CN110378371A - A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor - Google Patents
A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor Download PDFInfo
- Publication number
- CN110378371A CN110378371A CN201910503050.4A CN201910503050A CN110378371A CN 110378371 A CN110378371 A CN 110378371A CN 201910503050 A CN201910503050 A CN 201910503050A CN 110378371 A CN110378371 A CN 110378371A
- Authority
- CN
- China
- Prior art keywords
- nearest neighbor
- neighbor distance
- energy consumption
- average nearest
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000005265 energy consumption Methods 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000005856 abnormality Effects 0.000 title claims abstract description 35
- 230000002159 abnormal effect Effects 0.000 claims abstract description 21
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000003064 k means clustering Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 22
- 230000007717 exclusion Effects 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 230000009194 climbing Effects 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 229910021420 polycrystalline silicon Inorganic materials 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06Q50/04—Manufacturing
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Manufacturing & Machinery (AREA)
- General Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Water Supply & Treatment (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Public Health (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Quality & Reliability (AREA)
- Automation & Control Theory (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The present invention provides a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor, comprising: acquisition energy consumption data is simultaneously converted into alternation data；The time series characteristic value of energy consumption data is defined, and time series is divided into subsequence and is respectively mapped to four dimensional feature spaces；The average nearest neighbor distance Outlier factor of time subsequence is calculated separately in four dimensional feature spaces；The average nearest neighbor distance Outlier factor of subsequence is handled, the average nearest neighbor distance Outlier factor of time series is obtained；Average nearest neighbor distance outlier threshold is calculated according to average nearest neighbor distance Outlier factor, judges whether mode exception occur.A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor provided by the invention, the interference of exclusion mode abnormality detection, effectively improve the precision of mode abnormality detection, abnormal position is accurately positioned, the abnormal data occurred in phase step type and alternation energy consumption data is effectively detected out.
Description
Technical field
The present invention relates to energy consumption data abnormality detections and power consumption state monitoring technical field, more particularly to a kind of base
In the energy consumption method for detecting abnormality of average nearest neighbor distance Outlier factor.
Background technique
With the development of industrial information technology, many high energy consumption enterprises set up energy management system to realize energy consumption data
Real-time acquisition, the purpose is to optimize the energy operation and save production cost.Therefore, reality of the high energy consumption enterprise for energy consumption data
When detect and monitoring higher requirements are also raised that it is necessary to have real-times and intelligentized detection technique.High energy consumption machine
Working condition is complicated, long-term oepration at full load, and the probability for abnormal energy consumption occur is very high [1], and [2] thereby produce a large amount of exceptions
Energy consumption data.In process of production, the data of many high energy consumptions mechanical (such as hydraulic press, polycrystalline silicon reducing furnace) have step, hand over
The features such as replacing, be periodical.The abnormal energy consumption of machine is resulted even in along with the reduction of a large amount of energy loss and energy efficiency
Shutdown and imponderable safety accident, to influence the normal production of the entire production line.Therefore, develop it is a kind of reliable, quickly,
The abnormal energy consumption detection technique of automation has a very important significance.Using these new methods, manufacturing enterprise can be to high energy
Consumption machinery is monitored and handles, and avoids energy loss, improves source of mechanical energy efficiency.Mode is in energy consumption abnormal data extremely
A kind of typical case's anomaly pattern, institute's facing challenges are to develop a kind of reliable, mode abnormality detection skill fast and automatically at present
Art.
Currently, abnormality detection is generally by manual inspection, recording apparatus data, there is very big hysteresis quality in this.And
It is a changeless threshold value by the energy consumption alarm threshold that artificial experience is set, does not adapt to energy consumption data real-time change
Demand.Time series data method for detecting abnormality includes: the abnormality detection based on distance, the abnormality detection based on prediction, base
In the methods of abnormality detection of cluster [3] [4] [5].Traditional method for detecting abnormality is classified largely into local anomaly detection method
With global abnormal detection method.
Local anomaly detection method due to excessively focusing on local small variation, cause higher false detection rate and it is lower can
Scalability；And global abnormal detection method, due to ignoring local slight abnormality, rate of failing to report is very high.It is appropriate due to lacking
Optimization, their detection efficiency is low, bad adaptability [6].Therefore, phase step type and alternation cannot be effectively detected out in the prior art
The abnormal data occurred in energy consumption data.
Summary of the invention
The present invention is to overcome existing energy consumption method for detecting abnormality presence that phase step type and alternation can not be effectively detected out
It is abnormal to provide a kind of energy consumption based on average nearest neighbor distance Outlier factor for the technological deficiency of the abnormal data occurred in energy consumption data
Detection method.
In order to solve the above technical problems, technical scheme is as follows:
A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor, comprising the following steps:
S1: acquisition energy consumption data converts alternation data for energy consumption data by rain flow method；
S2: the time series characteristic value of energy consumption data is defined, and time series is divided into subsequence and is respectively mapped to four
Dimensional feature space；
S3: the average nearest neighbor distance Outlier factor of time subsequence is calculated separately in four dimensional feature spaces；
S4: the average nearest neighbor distance Outlier factor of subsequence is handled, the average nearest neighbor distance of time series is obtained
Outlier factor；
S5: average nearest neighbor distance outlier threshold is calculated according to average nearest neighbor distance Outlier factor, judges whether mode occur
It is abnormal.
Wherein, the step S1 specifically includes the following steps:
S11: acquisition energy consumption data obtains the time series X=(x of energy consumption data1,x2,...xn) and its cycle length CL,
Sliding window length m, the weight factor λ of off-note；
S12: alternation data, specific formula for calculation are converted for energy consumption data by rain flow method are as follows:
(Xi-Xi-1)(Xi-Xi+1) < 0i=2,3,4 ..., M-1；
Wherein, XiIt is the energy consumption data of collection in worksite；The endpoint of data segment is peak valley point, according to peak valley sequence, i.e. PV sequence
The standard of column is filtered, and is deleted all non-peak valley points, is deleted the interference data of ascents and descents, obtain alternation data.
Wherein, the step S2 specifically includes the following steps:
S21: according to sliding window length m by time series X=(x1,x2,...xn) intercept as multiple subsequences；
S22: four characteristic values of time subsequence are defined, are specifically included:
Subsequence height h is defined as:
H=xmax-xmin；
Subsequence mean valueIs defined as:
The variances sigma of subsequence2Is defined as:
The maximal contiguous distance m of subsequence is defined as:
M=max (| xi-xi-1|), (i=2,3,4...n)；
S23: time subsequence is mapped to four dimensional feature spacesIn.
Wherein, the step S3 first uses K-Means clustering method to cluster multi-energy data for 2 classes, then using average close
Neighborhood distance Outlier factor MNNDAF algorithm calculates separately the average nearest neighbor distance Outlier factor of 2 class time serieses.
Wherein, the MNNDAF algorithm specifically includes the following steps:
S31: it sets up an officeAnd pointFor in four dimensional feature spaces's
Arbitrary point, then the Euclidean distance of feature space indicates are as follows:
S32: defining average nearest neighbor distance MNND is average value of the point p at a distance from all Neighbor Points, remembers that the k- of point p is close
Neighborhood distance is k-dist (p)；If k ∈ N+,The then average nearest neighbor distance MNND of point p is defined as:
S33: it setsThe subspace of eigenvalue of four characteristic values of point c is C respectivelyh(h1,h2,...,
hn),Cσ(σ1,σ2,...,σn),Cm(m1,m2,...,mn), point c is calculated separately in four dimensional feature spacesAverage nearest neighbor distance MNND in subspace of eigenvalue, is denoted as MNND (c), MNNDh (c) respectively,MNND σ (c) and MNNDm (c) so far obtain the average nearest neighbor distance Outlier factor of point c are as follows:
MNNDAF (c)=MNND (c)+max { X }；
Wherein,λ is
The weight factor of off-note.
Wherein, the step S4 specifically includes the following steps:
S41: the average nearest neighbor distance Outlier factor value of subsequence is subjected to interpolation in the multi-energy data of 2 classes；
S42: the multi-energy data after interpolation is normalized；
S43: the multi-energy data after normalized is resequenced according to the sequence of initial data, obtains time series
Average nearest neighbor distance Outlier factor.
Wherein, the detailed process of normalized described in the step S42 are as follows:
A mapping from c to c' is established, so that the L of c'2Norm is 1, then has:
Wherein, feature value vector c (c1,c2,...cn) L2Norm is For normalizing
Change coefficient.
Wherein, in the step S5, the specific formula for calculation of average nearest neighbor distance outlier threshold δ are as follows:
Wherein, α is the mode exception coefficient set according to actual conditions, if MNNDAF > δ, by the time of energy consumption data
Sequence is judged as mode exception.
Compared with prior art, the beneficial effect of technical solution of the present invention is:
A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor provided by the invention, using rain flowmeter
Number method excludes the interference of mode abnormality detection；The precision that mode abnormality detection is improved using K-Means clustering method, to exception bits
It sets and is accurately positioned；The mode for accurately detecting out multi-energy data time series using MNNDAF algorithm is abnormal, by above-mentioned
The abnormal data occurred in phase step type and alternation energy consumption data is effectively detected out in method.
Detailed description of the invention
Fig. 1 is the flow diagram of this method；
Fig. 2 is original multi-energy data time series schematic diagram；
Fig. 3 is rain flow method treated multi-energy data time series schematic diagram；
Fig. 4 is mode abnormality detection result schematic diagram.
Specific embodiment
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent；
In order to better illustrate this embodiment, the certain components of attached drawing have omission, zoom in or out, and do not represent actual product
Size；
To those skilled in the art, it is to be understood that certain known features and its explanation, which may be omitted, in attached drawing
's.
The following further describes the technical solution of the present invention with reference to the accompanying drawings and examples.
Embodiment 1
As shown in Figure 1, a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor, including following step
It is rapid:
S1: acquisition energy consumption data converts alternation data for energy consumption data by rain flow method；
S2: the time series characteristic value of energy consumption data is defined, and time series is divided into subsequence and is respectively mapped to four
Dimensional feature space；
S3: the average nearest neighbor distance Outlier factor of time subsequence is calculated separately in four dimensional feature spaces；
S4: the average nearest neighbor distance Outlier factor of subsequence is handled, the average nearest neighbor distance of time series is obtained
Outlier factor；
S5: average nearest neighbor distance outlier threshold is calculated according to average nearest neighbor distance Outlier factor, judges whether mode occur
It is abnormal.
More specifically, the step S1 specifically includes the following steps:
S11: acquisition energy consumption data obtains the time series X=(x of energy consumption data1,x2,...xn) and its cycle length CL,
Sliding window length m, the weight factor λ of off-note；
S12: alternation data, specific formula for calculation are converted for energy consumption data by rain flow method are as follows:
(Xi-Xi-1)(Xi-Xi+1) < 0i=2,3,4 ..., M-1；
Wherein, XiIt is the energy consumption data of collection in worksite；The endpoint of data segment is peak valley point, according to peak valley sequence, i.e. PV sequence
The standard of column is filtered, and is deleted all non-peak valley points, is deleted the interference data of ascents and descents, obtain alternation data.
More specifically, the step S2 specifically includes the following steps:
S21: according to sliding window length m by time series X=(x1,x2,...xn) intercept as multiple subsequences；
S22: four characteristic values of time subsequence are defined, are specifically included:
Subsequence height h is defined as:
H=xmax-xmin；
Subsequence mean valueIs defined as:
The variances sigma of subsequence2Is defined as:
The maximal contiguous distance m of subsequence is defined as:
M=max (| xi-xi-1|), (i=2,3,4...n)；
S23: time subsequence is mapped to four dimensional feature spacesIn.
More specifically, the step S3 first uses K-Means clustering method to cluster multi-energy data for 2 classes, then using flat
Equal nearest neighbor distance Outlier factor MNNDAF algorithm calculates separately the average nearest neighbor distance Outlier factor of 2 class time serieses.
More specifically, the MNNDAF algorithm specifically includes the following steps:
S31: it sets up an officeAnd pointFor in four dimensional feature spaces's
Arbitrary point, then the Euclidean distance of feature space indicates are as follows:
S32: defining average nearest neighbor distance MNND is average value of the point p at a distance from all Neighbor Points, remembers that the k- of point p is close
Neighborhood distance is k-dist (p)；If k ∈ N+,The then average nearest neighbor distance MNND of point p is defined as:
S33: it setsThe subspace of eigenvalue of four characteristic values of point c is C respectivelyh(h1,h2,...,
hn),Cσ(σ1,σ2,...,σn),Cm(m1,m2,...,mn), point c is calculated separately in four dimensional feature spacesAverage nearest neighbor distance MNND in subspace of eigenvalue, is denoted as MNND (c), MNNDh (c) respectively,MNND σ (c) and MNNDm (c) so far obtain the average nearest neighbor distance Outlier factor of point c are as follows:
MNNDAF (c)=MNND (c)+max { X }；
Wherein,λ is
The weight factor of off-note.
More specifically, the step S4 specifically includes the following steps:
S41: the average nearest neighbor distance Outlier factor value of subsequence is subjected to interpolation in the multi-energy data of 2 classes；
S42: the multi-energy data after interpolation is normalized；
S43: the multi-energy data after normalized is resequenced according to the sequence of initial data, obtains time series
Average nearest neighbor distance Outlier factor.
More specifically, the detailed process of normalized described in the step S42 are as follows:
A mapping from c to c' is established, so that the L of c'2Norm is 1, then has:
Wherein, feature value vector c (c1,c2,...cn) L2Norm is For normalizing
Change coefficient.
More specifically, in the step S5, the specific formula for calculation of average nearest neighbor distance outlier threshold δ are as follows:
Wherein, α is the mode exception coefficient set according to actual conditions, if MNNDAF > δ, by the time of energy consumption data
Sequence is judged as mode exception.
In the specific implementation process, this method excludes the interference of mode abnormality detection using rain flow method；Using K-
Means clustering method improves the precision of mode abnormality detection, is accurately positioned to abnormal position；It is accurate using MNNDAF algorithm
Ground detects that the mode of multi-energy data time series is abnormal, and phase step type is effectively detected out by the above method and alternation performance consumes
The abnormal data occurred in data.
Embodiment 2
More specifically, on the basis of embodiment 1, proposed method for detecting abnormality is realized by MATLAB programming, compiled
Translate tool are as follows: MATLAB R2018a, running environment: Windows 7 or more version, hardware: client computer CPU3.3G
More than, memory 4,0M.
In the specific implementation process, it is first directed to collected multi-energy data, relevant parameter is adjusted, specifically includes: week
Phase length CL=100；Setting classification number, i.e., calculation n=2 in initial cluster；Sliding window length m=5；Normalization coefficientMode exception factor alpha=1.4.
As shown in Fig. 2, one shares 5 periods in original multi-energy data time series schematic diagram.Black oval mark
Three models exception has been outpoured, has needed detected.Wherein " climb " and " descending " position at data, and " climbing "
Preceding " examination squeezes " data are interference data, need before testing to be purged it.
In rain flow method treated multi-energy data time series fig. 3, it is shown that by rain flow method
Processing, the interference data at " examination squeezes " position before having disposed " climbing ", " descending " and " climbing "；It will " load is advanced " number
Translate up distance d according to section, effectively increase the discrimination of " idle running " and " load running ", K-Means clustering method into
Row classification；Such as Fig. 4 as can be seen that method for detecting abnormality proposed by the present invention, has correctly detected out " idle running " and " load
Mode at operation " is abnormal.
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all this
Made any modifications, equivalent replacements, and improvements etc., should be included in the claims in the present invention within the spirit and principle of invention
Protection scope within.
[1]Li L,Huang H,Zhao F,et al.Operation scheduling of multi–hydraulic
press system for energy consumption reduction[J].Journal of Cleaner
Production,2017,165.
[2]Gao M,Li X,Huang H,et al.Energy–saving Methods for Hydraulic
Presses Based on Energy Dissipation Analysis[J].Procedia Cirp,2016,48:331–
335.
[3]Ahmed M,Mahmood A N,Maher M J.Heart Disease Diagnosis Using Co–
clustering[M]//Scalable Information Systems.Springer International
Publishing,2014:61–70.
[4]Cecílio I M,Ottewill J R,Pretlove J,et al.Nearest neighbors method
for detecting transient disturbances in process and electromechanical systems
[J].Journal of Process Control,2014,24(9):1382–1393.
[5]Drugman T.Using mutual information in supervised temporal event
detection:Application to cough detection[J].Biomedical Signal Processing&
Control,2014,10(1):50–57.
[6]Ren H,Ye Z,Li Z.Anomaly detection based on a dynamic Markov model
[J].Information Sciences,2017,411.
Claims (8)
1. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor, which comprises the following steps:
S1: acquisition energy consumption data converts alternation data for energy consumption data by rain flow method；
S2: the time series characteristic value of energy consumption data is defined, and time series is divided into subsequence and is respectively mapped to four-dimensional spy
Levy space；
S3: the average nearest neighbor distance Outlier factor of time subsequence is calculated separately in four dimensional feature spaces；
S4: the average nearest neighbor distance Outlier factor of subsequence is handled, and the average nearest neighbor distance for obtaining time series is abnormal
The factor；
S5: average nearest neighbor distance outlier threshold is calculated according to average nearest neighbor distance Outlier factor, judges whether mode exception occur.
2. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 1, special
Sign is, the step S1 specifically includes the following steps:
S11: acquisition energy consumption data obtains the time series X=(x of energy consumption data1,x2,...xn) and its cycle length CL, sliding
Length of window m, the weight factor λ of off-note；
S12: alternation data, specific formula for calculation are converted for energy consumption data by rain flow method are as follows:
(Xi-Xi-1)(Xi-Xi+1) < 0i=2,3,4 ..., M-1；
Wherein, XiIt is the energy consumption data of collection in worksite；The endpoint of data segment is peak valley point, according to peak valley sequence, the i.e. mark of PV sequence
Standard is filtered, and is deleted all non-peak valley points, is deleted the interference data of ascents and descents, obtain alternation data.
3. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 2, special
Sign is, the step S2 specifically includes the following steps:
S21: according to sliding window length m by time series X=(x1,x2,...xn) intercept as multiple subsequences；
S22: four characteristic values of time subsequence are defined, are specifically included:
Subsequence height h is defined as:
H=xmax-xmin；
Subsequence mean valueIs defined as:
The variances sigma of subsequence2Is defined as:
The maximal contiguous distance m of subsequence is defined as:
M=max (| xi-xi-1|), (i=2,3,4...n)；
S23: time subsequence is mapped to four dimensional feature spacesIn.
4. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 3, special
Sign is that the step S3 first uses K-Means clustering method to cluster multi-energy data for 2 classes, then using average nearest neighbor distance
Outlier factor MNNDAF algorithm calculates separately the average nearest neighbor distance Outlier factor of 2 class time serieses.
5. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 4, special
Sign is, the MNNDAF algorithm specifically includes the following steps:
S31: it sets up an officeAnd pointFor in four dimensional feature spacesIt is any
Point, then the Euclidean distance of feature space indicates are as follows:
S32: defining average nearest neighbor distance MNND is average value of the point p at a distance from all Neighbor Points, the k- neighbour of note point p away from
From for k-dist (p)；If k ∈ N+,The then average nearest neighbor distance MNND of point p is defined as:
S33: it setsThe subspace of eigenvalue of four characteristic values of point c is C respectivelyh(h1,h2,...,hn),Cσ(σ1,σ2,...,σn),Cm(m1,m2,...,mn), point c is calculated separately in four dimensional feature spacesAverage nearest neighbor distance MNND in subspace of eigenvalue, is denoted as MNND (c), MNNDh (c) respectively,MNND σ (c) and MNNDm (c) so far obtain the average nearest neighbor distance Outlier factor of point c are as follows:
MNNDAF (c)=MNND (c)+max { X }；
Wherein,λ is abnormal
The weight factor of feature.
6. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 5, special
Sign is, the step S4 specifically includes the following steps:
S41: the average nearest neighbor distance Outlier factor value of subsequence is subjected to interpolation in the multi-energy data of 2 classes；
S42: the multi-energy data after interpolation is normalized；
S43: the multi-energy data after normalized is resequenced according to the sequence of initial data, obtains being averaged for time series
Nearest neighbor distance Outlier factor.
7. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 6, special
Sign is, the detailed process of normalized described in the step S42 are as follows:
A mapping from c to c' is established, so that the L of c'2Norm is 1, then has:
Wherein, feature value vector c (c1,c2,...cn) L2Norm is For normalization system
Number.
8. a kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor according to claim 6, special
Sign is, in the step S5, the specific formula for calculation of average nearest neighbor distance outlier threshold δ are as follows:
Wherein, α is the mode exception coefficient set according to actual conditions, if MNNDAF > δ, by the time series of energy consumption data
It is judged as mode exception.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910503050.4A CN110378371B (en) | 2019-06-11 | 2019-06-11 | Energy consumption abnormity detection method based on average neighbor distance abnormity factor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910503050.4A CN110378371B (en) | 2019-06-11 | 2019-06-11 | Energy consumption abnormity detection method based on average neighbor distance abnormity factor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110378371A true CN110378371A (en) | 2019-10-25 |
| CN110378371B CN110378371B (en) | 2022-12-16 |
Family
ID=68250136
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910503050.4A Active CN110378371B (en) | 2019-06-11 | 2019-06-11 | Energy consumption abnormity detection method based on average neighbor distance abnormity factor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110378371B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112381181A (en) * | 2020-12-11 | 2021-02-19 | 桂林电子科技大学 | Dynamic detection method for building energy consumption abnormity |
| CN112966017A (en) * | 2021-03-01 | 2021-06-15 | 北京青萌数海科技有限公司 | Abnormal subsequence detection method with indefinite length in time sequence |
| CN113048807A (en) * | 2021-03-15 | 2021-06-29 | 太原理工大学 | Air cooling unit backpressure abnormity detection method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005108607A1 (en) * | 2004-05-09 | 2005-11-17 | Technion Research & Development Foundation Ltd. | Compositions and methods for treating disorders associated with abnormal phosphate metabolism |
| CN102706563A (en) * | 2012-06-14 | 2012-10-03 | 哈尔滨工业大学 | Detection method for neighbor abnormities of gas turbine |
| CN106330624A (en) * | 2016-11-07 | 2017-01-11 | 国网江苏省电力公司南京供电公司 | Method for detecting power information network traffic abnormality |
| CN107818135A (en) * | 2017-09-26 | 2018-03-20 | 广东电网有限责任公司电力调度控制中心 | A kind of Wei Nuotu electric power big data method for detecting abnormality based on Grey Incidence |
| CN108435819A (en) * | 2018-05-29 | 2018-08-24 | 广东工业大学 | A kind of aluminum section extruder energy consumption method for detecting abnormality |
| US20180316707A1 (en) * | 2017-04-26 | 2018-11-01 | Elasticsearch B.V. | Clustering and Outlier Detection in Anomaly and Causation Detection for Computing Environments |
| WO2019012726A1 (en) * | 2017-07-14 | 2019-01-17 | Kabushiki Kaisha Toshiba | Abnormality detection device, abnormality detection method, and non-transitory computer readable medium |
-
2019
- 2019-06-11 CN CN201910503050.4A patent/CN110378371B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005108607A1 (en) * | 2004-05-09 | 2005-11-17 | Technion Research & Development Foundation Ltd. | Compositions and methods for treating disorders associated with abnormal phosphate metabolism |
| CN102706563A (en) * | 2012-06-14 | 2012-10-03 | 哈尔滨工业大学 | Detection method for neighbor abnormities of gas turbine |
| CN106330624A (en) * | 2016-11-07 | 2017-01-11 | 国网江苏省电力公司南京供电公司 | Method for detecting power information network traffic abnormality |
| US20180316707A1 (en) * | 2017-04-26 | 2018-11-01 | Elasticsearch B.V. | Clustering and Outlier Detection in Anomaly and Causation Detection for Computing Environments |
| WO2019012726A1 (en) * | 2017-07-14 | 2019-01-17 | Kabushiki Kaisha Toshiba | Abnormality detection device, abnormality detection method, and non-transitory computer readable medium |
| CN107818135A (en) * | 2017-09-26 | 2018-03-20 | 广东电网有限责任公司电力调度控制中心 | A kind of Wei Nuotu electric power big data method for detecting abnormality based on Grey Incidence |
| CN108435819A (en) * | 2018-05-29 | 2018-08-24 | 广东工业大学 | A kind of aluminum section extruder energy consumption method for detecting abnormality |
Non-Patent Citations (4)
| Title |
|---|
| 周珂仪: "动态系统长时间运行过程的异常变化检测", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 曾利云 等: "基于时间序列和聚类的挤压机能耗异常检测研究", 《机电工程技术》 * |
| 肖冬桂: "塔机疲劳剩余寿命预测系统研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 路世青 等: "港口起重机金属结构疲劳寿命分析与评价", 《物流工程三十年技术创新发展之道》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112381181A (en) * | 2020-12-11 | 2021-02-19 | 桂林电子科技大学 | Dynamic detection method for building energy consumption abnormity |
| CN112381181B (en) * | 2020-12-11 | 2022-10-04 | 桂林电子科技大学 | Dynamic detection method for building energy consumption abnormity |
| CN112966017A (en) * | 2021-03-01 | 2021-06-15 | 北京青萌数海科技有限公司 | Abnormal subsequence detection method with indefinite length in time sequence |
| CN112966017B (en) * | 2021-03-01 | 2023-11-14 | 北京青萌数海科技有限公司 | Abnormal subsequence detection method for indefinite length in time sequence |
| CN113048807A (en) * | 2021-03-15 | 2021-06-29 | 太原理工大学 | Air cooling unit backpressure abnormity detection method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110378371B (en) | 2022-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110362608A (en) | Energy consumption method for detecting abnormality based on rain flow method and local outlier factor | |
| Wu et al. | A multi-level-denoising autoencoder approach for wind turbine fault detection | |
| WO2022011754A1 (en) | Fault diagnosis method based on adaptive manifold embedded dynamic distribution alignment | |
| CN110378371A (en) | A kind of energy consumption method for detecting abnormality based on average nearest neighbor distance Outlier factor | |
| Mao et al. | Online detection of bearing incipient fault with semi-supervised architecture and deep feature representation | |
| Jia et al. | A deviation based assessment methodology for multiple machine health patterns classification and fault detection | |
| CN110895526A (en) | Method for correcting data abnormity in atmosphere monitoring system | |
| US8630962B2 (en) | Error detection method and its system for early detection of errors in a planar or facilities | |
| CN106600074B (en) | DFHSMM-based non-invasive power load monitoring method and system | |
| CN104994535B (en) | Sensor data stream method for detecting abnormality based on Multidimensional Data Model | |
| De Baets et al. | VI-based appliance classification using aggregated power consumption data | |
| CN106709816B (en) | Non-parametric regression analysis-based power load abnormal data identification and correction method | |
| CN117290802B (en) | Host power supply operation monitoring method based on data processing | |
| Salehi et al. | A relevance weighted ensemble model for anomaly detection in switching data streams | |
| CN113609901A (en) | Power transmission and transformation equipment fault monitoring method and system | |
| Xu et al. | A lof-based method for abnormal segment detection in machinery condition monitoring | |
| Zhao et al. | Collaborative fault detection for large-scale photovoltaic systems | |
| Sun et al. | Feature extraction and pattern identification for anemometer condition diagnosis | |
| CN115220396A (en) | Intelligent monitoring method and system for numerical control machine tool | |
| Liu et al. | A deep generative model based on CNN-CVAE for wind turbine condition monitoring | |
| Yan et al. | Robust event detection for residential load disaggregation | |
| Li et al. | Data driven condition monitoring of wind power plants using cluster analysis | |
| Liu et al. | Hierarchical context-aware anomaly diagnosis in large-scale PV systems using SCADA data | |
| Bhargava et al. | Anomaly detection in wireless sensor networks using S-Transform in combination with SVM | |
| Li et al. | A fusion framework using integrated neural network model for non-intrusive load monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |