring data, which exceeds limit or reaches threshold value, triggers alarm.And in actual motion, with energy situation climate, people
The conditions such as member, season influence, complex with energy data fluctuations situation, abnormal obvious by single static data is set to judge
It loses precisely.It is found by being tracked to the operation of more systems, abnormal system acquisition when occurring is less than and judging wrong report phenomenon by accident
It happens occasionally, not only can do nothing to help user accurately with energy, can also generate interference to user job in some cases.
Invention content
Separate-blas estimation technology in data mining is combined by the present invention with practical experience judgement, the accurate abnormal number of screening
According to realization is to the abnormal effective judgement with energy data of building in maximum magnitude, to overcome defect in the prior art.
The purpose of the present invention is achieved through the following technical solutions:
A kind of abnormal energy data detection method of building:
Step (1):Building energy consumption data are acquired by frequency acquisition of fixed time period, establish the time series with energy data
D:
D={ dt| t=t0,t1,t2,...,tn-1};
In formula, dtFor element, the gathered data in t moment is represented;
Step (2):It according to building actual conditions simultaneously based on industry target setting threshold value, compares with the threshold value, during extraction
Between abnormal elements in sequence, i.e., abnormal gathered data, using abnormal elements as the subsequence D of D1;It is removed in time series D
Subsequence D of the abnormal elements as D2:D2={ d1,d2,...,dp};
Step (3):Using the Outlier mining method based on deviation, with regular length m by D2It is split as s sub- sequences
Row, D2Subsequence D2sIt is expressed as:
D2s={ ds,ds+1,...,ds+m-1, 2≤m≤p
In formula, m is the element number of subsequence, and p represents D2Element number;
The smoothing factor of each subsequence is calculated with dissimilarity function, smoothing factor with defining threshold value is compared, is judged whether
Exceed:The subsequence exceeded is denoted as D21;Without departing from subsequence be denoted as normal data sequence;
Step (4):Time Series Similarity based on discrete Fourier transform searches D respectively1And D21In abnormal elements,
By D1And D21Multiple subsequences are split as with regular length respectively, with discrete Fourier transform by subsequence time series data from time domain
Spatial alternation obtains Fourier coefficient sequence, Fourier coefficient D to domain spacef:
In formula, element numbers of the n for subsequence, dtFor the element in subsequence, f=0 ..., n-1;
Given standard energy consumption sequence, given threshold, by D1And D21In the subsequence that splits out respectively with standard energy consumption sequence
Between calculate Euclidean distance, more than given threshold value sequence for abnormal data sequence, what it is less than or equal to threshold value is then normal data sequence
Row.
Building actual conditions and industry index include:Spring, period in autumn, air conditioning electricity should be less than standby power consumption amount;Night
Between the rest period, the electricity consumption of public building room lighting and should maintain extremely low energy consumption with water;Office building festivals or holidays electricity consumption should be small
In work daily power consumption;Commercial building non operation time section energy consumption should be less than business hours section energy consumption;Office building, teaching building etc. are built
Building the power consumption values fluctuation range during normal use should be relatively steady.
Preferably, dissimilarity function is defined as in step (3):
In step (4), the Euclidean distance of two sequences is expressed as:X in formulafAnd yfRespectively sequence X
With coefficients of the sequence Y after discrete Fourier transform.
Beneficial effects of the present invention:
The present invention provides a kind of methods of the abnormal data precise positioning in data to energy for building, may replace current energy
The empirical value determining method generally used in the industry of source, analysis method and discrete Fu by artificial intelligence in data mining
The application of vertical leaf transformation principle prevents the erroneous judgement to exception energy data, reports by mistake and fail to report, is provided with for energy unit and personnel
Abnormal use of effect can monitoring information.
Description of the drawings
Fig. 1 is the step flow chart of the method for the present invention.
Specific embodiment
Below in conjunction with specific embodiment, the invention will be further described.
Embodiment 1
A kind of abnormal energy data detection method of building:
Step (1):Building energy consumption data are acquired by frequency acquisition of fixed time period, establish the time series with energy data
D:
D={ dt| t=t0,t1,t2,...,tn-1};
In formula, dtFor element, the gathered data in t moment is represented;
Step (2):It according to building actual conditions simultaneously based on industry target setting threshold value, compares with the threshold value, during extraction
Between abnormal elements in sequence, i.e., abnormal gathered data, using abnormal elements as the subsequence D of D1;It is removed in time series D
Subsequence D of the abnormal elements as D2:D2={ d1,d2,...,dp};
Step (3):Using the Outlier mining method based on deviation, with regular length m by D2It is split as s sub- sequences
Row, D2Subsequence D2sIt is expressed as:
D2s={ ds,ds+1,...,ds+m-1, 2≤m≤p
In formula, m is the element number of subsequence, and p represents D2Element number;
The smoothing factor of each subsequence is calculated with dissimilarity function, smoothing factor with defining threshold value is compared, is judged whether
Exceed:The subsequence exceeded is denoted as D21;Without departing from subsequence be denoted as normal data sequence;
Step (4):Time Series Similarity based on discrete Fourier transform searches D respectively1And D21In abnormal elements,
By D1And D21Multiple subsequences are split as with regular length respectively, with discrete Fourier transform by subsequence time series data from time domain
Spatial alternation obtains Fourier coefficient sequence, Fourier coefficient D to domain spacef:
In formula, element numbers of the n for subsequence, dtFor the element in subsequence, f=0 ..., n-1;
Given standard energy consumption sequence, given threshold, by D1And D21In the subsequence that splits out respectively with standard energy consumption sequence
Between calculate Euclidean distance, more than given threshold value sequence for abnormal data sequence, what it is less than or equal to threshold value is then normal data sequence
Row.
Embodiment 2:
Such as the abnormal energy data detection method of building a kind of in embodiment 1, spring, period in autumn, air conditioning electricity should be less than
Standby power consumption amount;Night rest periods, the electricity consumption of public building room lighting and should maintain extremely low energy consumption with water;Office building section is false
Daily power consumption should be less than work daily power consumption;Commercial building non operation time section energy consumption should be less than business hours section energy consumption;Office
Power consumption values fluctuation range should be relatively steady during normal use for the building such as building, teaching building.
Embodiment 3:
Such as the abnormal energy data detection method of building a kind of in embodiment 1 or 2, dissimilarity function is defined as:
In the embodiment above, the Time Series Similarity that step (4) is converted based on discrete Fourier finding step again
(2), the subsequence D filtered out in (3)1And D21In abnormal elements.
The energy consumption data of building, time series there are certain law variation, such as seasonal variations generate air-conditioning
Electricity consumption data variation, Campus buildingss the energy consumption regularity between winter and summer vacations reduce, commercial building energy consumption rule during festivals or holidays
Property raising etc..It can be based on the variation of above-mentioned energy consumption data rule and make each class standard energy consumption sequence.
Principle is changed using discrete Fourier, when the method based on time series similarity analysis passes through multiple data points
Countershaft connects into curve, and line is expanded to by point, and the similarity or diversity factor between line and line are analyzed, and thus can will isolate
Energy consumption data be connected into it is regular with can behavior curve, therefrom find out outlier, position abnormal data.
By D1And D21Multiple subsequences are split as with regular length respectively, with Discrete Fourier Transform by ordinal number during subsequence
Domain space is transformed to according to from time domain space, obtains following fourier coefficient:
Wherein, element numbers of the n for subsequence, dtFor the element in subsequence, f=0 ..., n-1;
It is whether similar using two sequences of Euclidean distance measurement, if the Euclidean distance of two sequences is less than given threshold value ε,
Then think that the two sequences are similar, that is, meet equation below:
According to Parserval theories, following equation is also set up:
By D1And D21In the subsequence that splits out and calculate Euclidean distance between standard energy consumption sequence, more than the sequence of given threshold value
Abnormal data sequence is classified as, what it is less than or equal to threshold value is then normal data sequence.
Heretofore described threshold value can be provided by limited trials or experience.
Above example is merely to illustrate technical scheme of the present invention rather than limiting the scope of the invention, although
The present invention is explained in detail with reference to preferred embodiment, it will be understood by those of ordinary skill in the art that, it can be to this hair
Bright technical solution is modified or equivalent replacement, without departing from the spirit and scope of technical solution of the present invention.
Claims (3)
1. a kind of abnormal energy data detection method of building, it is characterised in that:Method is
Step (1):Building energy consumption data are acquired by frequency acquisition of fixed time period, establish the time series D with energy data:
D={ dt| t=t0,t1,t2,...,tn-1};
In formula, dtFor element, the gathered data in t moment is represented;
Step (2):According to building actual conditions and industry target setting threshold value is based on, is compared with the threshold value, extraction time sequence
Abnormal elements in row, using abnormal elements as the subsequence D of D1;The abnormal elements are removed in time series D as D's
Subsequence D2:D2={ d1,d2,...,dp};
Step (3):Using the Outlier mining method based on deviation, with regular length m by D2It is split as s subsequence, D2's
Subsequence D2sIt is expressed as:
D2s={ ds,ds+1,...,ds+m-1, 2≤m≤p
In formula, m is the element number of subsequence, and p represents D2Element number;
The smoothing factor of each subsequence is calculated with dissimilarity function, smoothing factor with defining threshold value is compared, is seen if fall out:
The subsequence exceeded is denoted as D21;Without departing from subsequence be denoted as normal data sequence;Dissimilarity function is defined as:
In formula, A is adjacent two numerical value dq-1, dqBetween difference, B is follow-up adjacent two numbers dq, dq+1Between difference, m is collection object
Length,It is xor operator, is defined as
Step (4):Time Series Similarity based on discrete Fourier transform searches D respectively1And D21In abnormal elements, by D1
And D21Multiple subsequences are split as with regular length respectively, with discrete Fourier transform by subsequence time series data from time domain space
Domain space is transformed to, obtains Fourier coefficient sequence, Fourier coefficient Df:
In formula, element numbers of the n for subsequence, dtFor the element in subsequence, f=0 ..., n-1;
Given standard energy consumption sequence, given threshold, by D1And D21In the subsequence that splits out calculate respectively and standard energy consumption sequence
Between Euclidean distance, Euclidean distance is more than the sequence of given threshold value for abnormal data sequence, less than or equal to threshold value then for just
Regular data sequence.
2. a kind of abnormal energy data detection method of building according to claim 1, it is characterised in that:Build actual conditions
And industry index includes:Spring, period in autumn, air conditioning electricity should be less than standby power consumption amount;Night rest periods, office building, religion
It learns building room lighting electricity consumption and should maintain extremely low energy consumption with water;Office building festivals or holidays electricity consumption should be less than work daily power consumption;
Commercial building non operation time section energy consumption should be less than business hours section energy consumption;Office building, the teaching building energy during normal use
Consumption value fluctuation range should be relatively steady.
3. a kind of abnormal energy data detection method of building according to claim 1, it is characterised in that:The Europe of two sequences
Formula distance is expressed as:X in formulafAnd yfRespectively sequence X and sequence Y be after discrete Fourier transform
Number.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510001544.4A CN104636999B (en) | 2015-01-04 | 2015-01-04 | A kind of abnormal energy data detection method of building |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510001544.4A CN104636999B (en) | 2015-01-04 | 2015-01-04 | A kind of abnormal energy data detection method of building |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104636999A CN104636999A (en) | 2015-05-20 |
CN104636999B true CN104636999B (en) | 2018-06-22 |
Family
ID=53215709
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510001544.4A Active CN104636999B (en) | 2015-01-04 | 2015-01-04 | A kind of abnormal energy data detection method of building |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104636999B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021120829A1 (en) * | 2019-12-20 | 2021-06-24 | 上海市建筑科学研究院有限公司 | Method for diagnosing energy consumption of lighting socket branches of buildings during non-operational periods |
Families Citing this family (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104915562B (en) * | 2015-06-11 | 2019-03-22 | 上海腾天节能技术有限公司 | Build efficiency diagnostic method and system |
CN106155867A (en) * | 2016-08-23 | 2016-11-23 | 珠海金智维信息科技有限公司 | The alarm method of monitoring performance data similarity tolerance and system |
CN106371939B (en) * | 2016-09-12 | 2019-03-22 | 山东大学 | A kind of time series data method for detecting abnormality and its system |
CN106384300B (en) * | 2016-09-27 | 2019-08-16 | 山东建筑大学 | Based on big data and the building of fuzzy model extremely with energy detection method and system |
CN109508846A (en) * | 2017-09-15 | 2019-03-22 | 新奥(中国)燃气投资有限公司 | A kind of detection method and device of unit data exception fluctuation |
CN108319568B (en) * | 2018-03-06 | 2021-08-06 | 江苏中科瀚星数据科技有限公司 | Method for positioning abnormal equipment state problem |
CN108536741B (en) * | 2018-03-09 | 2021-10-15 | 国网江苏省电力公司无锡供电公司 | Energy consumption abnormity monitoring method |
CN109359134B (en) * | 2018-08-30 | 2021-07-06 | 大连理工大学 | Method for identifying energy consumption implicit abnormal data of lighting socket based on data mining |
CN109409661A (en) * | 2018-09-19 | 2019-03-01 | 平安科技(深圳)有限公司 | Electronic device, the building Electrical Safety recognition methods based on cusum algorithm and storage medium |
CN109685362B (en) * | 2018-12-21 | 2021-04-23 | 吉林建筑大学 | Building legacy protection evaluation system and method based on intelligent network |
CN110362612B (en) * | 2019-07-19 | 2022-02-22 | 中国工商银行股份有限公司 | Abnormal data detection method and device executed by electronic equipment and electronic equipment |
CN113742387A (en) * | 2020-05-29 | 2021-12-03 | 中兴通讯股份有限公司 | Data processing method, device and computer readable storage medium |
CN112237433B (en) * | 2020-11-05 | 2021-07-23 | 山东大学 | Electroencephalogram signal abnormity monitoring system and method |
CN113917263B (en) * | 2021-10-09 | 2022-05-24 | 乐创达投资(广东)有限公司 | Method and system for secondary monitoring of abnormal energy consumption based on prediction data |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102289585A (en) * | 2011-08-15 | 2011-12-21 | 重庆大学 | Real-time monitoring method for energy consumption of public building based on data mining |
CN104199961A (en) * | 2014-09-19 | 2014-12-10 | 北京建筑技术发展有限责任公司 | Data mining based public building energy consumption monitoring platform data processing method |
-
2015
- 2015-01-04 CN CN201510001544.4A patent/CN104636999B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102289585A (en) * | 2011-08-15 | 2011-12-21 | 重庆大学 | Real-time monitoring method for energy consumption of public building based on data mining |
CN104199961A (en) * | 2014-09-19 | 2014-12-10 | 北京建筑技术发展有限责任公司 | Data mining based public building energy consumption monitoring platform data processing method |
Non-Patent Citations (2)
Title |
---|
基于相似性的时间序列异常检测研究;陈然;《中国优秀硕士学位论文全文数据库·信息科技辑》;20120615(第06期);第4-27页 * |
建筑能效当量能耗方法研究;李峥嵘;《同济大学学报(自然科学版)》;20100331;第38卷(第3期);第353-357页 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021120829A1 (en) * | 2019-12-20 | 2021-06-24 | 上海市建筑科学研究院有限公司 | Method for diagnosing energy consumption of lighting socket branches of buildings during non-operational periods |
Also Published As
Publication number | Publication date |
---|---|
CN104636999A (en) | 2015-05-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104636999B (en) | A kind of abnormal energy data detection method of building | |
CN103903408B (en) | Method for early warning and system are investigated in equipment fault | |
US10498585B2 (en) | Sensor data analytics and alarm management | |
US11002773B2 (en) | Monitoring apparatus, monitoring method, and storage medium | |
JP2014096050A (en) | Process monitoring diagnostic device and process monitoring diagnostic program | |
CN110322371A (en) | The area Gao Suntai multiplexing electric abnormality user based on multiple linear regression analysis detects localization method | |
CN103813355A (en) | Identification method for anomalous points of cooperative synchronization in distributed network | |
EP2837270A2 (en) | Fault detection, localization and performance monitoring of photosensors for lighting controls | |
US11222520B2 (en) | Precise predictive maintenance method of driver | |
KR101708709B1 (en) | Method for estimating customer baseline load using data mining and apparatus thereof | |
Yu et al. | Assessment of operating performance of chiller systems using cluster analysis | |
Zucker et al. | Improving energy efficiency of buildings using data mining technologies | |
CN106228280A (en) | Grid operating monitoring information identification categorizing system | |
Jeyasheeli et al. | An IOT design for smart lighting in green buildings based on environmental factors | |
CN110298765B (en) | Power distribution network power consumption abnormality detection method based on objective correlation factors | |
KR20190078685A (en) | Method of Anomaly Pattern Detection for Sensor Data using Increamental Clustering | |
Kiluk | Algorithmic acquisition of diagnostic patterns in district heating billing system | |
CN109842372A (en) | A kind of photovoltaic module fault detection method and system | |
CN110426996A (en) | Environmental pollution monitoring method based on big data and artificial intelligence | |
WO2018016098A1 (en) | Facility monitoring device and facility monitoring method | |
KR101997580B1 (en) | Data classification method based on correlation, and a computer-readable storege medium having program to perform the same | |
Chen et al. | Whole building system fault detection based on weather pattern matching and PCA method | |
US11507861B2 (en) | Machine learning techniques for improved water conservation and management | |
WO2016098805A1 (en) | Data-related information processing device and program | |
Bariya et al. | k-shapestream: Probabilistic streaming clustering for electric grid events |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 210046 6 building, 01 building 108, Gan Jia Bian, Qixia District, Nanjing, Jiangsu. Applicant after: Jiangsu alliance wisdom energy Limited by Share Ltd Address before: 210046 6 building, 01 building 108, Gan Jia Bian, Qixia District, Nanjing, Jiangsu. Applicant before: Nanjing Lianhong Automatization System Engineering Co., Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |