CN104166718A - Bad data detection and recognition method suitable for large power grid - Google Patents
Bad data detection and recognition method suitable for large power grid Download PDFInfo
- Publication number
- CN104166718A CN104166718A CN201410407189.6A CN201410407189A CN104166718A CN 104166718 A CN104166718 A CN 104166718A CN 201410407189 A CN201410407189 A CN 201410407189A CN 104166718 A CN104166718 A CN 104166718A
- Authority
- CN
- China
- Prior art keywords
- data
- sus
- bad
- suspicious
- bad 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
- 238000000034 method Methods 0.000 title claims abstract description 95
- 238000001514 detection method Methods 0.000 title claims abstract description 48
- 238000005259 measurement Methods 0.000 claims abstract description 24
- 238000004458 analytical method Methods 0.000 claims abstract description 6
- 238000013480 data collection Methods 0.000 claims description 45
- 239000011159 matrix material Substances 0.000 claims description 35
- 230000035945 sensitivity Effects 0.000 claims description 10
- 238000012360 testing method Methods 0.000 claims description 4
- 239000012141 concentrate Substances 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 abstract 2
- 230000003044 adaptive effect Effects 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 4
- 238000011109 contamination Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Primary Health Care (AREA)
- Computational Linguistics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention relates to a bad data detection and recognition method suitable for a large power grid. The method is applied to evaluation of the state of an electric power system. The method comprises the following steps that bad data are detected, and suspect data sets in the bad data are determined; related suspect data sets in the suspect data sets are determined; bad data in the suspect data sets are determined; the bad data in the suspect data sets are deleted. The method is widely applied to current provincial-level power grids and a national power grid, can be adaptive to online analysis requirements of the large power grid in the future, and can quickly and accurately delete bad data in measurement information, and improve state evaluation precision.
Description
Technical field:
The present invention relates to a kind of bad data detection and identification method, more specifically relate to a kind of bad data detection and identification method that is applicable to large electrical network.
Background technology:
Power system state estimation is the important component part of MODERN ENERGY management system, and its metric data major part derives from SCADA system, and information, except containing normal measurement noise, also may contain bad data.The existence of bad data, will cause estimated result to be polluted, and even make it serious distortion.The detection and identification of bad data is one of critical function of Power system state estimation, and its object is to get rid of and measures a small amount of bad data accidentally occurring in sampled data, to improve the reliability of state estimation.
Power system state estimation is to utilize the redundance of real-time measurement system to improve data precision, gets rid of the caused error message of random disturbance, and then the running status of estimation or prognoses system.The method of the bad data detection and identification based on state estimation mainly contains residual error search procedure, Non quadratic criteria method, zero residual error method and estimates identification method.These methods are mainly using weighted residual or residual value as eigenwert, suppose that it obeys a certain probability distribution, and determine a threshold value according to certain level of confidence, carry out test of hypothesis.Find after suspicious measurement data, it is got rid of from measurement data or reduce its weights, obtain new state estimation value.The shortcoming that may exist in above detection and identification method: may occur that residual contamination and residual error flood phenomenon, thereby cause undetected or flase drop, affect the effect of identification, further affect estimation effect.Because algorithm adopts non-linear residual equation, in identification process, need to carry out repeatedly state estimation, so calculated amount is very big, and state estimation has certain requirement to real-time, unsuitable as online real-time estimation method.In addition adopt linearization residual equation, utilize the estimated value of the submatrix computation and measurement of residual sensitivity matrix, because sensitivity matrix is the full battle array of higher-dimension, therefore this method calculated amount is still very large.In addition, when occurring in the situation of a plurality of bad datas, the phenomenon of the identification that makes often can to make a mistake in this way.For the defect in the whole bag of tricks, also there is much improved research branch in state estimation method.
Traditional detection and discrimination method are usingd weighted residual rw or residual rn as bad data identification method, obtain after suspicious metric data collection, reduce one by one its weights or directly from metric data, reject, then re-starting state estimation, so circulation is until meet the condition of convergence.The shortcoming of these class methods is that calculated amount is very large, and computing velocity is slow; If define detection number of times, easily occur that again residual contamination and residual error flood phenomenon, cause undetected or flase drop phenomenon, affect identification effect.Especially few at metric data, and data value is inaccurate, and in insecure situation (in power distribution network), it is even more serious that residual contamination and residual error are flooded phenomenon.
Power system state estimation at home and abroad develops decades, and the research of bad data detection and identification is also never stagnated, but still the effective ways of a good detection and identification bad data are not proposed, a kind of bad data detection and identification method that is applicable to large electrical network is now proposed to overcome above-mentioned defect.
Summary of the invention:
The object of this invention is to provide a kind of bad data detection and identification method that is applicable to large electrical network, the method is widely used in current net provincial power network and nationwide integrated power grid, and can adapt to following scale grid line analysis demand, can reject fast and accurately the bad data in measurement information, improve precision of state estimation.
For achieving the above object, the present invention by the following technical solutions: a kind of bad data detection and identification method that is applicable to large electrical network, described method is used in Power system state estimation process; Said method comprising the steps of:
(1) determine the suspicious data collection SUS of the whole network;
(2) determine the associated suspicious data collection sus[n that described suspicious data is concentrated];
(3) the associated suspicious data collection of identification sus[n] in bad data.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, the deterministic process of described step (1) is:
(1-1) carry out the state estimation of the whole network;
(1-2) pass through r
ndetection method is determined suspicious data collection SUS;
Described r
nthe testing process of detection method is:
In formula, H
0be that i metric data is not suspicious metric data, H
1be that i metric data is suspicious metric data; r
n,ibe i standardized residual,
it is the threshold value of i standardized residual.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, the threshold value of described standardized residual is determined by following steps:
Definition W and W
nthe sensitivity matrix that is respectively m * m rank residual sensitivity matrix and m * m rank standardized residual, has
Standardized residual r
nfor:
r
N=W
Nv (3)
Wherein, diagonal matrix D=diag[WR], the weight matrix that R is and measurement is corresponding, v is the m dimension error vector that contains bad data;
Under normal measurement condition, there is standardized residual r
ncovariance matrix be:
Wherein, E represents expectation function, r
n,zfor the standardized residual of all measurements, therefore have
So when getting probability of false detection P
e=0.005, get the threshold value of described standardized residual
for
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, the deterministic process of described step (2) is:
(2-1) search of incidence matrix;
(2-2) finally form associated suspicious data collection.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, described step (2-1) is in Power system state estimation, and the associated suspicious data search procedure based on Jacobi matrix is:
Step (2-1-1): the number n that associated suspicious data collection is set is 1;
Step (2-1-2): newly-built n associated suspicious data collection sus[n];
Step (2-1-3): judge whether all suspicious data collection of the whole network SUS is empty, if be not idle running step (2-1-4), empty if, search procedure finishes;
Step (2-1-4): take out the data of standardized residual maximum from SUS, be assumed to be i data, these data are rejected from SUS, and join sus[n] in;
Step (2-1-5): search for the nonzero element in Jacobi matrix i row element, and record the row number of these nonzero elements, form set LOR;
Step (2-1-6): take out successively row number from LOR, and search for the row that Jacobi matrix is corresponding, in each row of record, the line number of non-zero element, forms set ROW;
Step (2-1-7): judge successively in ROW, whether data corresponding to line number belong to S set US, if there are the data that belong to S set US, all data that belong to S set US are joined to associated suspicious data collection sus[n] in, the data that these belonged to S set US are rejected simultaneously from SUS, go to step (2-1-8); If there is no the data that belong to S set US, n associated suspicious measurement collection sus[n] form completely, carry out n=n+1, and go to step (2-1-2);
Step (2-1-8): take out in step (2-1-7) and belong to line number corresponding to data of S set US, and search for the non-zero element in Jacobi matrix corresponding row element, record non-zero element column number, form LOR, go to step (2-1-6);
By above 8 steps, form n associated suspicious data collection of described step (2-2).
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, according to described associated suspicious data collection sus[n] calculate each sus[n] weighted residual quadratic sum J (x):
J(x)=[z-h(x)]
TR
-1[z-h(x)] (7)
Wherein, z is associated suspicious data, and R is corresponding diagonal angle weight matrix, and h (x) is the computing function of associated suspicious data.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, described step (3) is determined bad data by double-deck bad data identification method.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, described double-deck bad data identification method process is:
Step (8-1): the suspicious measurement collection of current association to be processed sequence number k=1 is set;
Step (8-2): take out k associated suspicious data collection, calculate the weighted residual quadratic sum of all data in this set, be designated as J (x);
Step (8-3): take out k associated suspicious data and concentrate standardized residual maximum and not by the data sus[k of identification, i],, based on Givens orthogonal row converter technique, revise fast state estimation factor table and delete fast described suspicious data sus[k, i];
Step (8-4): associated suspicious data collection sus[k described in rapid solving] in all data described weighted residual quadratic sum J ' (x);
Step (8-5): described in identification, whether suspicious data is bad data;
If described suspicious data sus[k, i] be bad data, and go to step (8-6); If described suspicious data sus[k, i] not bad data, recover described suspicious data sus[k, i] to associated suspicious data collection sus[k] in, put identification mark, and by the quick modifying factor sublist of Givnes orthogonal row converter technique, then go to step (8-3);
Step (8-6): described in identification, associated suspicious data is concentrated and whether also had bad data;
If described associated suspicious data is concentrated, there is no bad data, continued next associated suspicious data set analysis, carried out k=k+1, and go to step (8-7), k is associated suspicious data collection number; If described associated suspicious data is concentrated, also there is bad data, need to continue identification, go to step (8-3);
Step (8-7): judge whether k is greater than n, if so, bad data identification process finishes; If not, go to step (8-2).
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, described step (8-5) judges whether to exist bad data by following formula:
| J (x)-J ' (x) | > ε
1, ε
1for threshold value; If inequality is set up, described data are bad data, if inequality is false, described data do not exist bad data;
Described step (8-6) judges the concentrated bad data that whether also exists of described associated suspicious data by following formula:
J ' is < ε (x)
2, ε
2for threshold value; If inequality is set up, described associated suspicious data is concentrated has not had bad data; If inequality is false, described associated suspicious data is concentrated and is also had bad data.
The invention provides a kind of bad data detection and identification method that is applicable to large electrical network, after completing detection and identification process, according to the described bad data picking out, the data respective weights of described Power system state estimation is revised, and carry out iteratively faster calculating, obtain described Power system state estimation data section accurately.
With immediate prior art ratio, the invention provides technical scheme and there is following excellent effect
1, the present invention is the important component part that power state is estimated, is requisite measure and the gordian technique that improves precision of state estimation;
2, the present invention can fast and accurately reject the bad data that error is larger, accurately in the bad data of location, can avoid unnecessary search;
3, the present invention can improve the sensitivity of bad data detection and identification and accuracy, finally improves state estimation computational accuracy;
4, the present invention has more wide application prospect, is widely used in current net provincial power network and nationwide integrated power grid, and can adapt to following scale grid line analysis demand;
5, the present invention is in the Demonstration Application of scheduling institutions at different levels, can further promote accuracy and computing velocity that intelligent grid supporting system technology state estimation is calculated, support the ability of the becoming more meticulous of intelligent grids scheduling at different levels, lean and integrative operation comprehensively;
6, the present invention can effectively improve the real-time of ultra-large powernet analytical calculation, for safety, high-quality and the economical operation of the large electrical network of extra-high voltage provides strong technical support;
7, the present invention makes dispatcher follow the tracks of in time operation of power networks state by accurate status estimated result, the operation conditions that notes abnormalities is also made control decision, avoid fault spread, further promote intelligent grid supporting system technology technical merit and operation stability, will further promote dispatching of power netwoks and control the ability of large electrical network;
8, the present invention ensures large power grid security, stable, high-quality, economical operation, to promoting electrical power services quality and guaranteeing that social stable development has important realistic meaning.
Accompanying drawing explanation
Fig. 1 is straightforward procedure process flow diagram of the present invention;
Fig. 2 is concrete grammar process flow diagram of the present invention.
Embodiment
Below in conjunction with embodiment, the invention will be described in further detail.
Embodiment 1:
As shown in Figure 1-2, a kind of bad data detection and identification method that is applicable to large electrical network of the invention of this example, described method is used in Power system state estimation process; Said method comprising the steps of:
(1) determine the suspicious data collection SUS of the whole network;
(2) determine the associated suspicious data collection sus[n that described suspicious data is concentrated];
(3) the associated suspicious data collection of identification sus[n] in bad data.
The deterministic process of described step (1) is:
(1-1) carry out the state estimation of the whole network.
(1-2) r
ndetection method is determined suspicious data collection SUS
Described step (1-1) is done the state estimation an of the whole network by original measurement.
Described r
nthe testing process of detection method is:
In formula, H
0be that i metric data is not suspicious metric data, H
1be that i metric data is suspicious metric data; r
n,ibe i standardized residual,
it is the threshold value of i standardized residual.
The threshold value of described standardized residual is determined by following steps:
Definition W and W
nthe sensitivity matrix that is respectively m * m rank residual sensitivity matrix and m * m rank standardized residual, has
Standardized residual r
nfor:
r
N=W
Nv (3)
Wherein, diagonal matrix D=diag[WR], the weight matrix that R is and measurement is corresponding, v is the m dimension error vector that contains bad data;
Therefore,, under normal measurement condition, there is standardized residual r
ncovariance matrix be:
Wherein, E represents expectation function, r
n,zfor the standardized residual of all measurements, therefore have
So when getting probability of false detection P
e=0.005, get the threshold value of described standardized residual
for
The deterministic process of described step (2) is:
(2-1) search of incidence matrix;
(2-2) finally form associated suspicious data collection.
Described step (2-1) is in Power system state estimation, and the associated suspicious data search procedure based on Jacobi matrix is:
Step (2-1-1): the number n that associated suspicious data collection is set is 1;
Step (2-1-2): newly-built n associated suspicious data collection sus[n];
Step (2-1-3): judge whether all suspicious data collection of the whole network SUS is empty, if be not idle running step (2-1-4), empty if, search procedure finishes;
Step (2-1-4): take out the data of standardized residual maximum from SUS, be assumed to be i data, these data are rejected from SUS, and join sus[n] in;
Step (2-1-5): search for the nonzero element in Jacobi matrix i row element, and record the row number of these nonzero elements, form set LOR;
Step (2-1-6): take out successively row number from LOR, and search for the row that Jacobi matrix is corresponding, in each row of record, the line number of non-zero element, forms set ROW;
Step (2-1-7): judge successively in ROW, whether data corresponding to line number belong to S set US, if there are the data that belong to S set US, all data that belong to S set US are joined to associated suspicious data collection sus[n] in, the data that these belonged to S set US are rejected simultaneously from SUS, go to step (2-1-8); If there is no the data that belong to S set US, n associated suspicious measurement collection sus[n] form completely, carry out n=n+1, and go to step (2-1-2);
Step (2-1-8): take out in step (2-1-7) and belong to line number corresponding to data of S set US, and search for the non-zero element in Jacobi matrix corresponding row element, record non-zero element column number, form LOR, go to step (2-1-6);
By above 8 steps, form n associated suspicious data collection of described step (2-2).
According to described associated suspicious data collection sus[n] calculate each sus[n] weighted residual quadratic sum J (x):
J(x)=[z-h(x)]
TR
-1[z-h(x)] (7)
Wherein, z is associated suspicious data, and R is corresponding diagonal angle weight matrix, and h (x) is the computing function of associated suspicious data.
Described double-deck bad data identification method process is:
Step (8-1): the suspicious measurement collection of current association to be processed sequence number k=1 is set;
Step (8-2): take out k associated suspicious data collection, calculate the weighted residual quadratic sum of all data in this set, be designated as J (x);
Step (8-3): take out k associated suspicious data and concentrate standardized residual maximum and not by the data sus[k of identification, i],, based on Givens orthogonal row converter technique, revise fast state estimation factor table and delete fast described suspicious data sus[k, i];
Step (8-4): associated suspicious data collection sus[k described in rapid solving] in all data described weighted residual quadratic sum J ' (x);
Step (8-5): described in identification, whether suspicious data is bad data;
If described suspicious data sus[k, i] be bad data, and go to step (8-6); If described suspicious data sus[k, i] not bad data, recover described suspicious data sus[k, i] to associated suspicious data collection sus[k] in, put identification mark, and by the quick modifying factor sublist of Givnes orthogonal row converter technique, then go to step (8-3);
Step (8-6): described in identification, associated suspicious data is concentrated and whether also had bad data;
If described associated suspicious data is concentrated, there is no bad data, continued next associated suspicious data set analysis, carried out k=k+1, and go to step (8-7), k is associated suspicious collection number; If described associated suspicious data is concentrated, also there is bad data, need to continue identification, go to step (8-3);
Step (8-7): judge whether k is greater than n, if so, bad data identification process finishes; If not, go to step (8-2).
Described step (8-5) judges whether to exist bad data by following formula:
| J (x)-J ' (x) | > ε
1, ε
1for threshold value; If inequality is set up, described data are bad data, if inequality is false, described data do not exist bad data;
Described step (8-6) judges the concentrated bad data that whether also exists of described associated suspicious data by following formula:
J ' is < ε (x)
2, ε
2for threshold value; If inequality is set up, described associated suspicious data is concentrated has not had bad data; If inequality is false, described associated suspicious data is concentrated and is also had bad data.
After completing detection and identification process, according to the described bad data picking out, the data respective weights of described Power system state estimation is revised, and carried out iteratively faster calculating, obtain described Power system state estimation data section accurately.
Adopt associated suspicious measurement Fast search technique, dwindle bad data identification scope, improved bad data identification speed and sensitivity; By the double-deck identification technique of heterogeneous pass bad data, greatly improved heterogeneous pass bad data identification accuracy and precision, finally obtain described Power system state estimation data section accurately.
Finally should be noted that: above embodiment is only in order to illustrate that technical scheme of the present invention is not intended to limit; although those of ordinary skill in the field are to be understood that with reference to above-described embodiment: still can modify or be equal to replacement the specific embodiment of the present invention; these do not depart from any modification of spirit and scope of the invention or are equal to replacement, within the claim protection domain of the present invention all awaiting the reply in application.
Claims (10)
1. be applicable to a bad data detection and identification method for large electrical network, described method is used in Power system state estimation process; It is characterized in that: said method comprising the steps of:
(1) determine the suspicious data collection SUS of the whole network;
(2) determine the associated suspicious data collection sus[n that described suspicious data is concentrated];
(3) the associated suspicious data collection of identification sus[n] in bad data.
2. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 1, is characterized in that: the deterministic process of described step (1) is:
(1-1) carry out the state estimation of the whole network;
(1-2) pass through r
ndetection method is determined suspicious data collection SUS;
Described r
nthe testing process of detection method is:
In formula, H
0be that i metric data is not suspicious metric data, H
1be that i metric data is suspicious metric data; r
n,ibe i standardized residual,
it is the threshold value of i standardized residual.
3. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 2, is characterized in that: the threshold value of described standardized residual is determined by following steps:
Definition W and W
nthe sensitivity matrix that is respectively m * m rank residual sensitivity matrix and m * m rank standardized residual, has
Standardized residual r
nfor:
r
N=W
Nv (3)
Wherein, diagonal matrix D=diag[WR], the weight matrix that R is and measurement is corresponding, v is the m dimension error vector that contains bad data;
Under normal measurement condition, there is standardized residual r
ncovariance matrix be:
Wherein, E represents expectation function, r
n,zfor the standardized residual of all measurements, therefore have
So when getting probability of false detection P
e=0.005, get the threshold value of described standardized residual
for
4. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 1, is characterized in that: the deterministic process of described step (2) is:
(2-1) search of incidence matrix;
(2-2) finally form associated suspicious data collection.
5. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 4, is characterized in that: described step (2-1) is in Power system state estimation, and the associated suspicious data search procedure based on Jacobi matrix is:
Step (2-1-1): the number n that associated suspicious data collection is set is 1;
Step (2-1-2): newly-built n associated suspicious data collection sus[n];
Step (2-1-3): judge whether all suspicious data collection of the whole network SUS is empty, if be not idle running step (2-1-4), empty if, search procedure finishes;
Step (2-1-4): take out the data of standardized residual maximum from SUS, be assumed to be i data, these data are rejected from SUS, and join sus[n] in;
Step (2-1-5): search for the nonzero element in Jacobi matrix i row element, and record the row number of these nonzero elements, form set LOR;
Step (2-1-6): take out successively row number from LOR, and search for the row that Jacobi matrix is corresponding, in each row of record, the line number of non-zero element, forms set ROW;
Step (2-1-7): judge successively in ROW, whether data corresponding to line number belong to S set US, if there are the data that belong to S set US, all data that belong to S set US are joined to associated suspicious data collection sus[n] in, the data that these belonged to S set US are rejected simultaneously from SUS, go to step (2-1-8); If there is no the data that belong to S set US, n associated suspicious measurement collection sus[n] form completely, carry out n=n+1, and go to step (2-1-2);
Step (2-1-8): take out in step (2-1-7) and belong to line number corresponding to data of S set US, and search for the non-zero element in Jacobi matrix corresponding row element, record non-zero element column number, form LOR, go to step (2-1-6);
By above 8 steps, form n associated suspicious data collection of described step (2-2).
6. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 5, is characterized in that: according to described associated suspicious data collection sus[n] calculate each sus[n] weighted residual quadratic sum J (x):
J(x)=[z-h(x)]
TR
-1[z-h(x)] (7)
Wherein, z is associated suspicious data, and R is corresponding diagonal angle weight matrix, and h (x) is the computing function of associated suspicious data.
7. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 6, is characterized in that: described step (3) is determined bad data by double-deck bad data identification method.
8. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 7, is characterized in that: described double-deck bad data identification method process is:
Step (8-1): the suspicious measurement collection of current association to be processed sequence number k=1 is set;
Step (8-2): take out k associated suspicious data collection, calculate the weighted residual quadratic sum of all data in this set, be designated as J (x);
Step (8-3): take out k associated suspicious data and concentrate standardized residual maximum and not by the data sus[k of identification, i],, based on Givens orthogonal row converter technique, revise fast state estimation factor table and delete fast described suspicious data sus[k, i];
Step (8-4): associated suspicious data collection sus[k described in rapid solving] in all data described weighted residual quadratic sum J ' (x);
Step (8-5): described in identification, whether suspicious data is bad data;
If described suspicious data sus[k, i] be bad data, and go to step (8-6); If described suspicious data sus[k, i] not bad data, recover described suspicious data sus[k, i] to associated suspicious data collection sus[k] in, put identification mark, and by the quick modifying factor sublist of Givnes orthogonal row converter technique, then go to step (8-3);
Step (8-6): described in identification, associated suspicious data is concentrated and whether also had bad data;
If described associated suspicious data is concentrated, there is no bad data, continued next associated suspicious data set analysis, carried out k=k+1, and go to step (8-7), k is associated suspicious data collection number; If described associated suspicious data is concentrated, also there is bad data, need to continue identification, go to step (8-3);
Step (8-7): judge whether k is greater than n, if so, bad data identification process finishes; If not, go to step (8-2).
9. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 8, is characterized in that: described step (8-5) judges whether to exist bad data by following formula:
| J (x)-J ' (x) | > ε
1, ε
1for threshold value; If inequality is set up, described data are bad data, if inequality is false, described data do not exist bad data;
Described step (8-6) judges the concentrated bad data that whether also exists of described associated suspicious data by following formula:
J ' is < ε (x)
2, ε
2for threshold value; If inequality is set up, described associated suspicious data is concentrated has not had bad data; If inequality is false, described associated suspicious data is concentrated and is also had bad data.
10. a kind of bad data detection and identification method that is applicable to large electrical network as claimed in claim 1, it is characterized in that: after completing detection and identification process, according to the described bad data picking out, the data respective weights of described Power system state estimation is revised, and carry out iteratively faster calculating, obtain described Power system state estimation data section accurately.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410407189.6A CN104166718B (en) | 2014-08-18 | 2014-08-18 | A kind of bad data detection and identification method suitable for bulk power grid |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410407189.6A CN104166718B (en) | 2014-08-18 | 2014-08-18 | A kind of bad data detection and identification method suitable for bulk power grid |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104166718A true CN104166718A (en) | 2014-11-26 |
CN104166718B CN104166718B (en) | 2017-11-03 |
Family
ID=51910531
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410407189.6A Active CN104166718B (en) | 2014-08-18 | 2014-08-18 | A kind of bad data detection and identification method suitable for bulk power grid |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104166718B (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104766175A (en) * | 2015-04-16 | 2015-07-08 | 东南大学 | Power system abnormal data identifying and correcting method based on time series analysis |
CN106295850A (en) * | 2016-07-15 | 2017-01-04 | 广东电网有限责任公司电力科学研究院 | A kind of bad data recognition method and system based on historical metrology |
CN106887842A (en) * | 2017-04-11 | 2017-06-23 | 广东电网有限责任公司电力调度控制中心 | A kind of bad data discrimination method based on XML/E format multiple-source electric network datas |
CN107133721A (en) * | 2017-04-17 | 2017-09-05 | 国网江苏省电力公司 | Power distribution network based on gray theory measures bad data relation factor analysis method |
CN107506824A (en) * | 2017-08-31 | 2017-12-22 | 广东工业大学 | Bad the observation data detection method and device of a kind of power distribution network |
CN108182257A (en) * | 2017-12-29 | 2018-06-19 | 东北电力大学 | A kind of GSA bad data detection and identification methods based on the optimization of areal concentration statistical method |
CN108333468A (en) * | 2018-01-05 | 2018-07-27 | 南京邮电大学 | The recognition methods of bad data and device under a kind of active power distribution network |
CN111931318A (en) * | 2020-07-07 | 2020-11-13 | 中国电力科学研究院有限公司 | Power supply path analysis method and system based on graph calculation |
CN112230087A (en) * | 2020-10-13 | 2021-01-15 | 全球能源互联网研究院有限公司 | Linear state estimation method and device, electronic equipment and storage medium |
CN114372235A (en) * | 2021-12-30 | 2022-04-19 | 东方电子股份有限公司 | Power system state estimation method for preventing singular transformation matrix |
CN114938030A (en) * | 2022-06-15 | 2022-08-23 | 华北电力大学 | Bad data processing method and device for electricity-heat comprehensive energy system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110191076A1 (en) * | 2008-10-09 | 2011-08-04 | Shunji Maeda | Error detection method and system |
CN103324847A (en) * | 2013-06-17 | 2013-09-25 | 西南交通大学 | Method for detecting and identifying dynamic bad data of electric power system |
CN103745109A (en) * | 2014-01-10 | 2014-04-23 | 国家电网公司 | Bad data detection and identification method based on measurement of PMU (Phasor Measurement Unit) and measurement of SCADA (Supervisory Control and Data Acquisition) |
-
2014
- 2014-08-18 CN CN201410407189.6A patent/CN104166718B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110191076A1 (en) * | 2008-10-09 | 2011-08-04 | Shunji Maeda | Error detection method and system |
CN103324847A (en) * | 2013-06-17 | 2013-09-25 | 西南交通大学 | Method for detecting and identifying dynamic bad data of electric power system |
CN103745109A (en) * | 2014-01-10 | 2014-04-23 | 国家电网公司 | Bad data detection and identification method based on measurement of PMU (Phasor Measurement Unit) and measurement of SCADA (Supervisory Control and Data Acquisition) |
Non-Patent Citations (2)
Title |
---|
周小宝: "电力系统状态估计不良数据检测与辨识方法研究与应用", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
张永超: "电力系统不良数据检测和辨识方法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104766175A (en) * | 2015-04-16 | 2015-07-08 | 东南大学 | Power system abnormal data identifying and correcting method based on time series analysis |
CN106295850A (en) * | 2016-07-15 | 2017-01-04 | 广东电网有限责任公司电力科学研究院 | A kind of bad data recognition method and system based on historical metrology |
CN106887842A (en) * | 2017-04-11 | 2017-06-23 | 广东电网有限责任公司电力调度控制中心 | A kind of bad data discrimination method based on XML/E format multiple-source electric network datas |
CN107133721A (en) * | 2017-04-17 | 2017-09-05 | 国网江苏省电力公司 | Power distribution network based on gray theory measures bad data relation factor analysis method |
CN107506824B (en) * | 2017-08-31 | 2021-01-26 | 广东工业大学 | Method and device for detecting bad observation data of power distribution network |
CN107506824A (en) * | 2017-08-31 | 2017-12-22 | 广东工业大学 | Bad the observation data detection method and device of a kind of power distribution network |
CN108182257A (en) * | 2017-12-29 | 2018-06-19 | 东北电力大学 | A kind of GSA bad data detection and identification methods based on the optimization of areal concentration statistical method |
CN108333468A (en) * | 2018-01-05 | 2018-07-27 | 南京邮电大学 | The recognition methods of bad data and device under a kind of active power distribution network |
CN111931318A (en) * | 2020-07-07 | 2020-11-13 | 中国电力科学研究院有限公司 | Power supply path analysis method and system based on graph calculation |
CN111931318B (en) * | 2020-07-07 | 2022-09-27 | 中国电力科学研究院有限公司 | Power supply path analysis method and system based on graph calculation |
CN112230087A (en) * | 2020-10-13 | 2021-01-15 | 全球能源互联网研究院有限公司 | Linear state estimation method and device, electronic equipment and storage medium |
CN112230087B (en) * | 2020-10-13 | 2022-08-05 | 全球能源互联网研究院有限公司 | Linear state estimation method and device, electronic equipment and storage medium |
CN114372235A (en) * | 2021-12-30 | 2022-04-19 | 东方电子股份有限公司 | Power system state estimation method for preventing singular transformation matrix |
CN114938030A (en) * | 2022-06-15 | 2022-08-23 | 华北电力大学 | Bad data processing method and device for electricity-heat comprehensive energy system |
CN114938030B (en) * | 2022-06-15 | 2024-08-06 | 华北电力大学 | Bad data processing method and device for electric-thermal comprehensive energy system |
Also Published As
Publication number | Publication date |
---|---|
CN104166718B (en) | 2017-11-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104166718A (en) | Bad data detection and recognition method suitable for large power grid | |
CN103324847B (en) | Electrical Power System Dynamic bad data detection and identification method | |
CN103326358B (en) | Electric power system dynamic state estimation method based on synchronous phase-angle measuring device | |
Weimer et al. | Distributed detection and isolation of topology attacks in power networks | |
CN104836223B (en) | Power grid parameter error and bad data coordinated identification and estimation method | |
Mirshekali et al. | Linear and nonlinear fault location in smart distribution network under line parameter uncertainty | |
CN103576053A (en) | Voltage sag source locating method based on limited electric energy quality monitoring points | |
CN109345408A (en) | Abnormal data discriminating method based on shot and long term memory network | |
CN103544542A (en) | Power system transient stability margin predicting method | |
Veerakumar et al. | PMU-based real-time distribution system state estimation considering anomaly detection, discrimination and identification | |
CN104504247B (en) | A kind of RAIM methods of the double satellite failures of GPS | |
CN105183938A (en) | Bad data identification and estimation method for power grid | |
Burr et al. | Revisiting statistical aspects of nuclear material accounting | |
Chen et al. | An improved recursive Bayesian approach for transformer tap position estimation | |
Weng et al. | Robust state-estimation procedure using a least trimmed squares pre-processor | |
CN113721182B (en) | Method and system for evaluating reliability of online performance monitoring result of power transformer | |
Kumar et al. | Deep Learning based Fault Detection in Power Transmission Lines | |
Mahato et al. | Electric power system transient stability assessment based on Bi-LSTM attention mechanism | |
CN108459991A (en) | A method of obtaining equipment dependability data | |
CN115758284B (en) | Photovoltaic hot spot fault detection method and system based on fusion kernel entropy and information gain | |
Chabane et al. | Sensor fault detection and diagnosis using zonotopic set-membership estimation | |
CN107204616A (en) | Power system probabilistic state estimation method based on the pseudo- spectrometry of adaptive sparse | |
US10488447B2 (en) | Method for state estimation of a distribution network based on real time measurement values | |
Rashed et al. | State estimation in the presence of cyber attacks using distributed partition technique | |
CN115795222A (en) | Grid bad parameter identification and correction method based on synchronous phasor measurement |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |