CN116148753A - Intelligent electric energy meter operation error monitoring system - Google Patents
Intelligent electric energy meter operation error monitoring system Download PDFInfo
- Publication number
- CN116148753A CN116148753A CN202310408981.2A CN202310408981A CN116148753A CN 116148753 A CN116148753 A CN 116148753A CN 202310408981 A CN202310408981 A CN 202310408981A CN 116148753 A CN116148753 A CN 116148753A
- Authority
- CN
- China
- Prior art keywords
- ammeter
- value
- data
- user
- singular value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 claims abstract description 90
- 230000005611 electricity Effects 0.000 claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000005259 measurement Methods 0.000 claims abstract description 22
- 238000004891 communication Methods 0.000 claims abstract description 12
- 238000001514 detection method Methods 0.000 claims description 32
- 238000004364 calculation method Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 20
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 238000012937 correction Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000007689 inspection Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004171 remote diagnosis Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
- G01R35/04—Testing or calibrating of apparatus covered by the other groups of this subclass of instruments for measuring time integral of power or current
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention provides an intelligent ammeter operation error monitoring system, and belongs to the field of intelligent ammeter error monitoring base numbers. The system comprises: the system comprises a user ammeter, a district total ammeter and a monitoring platform for carrying out data communication with the user ammeter and the district total ammeter through the Internet; and the monitoring platform reads the metering values of the user ammeter and the total ammeter of the district in real time, performs singular value identification on the metering values of the user ammeter, and calculates the metering error of the user ammeter based on the metering values of the total ammeter of the district and the metering values of the user ammeter after singular value processing. The method solves the problems of low error estimation precision, low operation speed, poor real-time performance and the like in the prior art. The invention can eliminate the influence of the singular value on the error estimation precision by identifying the singular value appearing in the meter measurement value caused by the actions such as user electricity stealing and the like and performing singular value processing.
Description
Technical Field
The invention belongs to the technical field of intelligent ammeter error monitoring, and particularly relates to an intelligent ammeter operation error monitoring system.
Background
At present, the traditional manual meter reading mode is replaced by automatic acquisition, so that the work load of manual on-site meter reading is greatly reduced, and the work of on-site inspection of the operation working conditions of the metering device of the electricity consumers, particularly the residents in the low-voltage transformer area, is also greatly reduced. The intelligent ammeter is used as a metering tool for consuming electric energy by a user, and the operation reliability of the intelligent ammeter not only influences the operation income of a power grid company, but also is directly related to the actual benefits of thousands of households. In order to strengthen dynamic management of a transformer area and improve the service level of a power grid, it is imperative to search for a high-efficiency and accurate remote diagnosis method for the operation error of an intelligent electric energy meter. The state monitoring of the traditional intelligent electric energy meter is realized by adding on-line monitoring equipment, and the on-line detection of the voltage, current, power, load, electric quantity and other data of the metering device and the secondary circuit at the earliest, so as to realize on-line error monitoring and alarming of a monitoring target. Although the method improves the management work efficiency, the method brings about the increase of equipment purchasing, operation and maintenance costs. Under the background, the large data analysis technology can be utilized, based on the data such as the power consumption of the users in the transformer area, the total power consumption of the transformer area, the network loss, the user files, the household meter relation and the like in the power consumption information acquisition system, a plurality of dimensions such as the condition, the household transformer relation, the power consumption, the network loss and the like are acquired from the transformer area, the statistics rule of the total power consumption and the electric quantity of each sub-meter under the same transformer area is researched, an intelligent meter operation error calculation model and an analysis model are established, the operation errors of each intelligent meter under the transformer area are calculated, all intelligent meter operation health conditions under the transformer area are obtained, the remote diagnosis and evaluation of the intelligent meter operation errors are realized, the effective technical means is provided for the power consumption inspection work, the suspected metering points with large workload and lacking pertinence are timely found, the on-site inspection high-efficiency and the on-site inspection work of the intelligent meter is realized, and the on-site inspection work is pertinently developed.
Currently, electric workers have achieved some results in the research of remote error estimation of smart meters. Most of the methods are based on a generalized flow conservation model, and a high-dimensional linear equation set is solved by adopting a trigonometric decomposition method, a least square method and the like. However, this type of method has the following problems: the data acquisition quality of the current low-voltage distribution transformer area is low, and phenomena of time mark dislocation, false alarm, bad value and the like exist, so that the intelligent ammeter has low error estimation precision and low operation speed and poor instantaneity; in particular, the existing method does not consider the influence of singular values (such as the electricity consumption value under the abnormal conditions such as user power stealing) on the error estimation precision.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides an intelligent electric energy meter operation error monitoring system.
In order to achieve the above object, the present invention adopts the following technical scheme.
An intelligent ammeter operation error monitoring system, comprising: the system comprises a user ammeter, a district total ammeter and a monitoring platform for carrying out data communication with the user ammeter and the district total ammeter through the Internet; the monitoring platform reads metering values of the user ammeter and the total ammeter of the station area in real time, performs singular value identification on the metering values of the user ammeter, and calculates metering errors of the user ammeter based on the metering values of the total ammeter of the station area and the metering values of the user ammeter after singular value processing;
the method for carrying out singular value identification on the metering value of the user ammeter comprises the following steps:
based onNSingular value detection is carried out on the metering value by different methods to obtainNA set of preliminary singular value data,N≥2；
counting the number of each metering value belonging to the preliminary singular value data setM；
Will measure the value ofMThe value is compared with a set threshold value,Mthe measured value whose value exceeds the set threshold is a singular value.
Further, the data obtained by the monitoring platform from the user electric meter and the total electric meter of the district further comprises file data of the user electric meter and the total electric meter of the district.
Further, the method for calculating the metering error of the user ammeter comprises the following steps:
building an ideal power consumption model of the station area:
total electric meter metering value of station area = actual electricity consumption of user + line loss of station area + fixed loss
Establishing a power consumption approximation model of a platform region in a data acquisition period:
in the method, in the process of the invention,y、z、crespectively the measurement value, the line loss and the fixed loss of the total ammeter of the station area in one data acquisition period,
、/>
respectively the firstiMetering value sum of individual user ammeteriThe actual power usage of the individual user, wherein,i=1,2,…,n，nthe number of the user electricity meters for the transformer area; />
、b、cFor the value to be fixed, the solving method is as follows:
list bynEquation set composed of electric approximation models of the station area in +2 data acquisition period:
in the method, in the process of the invention,
、/>
、/>
respectively the firstjWithin the data acquisition periodiMetering value of individual user ammeter and meter of total ammeter of station areaThe magnitude and the fixed loss of the material,j=1,2,…,n+2；
writing the set of equations into a matrix equation:AX=Ywherein, the method comprises the steps of, wherein,
solving the matrix equation to obtain:
or->
Thereby obtaining->
、b、c；
According to the first data acquisition periodiMetering value calculation of individual user ammeteriActual power consumption of individual users:
；
the first data acquisition periodiThe relative errors of the individual consumer electricity meters are:
in the method, in the process of the invention,
is the firstiThe relative error of the individual consumer's electricity meters,i=1,2,…,n。
further, the line loss of the station area in one data acquisition period is as follows:z=by。
further, the method for processing the singular values in error calculation includes:
acquiring historical data of a singular value user ammeter;
calculating an average value of the measurement values in one data acquisition period based on the historical data;
and replacing the corresponding singular value with the average value.
Further, the method for processing the singular values in error calculation includes:
calculating the average value of all non-singular value user electricity meter measurement values in a data acquisition period;
and replacing the corresponding singular value with the average value.
Further, the method for processing the singular values in error calculation includes:
calculating the average value of all non-singular value user electricity meter measurement values in one data acquisition periodx - ；
Metering value of total ammeter in station areayThe correction is as follows:
，his the number of singular values;
and carrying out error calculation based on the corrected metering value of the total ammeter of the district and the metering value of the non-singular value user ammeter.
Still further, theNThe method comprises the steps that (1) 4,4 singular value detection methods are respectively a singular value detection method based on a K-means algorithm, a local anomaly factor algorithm, an isolated forest algorithm and an SVM algorithm, and preliminary singular value data sets D1, D2, D3 and D4 are respectively obtained based on the 4 singular value detection methods; if a data belongs to any two or more of the preliminary singular value data sets D1, D2, D3 and D4, the data is singular values.
Further, the singular value detection method based on the K-means algorithm comprises the following steps:
s11, setting the number of cluster clustersk；
S12, slavemOptionally in the datakThe data is used as an initial clustering center
；
S13, calculating each data and the aggregationDistance between class centers
And respectively classifying each data into the class of the nearest cluster center to obtainkClusters of (a) wherein->
Represent the firstiI=1, 2, … m,
represent the firstjThe cluster centers, j= =1, 2, … k;
s14, taking the average value of the data of each cluster as a new cluster center, repeating S13 to obtain a new cluster centerkClustering clusters, ifkThe cluster clusters are not changed any more, and S15 is switched; otherwise, turning to S13;
s15, if the objective function converges or meets the termination condition, obtainingkClustering clusters; otherwise, turning to S12;
s16, countingkThe data in the clusters with the least data form a preliminary singular value data set D1.
Compared with the prior art, the invention has the following beneficial effects.
According to the invention, the monitoring platform for carrying out data communication with the user electric meter and the district total electric meter through the Internet is arranged, the monitoring platform reads the metering values of the user electric meter and the district total electric meter in real time, carries out singular value identification on the metering values of the user electric meter, calculates the metering error of the user electric meter based on the district total electric meter metering values and the user electric meter metering values after singular value processing, and realizes automatic monitoring of the operation error of the intelligent electric meter. The method solves the problems of low error estimation precision, low operation speed, poor real-time performance and the like in the prior art. The invention can eliminate the influence of the singular value on the error estimation precision by identifying the singular value appearing in the meter measurement value caused by the actions such as user electricity stealing and the like and performing singular value processing.
Drawings
Fig. 1 is a block diagram of an intelligent ammeter operation error monitoring system according to an embodiment of the present invention. In the figure: 1-a total ammeter of a platform area, 2-a user ammeter and 3-a monitoring platform.
Fig. 2 is a schematic distribution diagram of a smart meter in a station area.
Detailed Description
The present invention will be further described with reference to the drawings and the detailed description below, in order to make the objects, technical solutions and advantages of the present invention more apparent. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Fig. 1 is a block diagram of an intelligent ammeter operation error monitoring system according to an embodiment of the present invention, including: the monitoring platform 3 is used for carrying out data communication with the user ammeter 2 and the district total ammeter 1 through the Internet; the monitoring platform 3 reads metering values of the user ammeter 2 and the district total ammeter 1 in real time, performs singular value identification on the metering values of the user ammeter 2, and calculates metering errors of the user ammeter 2 based on the district total ammeter 1 metering values and the user ammeter 2 metering values after singular value processing;
the method for carrying out singular value identification on the metering value of the user ammeter 2 comprises the following steps:
based onNSingular value detection is carried out on the metering value by different methods to obtainNA set of preliminary singular value data,N≥2；
counting the number of each metering value belonging to the preliminary singular value data setM；
Will measure the value ofMThe value is compared with a set threshold value,Mthe measured value whose value exceeds the set threshold is a singular value.
In this embodiment, error measurement is mainly performed on the user smart meter of each transformer station, and fig. 2 is a schematic diagram of the topological relation of the typical power distribution station smart meter. In fig. 2, a total electric meter 1 of a district is installed under each district distribution transformer, and a total plurality of user electric meters 2 are connected. The reading of the smart meter can be automatically obtained through a smart meter remote meter reading system (equivalent to the monitoring platform 3 in fig. 1).
The system in this embodiment mainly comprises a monitoring platform 3, a total electric meter 1 in a platform area and a user electric meter 2. The monitoring platform 3 is in data communication with the total electric meter 1 and the user electric meter 2 through the internet. Each of the sections is described separately below.
The consumer electricity meter 2 is installed at the home or workplace of each consumer and is used for measuring the electricity consumption (in degrees or kilowatt-hours) of the consumer. The current user electric meter 2 (including the total electric meter 1 of the district) adopts intelligent electric meters. The intelligent ammeter is a kilowatt-hour meter which uses an intelligent chip (such as a CPU) as a core and has the functions of electric power metering, time counting, fee counting, communication with an upper computer, electricity management and the like by using a more computer technology, a communication technology and the like. The user electricity meter 2 of the present embodiment performs data communication with the monitoring platform 3 (upper computer) through the internet by using the built-in communication module.
The total ammeter 1 of the transformer area is arranged at the output end of the distribution transformer of the transformer area and is used for measuring the total electricity consumption of the transformer area. Theoretically, the metering value of the total ammeter 1 in the transformer area is equal to the sum of metering values of all the user ammeter 2 in the transformer area; however, in practice, the power grid structure is complex, and there are also transmission line loss (line loss for short) and fixed loss of the electric meter in the power grid, and there is a certain measurement error between the user electric meter 2 and the total electric meter 1 in the area, and in general, the measurement value of the total electric meter 1 in the area is obviously greater than the sum of the measurement values of the user electric meters 2. The total electric meter 1 in the area is usually checked regularly and can be regarded as a standard electric meter with no error in reading. The error calculation of the present embodiment is also performed on the assumption that the total electric meter 1 of the cell is error-free. Likewise, the total electric meter 1 of the platform area also adopts a smart electric meter, and utilizes an internal communication module to carry out data communication with the monitoring platform 3 (upper computer) through the internet.
The monitoring platform 3 is a data processing and monitoring center of the system and is used for monitoring the user ammeter 2 and the total ammeter 1 in the transformer area. The monitoring platform 3 is generally implemented by a smart meter remote meter reading system, acquires metering data of the user meter 2 and the total meter 1 of the station area in real time through the internet, and calculates a measurement error of the user meter 2 based on the metering data. Unlike the prior art, the present embodiment first performs singular value recognition on the metering value of the user ammeter 2. The singular value is data which is greatly different from a normal value or which violates a natural law, such as data which is far smaller than an average value (power is not used for a long time), power consumption is negative (power theft), and the like. When singular values are present, they will have a significant impact on the results of the data processing if not processed. For the present embodiment, when the singular value exists in the measured value of the user electric meter 2, not only the error calculation of the singular value user electric meter 2 is wrong, but also the error calculation deviation of the non-singular value user electric meter 2 is obviously increased. Therefore, in the embodiment, the singular value identification is performed on the metering value of the user ammeter 2, the identified singular value is processed, and the error calculation is performed based on the data processed by the singular value, so that the influence of the singular value can be greatly reduced, and the error calculation precision is improved. It should be noted that, the result of performing the error calculation of the user ammeter 2 based on the data after the singular value processing in this embodiment is only suitable for the non-singular value user ammeter 2, and is not suitable for the singular value user ammeter 2. For the singular value user ammeter 2, the cause of the occurrence of the singular value should be found out by combining manual field investigation, and related measures are taken to eliminate the singular value, and then error calculation can be performed.
The embodiment also provides a technical scheme of singular value recognition. In order to improve the accuracy of singular value recognition, the embodiment is based on a plurality of kinds ofNSingular value detection of data is performed by different methods. Each method obtains a preliminary singular value data set, and obtainsNA set of sets. And then further counting the number of the data belonging to the preliminary singular value data setM，M≤N。MThe data of =0 must not be singular values, and in order to reduce the statistical effort, only the data corresponding to the data in the preliminary singular value data set may be countedM. Obviously, whenNWhen the fixing device is used for fixing the fixing device,Mthe larger the value the more likely the data is singular values; it is precisely the case that the number of the devices,M/Nthe larger the data the more likely it is to be singular values. This example is by comparisonMAnd the size of the set threshold value is used for judging whether the data to be detected is a singular value, ifMIf the set threshold is exceeded, the data to be detected is considered to beSingular values; otherwise not singular values. The set threshold is determined empirically andNis related to the size of (a). Of course, can also be according toM/NAnd (3) whether the size of the code exceeds a set threshold value or not, and carrying out singular value recognition.
As an alternative embodiment, the data obtained by the monitoring platform 3 from the user electric meter 2 and the total electric meter 1 of the district also includes the file data of the user electric meter 2 and the total electric meter 1 of the district.
The present embodiment shows the data acquired by the monitoring platform 3 from the user electricity meter 2 and the total station electricity meter 1. In addition to the electric energy measurement value data of the user electric meter 2 and the district total electric meter 1, the data obtained by the monitoring platform 3 from the user electric meter 2 and the district total electric meter 1 also includes the archive data of the user electric meter 2 and the district total electric meter 1, such as electric meter numbers, parameters, and the like.
As an alternative embodiment, the method for calculating the metering error of the user ammeter 2 includes:
building an ideal power consumption model of the station area:
total electric meter 1 measurement value of the transformer area = actual electricity consumption of a user + line loss of the transformer area + fixed loss
Establishing a power consumption approximation model of a platform region in a data acquisition period:
in the method, in the process of the invention,y、z、crespectively the measurement value, the line loss and the fixed loss of the total ammeter 1 of the transformer area in one data acquisition period,
、/>
respectively the firstiMetering value sum of individual user's ammeter 2iThe actual power usage of the individual user, wherein,i=1,2,…,n，nthe number of user electricity meters 2 for the transformer area; />
、b、cFor the value to be fixed, the solving method is as follows: />
List bynEquation set composed of electric approximation models of the station area in +2 data acquisition period:
in the method, in the process of the invention,
、/>
、/>
respectively the firstjWithin the data acquisition periodiThe metering value of the individual consumer electricity meters 2, the metering value of the total electricity meters 1 of the district and the fixed loss,j=1,2,…,n+2；
writing the set of equations into a matrix equation:AX=Ywherein, the method comprises the steps of, wherein,
solving the matrix equation to obtain:
or->
Thereby obtaining->
、b、c；
According to the first data acquisition periodiCalculation of the metering value of the individual consumer's ammeter 2iActual power consumption of individual users:
；
data acquisitionIn the period ofiThe relative errors of the individual consumer meters 2 are:
in the method, in the process of the invention,
is the firstiThe relative error of the individual consumer meters 2,i=1,2,…,n。
the embodiment provides a technical scheme for calculating the metering error of the user ammeter 2. Firstly, building an ideal power consumption model of a platform region according to the topological relation of a intelligent ammeter of the platform region: total electric meter 1 measurement value of the station area = actual electricity consumption of the user + station area line loss + fixed loss. And then performing approximation processing on the ideal model, and establishing an electricity approximation model of the transformer area in a data acquisition period. The approximation process includes: firstly, the actual power consumption of each user
Metering value with the consumer electricity meter 2>
Ratio of (2)
//>
Regarded as constant->
The method comprises the steps of carrying out a first treatment on the surface of the Secondly, the line loss of the transformer areazMetering value of total ammeter 1 in station areayRatio of (2)z/yConsidered as a constantbThe method comprises the steps of carrying out a first treatment on the surface of the Thirdly, regarding the fixed loss of the station area as a constantc。/>
、b、cTo be fixed, due toi=1,2,…,nCo-mingling withn+2 waiting values, thus requiring construction to includenA +2 approximation model equation set is solved, and the equation set taking the waiting value as an unknown variable isThe waiting value can be obtained. The embodiment is by listingnThe power consumption approximation model of the station area with +2 data acquisition period is obtained bynA system of +2 equations and writing the system of equations in the form of a matrix equationAX=YWhereinXTo be evaluated +.>
、b、cComposition%n+2) x 1 order matrix,A、Yrespectively bynThe metering value of the user ammeter 2 and the metering value of the total ammeter 1 in the station area in +2 data acquisition periodn+2)×(nSum of +2) ordern+2) x 1 order matrix. The solving method of the matrix equation is as follows: if it isAReversible, then
The method comprises the steps of carrying out a first treatment on the surface of the If it isAIrreversible, then get approximate solution->
. Has->
Can be based on the first data acquisition periodiMetering value of individual consumer electricity meter 2x i Calculate the firstiActual power consumption of individual users: />
. Further get the firstiRelative error of the individual consumer meters 2: />
Can also be written as +.>
。
As an alternative embodiment, the line loss of the station area in one data acquisition period is:z=by。
the embodiment provides a technical scheme for calculating the line loss of the station area. The previous embodiment has obtained the ratio of the line loss of the district to the measured value of the district total ammeter 1bOn the basis of which a single can be easily obtainedThe line loss of the station area in the data acquisition period is as follows:z=by。
as an alternative embodiment, the method for processing singular values in performing error calculation includes:
acquiring historical data of a singular value user ammeter 2;
calculating an average value of the measurement values in one data acquisition period based on the historical data;
and replacing the corresponding singular value with the average value.
The embodiment provides a technical scheme of singular value processing. The present embodiment processes the singular values based on the history data of the identified singular value user meter 2 measurement values, provided of course that the history data is not singular values. Specifically, an average value of the measured values in one data acquisition period is calculated based on the historical data, and then the corresponding singular value is replaced by the average value.
As an alternative embodiment, the method for processing singular values in performing error calculation includes:
calculating the average value of all non-singular value user ammeter 2 metering values in a data acquisition period;
and replacing the corresponding singular value with the average value.
The embodiment provides another technical scheme of singular value processing. The singular value processing method of the previous embodiment is only possible if the history of the measured values of the singular value consumer electric meter 2 is not a singular value, and is not possible if the history is also a singular value. For this reason, the present embodiment does not use the history data any more, but solves for the average value of the measured values of the other non-singular value user meters 2, and replaces the corresponding singular value with the average value.
As an alternative embodiment, the method for processing singular values in performing error calculation includes:
calculating the average value of all non-singular value user ammeter 2 measurement values in one data acquisition periodx - ；
Metering value of total ammeter 1 in station areayThe correction is as follows:
，his the number of singular values;
and carrying out error calculation based on the corrected metering value of the total electric meter 1 of the district and the metering value of the non-singular value user electric meter 2.
The present embodiment provides yet another technical solution for singular value processing. Unlike the previous two embodiments, which replace singular values, the present embodiment eliminates the measured value data of the singular value user meter 2, which is equivalent to the reduction of the user meters 2 participating in modeling. However, since the measured value of the total electric meter 1 still includes the actual power consumption of the singular value user, the measured value of the total electric meter 1 needs to be corrected, that is, the actual power consumption of the singular value user is subtracted from the measured value of the total electric meter 1. Since the actual power consumption of the singular value user is unknown, an approximation process is required. In this embodiment, the average value of the measured values of all the non-singular value user electricity meters 2 is regarded as the actual electricity consumption of each singular value user, and the measured value of the total electricity meter 1 in the station area is corrected.
As an alternative embodiment, theNThe method comprises the steps that (1) 4,4 singular value detection methods are respectively a singular value detection method based on a K-means algorithm, a local anomaly factor algorithm, an isolated forest algorithm and an SVM algorithm, and preliminary singular value data sets D1, D2, D3 and D4 are respectively obtained based on the 4 singular value detection methods; if a data belongs to any two or more of the preliminary singular value data sets D1, D2, D3 and D4, the data is singular values.
The present embodiment defines a plurality of singular value detection methods. In this embodiment, n=4, that is, 4 singular value detection methods are adopted, which are respectively a singular value detection method based on a K-means algorithm, a local anomaly factor algorithm, an isolated forest algorithm and an SVM algorithm. The 4 algorithms are all well established prior art, but the 4 algorithms are not dedicated to singular value detection. The present embodiment is slightly improved on the basis of the original algorithm, and uses them as singular value detection, resulting in 4 preliminary singular value data sets D1, D2, D3 and D4. The threshold is then set to 2, i.e. a data is singular if it belongs to any two or more of the preliminary sets of singular value data D1, D2, D3 and D4.
As an alternative embodiment, the singular value detection method based on the K-means algorithm includes:
s11, setting the number of cluster clustersk；
S12, slavemOptionally in the datakThe data is used as an initial clustering center
；
S13, calculating the distance between each data and the clustering center
And respectively classifying each data into the class of the nearest cluster center to obtainkClusters of (a) wherein->
Represent the firstiData, i=1, 2, … m,>
represent the firstjThe cluster centers, j= =1, 2, … k;
s14, taking the average value of the data of each cluster as a new cluster center, repeating S13 to obtain a new cluster centerkClustering clusters, ifkThe cluster clusters are not changed any more, and S15 is switched; otherwise, turning to S13;
s15, if the objective function converges or meets the termination condition, obtainingkClustering clusters; otherwise, turning to S12;
s16, countingkThe data in the clusters with the least data form a preliminary singular value data set D1.
The embodiment provides a singular value detection method based on a K-means algorithm. The K-means clustering algorithm is the most basic and widely used clustering analysis method at present, and is a typical function clustering method taking the distance between a data point and the center of a cluster as an optimization target. According to the core thought of the clustering algorithm, the clustering algorithm should make the similarity among clusters as far as possibleIt is possible to maximize the principle that the similarity between clusters is minimized. The K-means algorithm selects the desired cluster center K, and the cluster center is continuously iterated and recalculated to minimize the variance in the whole cluster, so that clusters which are relatively compact and mutually independent are taken as the final target of the algorithm. And obtaining an extremum by using a function method, and adjusting the iteration frequency threshold value to obtain the optimal clustering effect. The present example giveskAfter clustering, respectively countingkAnd regarding the data in the clusters with the least data number as outliers, and forming the outliers into a preliminary singular value data set D1.
As an alternative embodiment, the singular value detection method based on the local anomaly factor algorithm includes:
s21, calculating any pointpIs the first of (2)kFirst of the points in the neighborhoodkThe distance can be reached:
in the method, in the process of the invention,
is a neighborhood pointoIs the first of (2)kDistance d (o, p) is the neighborhood pointoTo the pointpIs a distance of (2);
s22, calculating any pointpIs the local first of (2)kLocal reachable density:
in the method, in the process of the invention,
is thatpPoint NokDistance neighborhood->
The number of data points in the memory;
s23, calculating any pointpIs the first of (2)kLocal outlier factor:
s24, selecting data points corresponding to the largest local outliers to form a preliminary singular value data set D2.
The embodiment provides a singular value detection method based on a local anomaly factor algorithm. The local anomaly factor algorithm (Local Outlier Factor, LOF) belongs to one of the proximity-based approaches. Proximity-based methods are roughly classified into three categories, respectively: clustering methods such as KNN; a distance-based method; a density-based method; LOF belongs to one of the third class of density-based methods. The local anomaly factor algorithm reflects the degree of anomaly of a sample by calculating a "local reachable density", the greater the local reachable density of a sample point, the more likely that this point is an anomaly. In this embodiment, after local outliers of each data point are obtained by LOF, the local outliers are maximizednAnd data points corresponding to the local outliers form a preliminary singular value data set D2.
As an alternative embodiment, the singular value detection method based on the isolated forest algorithm includes:
s31, constructing an isolated tree;
s311, slave data set x=
Random decimation inΨThe sample points are used as root nodes of the tree;
s312, randomly selecting attributesqAttributes ofqThe data in (a) is recorded as
In the attribute ofqRandomly selecting a value between the maximum and minimum values of (2)pAs a cutting point;
s313 topDividing nodes when
<pWhen data is->
Is put at the place whereLeft branch of front node; when (when)
≥pWhen data is->
Put on the right branch;
s314, repeating S312 and S313 in the left and right branches, and continuously constructing new child nodes until the recursion ends when any one of the following conditions is met:
condition 1: the data set X only comprises one piece of data or all the data are the same;
condition 2: has reached the maximum height of the tree
；
S32, calculating the height of the sample in each tree;
by dataxTraversing an isolated tree to calculate data pointsxThe number of layers, i.e. the height, from the root node to the leaf node finally falls on the treeh(x) The calculation formula is as follows:
h(x)=e+C(n)
in the method, in the process of the invention,eis data ofxThe number of edges passing from the root node to the leaf node of the orphan tree,C(n) Is a correction value, expressed in a single plantnThe average path length of the binary tree constructed by the sample data is calculated as follows:
in the method, in the process of the invention,ξis Euler constant, and the value is 0.5772156649;
s33, calculating an outlier score, wherein the calculation formula is as follows:
in the method, in the process of the invention,E(h(x) Is) ish(x) Is used as a reference to the desired value of (a),s(x,n) Is data ofxIs used to determine the outlier score of (c),s(x,n) The value range of (1) is (0, 1)]；
S34, wills(x,n) The data greater than the set threshold constitute a preliminary singular value data set D3.
The embodiment provides a singular value detection method based on an isolated forest algorithm. The isolated forest algorithm iForest is an unsupervised learning algorithm, is generally used for detecting the abnormality of the structured data, and identifies the abnormality through the outlier in the isolated data. In general, the anomaly detection firstly determines what is normal content and abnormal content, and then obtains a detection result to judge. However, unlike this principle, the iferst algorithm does not first define normal and abnormal boundaries, nor does it calculate a point-based distance. The iferst is a decision tree-based algorithm, and is composed of a plurality of isolated binary trees ifere, and for each ifere, a segmentation value is randomly selected from a feature set extracted from data, then a maximum value and a minimum value under the feature in the data set are randomly selected, the data are segmented through the segmentation value, so that left and right subtrees are constructed, until the data cannot be segmented or the tree height limit is reached, and the segmentation mode can enable abnormal data points to be closer to a root node in the ifere, namely the path from a leaf node where the data points are located to the root node is shorter, so that the abnormal data points and normal data points are separated. The present embodiment calculates the outlier score for each data using ifersts(x, n) Will bes(x,n) The data greater than the set threshold constitute a preliminary singular value data set D3.
As an alternative embodiment, the singular value detection method based on the SVM algorithm includes:
s41, setting a data set
The weight vector of the largest interval classification hyperplane for the two classifications is +.>
The deviation vector isbThe method comprises the steps of carrying out a first treatment on the surface of the The classification mechanism is as follows: when wx + b > 1,y=1; when wx + b < -1,y=-1；
s42, solving w which enables the minimum value of the objective function by adopting Lagrange method,b：
In the method, in the process of the invention,nfor the number of samples, ζ is the relaxation term,Cin order to penalize the parameters,
alpha is Lagrange multiplier;
s43, will satisfy
Is->
A preliminary set of singular value data D4 is composed,
。
the embodiment provides a singular value detection method based on an SVM algorithm. The core content of the support vector machine SVM is proposed in 1992 to 1995, and is based on the structural risk minimization principle of the statistical theory and the VC dimension theory, and an optimal model is sought between the complexity of the model and the learning capacity of the model according to limited samples, so that the model is optimally popularized, namely generalization. Support vector machines are good at solving the problem of data space linearity inseparability, mainly by relaxation of variable and kernel function techniques. For the classification of data, it is desirable to find the best classifying plane, which can separate all classes of data, i.e. the plane with the largest data point interval belonging to different classes, called the maximum interval hyperplane, this classifier is called the maximum interval classifier. The classification interval is the largest, namely the confidence range in the popularization is the smallest, and the structural risk minimization principle is satisfied. In the embodiment, the maximum interval hyperplane equation y=wx+b is obtained by using an SVM algorithm, and then the gas data to be detected are obtained
Substitution into the equation will satisfy +.>
Is->
A preliminary singular value data set D4 is composed.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any changes or substitutions easily contemplated by those skilled in the art within the scope of the present invention should be included in the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (10)
1. An intelligent ammeter operation error monitoring system, characterized by comprising: the system comprises a user ammeter, a district total ammeter and a monitoring platform for carrying out data communication with the user ammeter and the district total ammeter through the Internet; the monitoring platform reads metering values of the user ammeter and the total ammeter of the station area in real time, performs singular value identification on the metering values of the user ammeter, and calculates metering errors of the user ammeter based on the metering values of the total ammeter of the station area and the metering values of the user ammeter after singular value processing;
the method for carrying out singular value identification on the metering value of the user ammeter comprises the following steps:
based onNSingular value detection is carried out on the metering value by different methods to obtainNA set of preliminary singular value data,N≥2；
counting the number of each metering value belonging to the preliminary singular value data setM；
Will measure the value ofMThe value is compared with a set threshold value,Mthe measured value whose value exceeds the set threshold is a singular value.
2. The intelligent ammeter operation error monitoring system of claim 1 wherein the data obtained by the monitoring platform from the utility meter and the total power meter area further comprises archival data of the utility meter and the total power meter area.
3. The intelligent ammeter operation error monitoring system of claim 1, wherein the method of calculating the ammeter metering error comprises:
building an ideal power consumption model of the station area:
total electric meter metering value of station area = actual electricity consumption of user + line loss of station area + fixed loss
Establishing a power consumption approximation model of a platform region in a data acquisition period:
in the method, in the process of the invention,y、z、crespectively the measurement value, the line loss and the fixed loss of the total ammeter of the station area in one data acquisition period,
、
respectively the firstiMetering value sum of individual user ammeteriThe actual power usage of the individual user, wherein,i=1,2,…,n，nthe number of the user electricity meters for the transformer area; />
、b、cFor the value to be fixed, the solving method is as follows:
list bynEquation set composed of electric approximation models of the station area in +2 data acquisition period:
in the method, in the process of the invention,
、/>
、/>
respectively the firstjWithin the data acquisition periodiThe metering value of the individual user electricity meters, the metering value of the total electricity meters of the station area and the fixed loss,j=1,2,…,n+2；
writing the set of equations into a matrix equation:AX=Ywherein, the method comprises the steps of, wherein,
solving the matrix equation to obtain:
or->
Thereby obtaining->
、b、c；
According to the first data acquisition periodiMetering value calculation of individual user ammeteriActual power consumption of individual users:
；
the first data acquisition periodiThe relative errors of the individual consumer electricity meters are:
in the method, in the process of the invention,
is the firstiThe relative error of the individual consumer's electricity meters,i=1,2,…,n。
4. the intelligent ammeter operation error monitoring system according to claim 3 wherein the line loss of the station area in one data acquisition period is:z=by。
5. the intelligent ammeter operation error monitoring system according to claim 3 wherein the method of processing singular values in performing the error calculation comprises:
acquiring historical data of a singular value user ammeter;
calculating an average value of the measurement values in one data acquisition period based on the historical data;
and replacing the corresponding singular value with the average value.
6. The intelligent ammeter operation error monitoring system according to claim 3 wherein the method of processing singular values in performing the error calculation comprises:
calculating the average value of all non-singular value user electricity meter measurement values in a data acquisition period;
and replacing the corresponding singular value with the average value.
7. The intelligent ammeter operation error monitoring system according to claim 3 wherein the method of processing singular values in performing the error calculation comprises:
calculating the average value of all non-singular value user electricity meter measurement values in one data acquisition periodx - ；
Metering value of total ammeter in station areayThe correction is as follows:y-hx - ，his the number of singular values;
and carrying out error calculation based on the corrected metering value of the total ammeter of the district and the metering value of the non-singular value user ammeter.
8. The intelligent ammeter operation error monitoring system of claim 1 wherein theNThe method comprises the steps that (1) 4,4 singular value detection methods are respectively a singular value detection method based on a K-means algorithm, a local anomaly factor algorithm, an isolated forest algorithm and an SVM algorithm, and preliminary singular value data sets D1, D2, D3 and D4 are respectively obtained based on the 4 singular value detection methods; if a data belongs to any two or more of the preliminary singular value data sets D1, D2, D3 and D4, the data is singular values.
9. The intelligent ammeter operation error monitoring system according to claim 8, wherein the K-means algorithm based singular value detection method comprises:
s11, setting the number of cluster clustersk；
S12, slavemOptionally in the datakThe data is used as an initial clustering center
；
S13, calculating the distance between each data and the clustering center
And respectively classifying each data into the class of the nearest cluster center to obtainkClusters of (a) wherein->
Represent the firstiI=1, 2, … m,
represent the firstjCluster centers, j=1, 2, … k;
s14, taking the average value of the data of each cluster as a new cluster center, repeating S13 to obtain a new cluster centerkClustering clusters, ifkThe cluster clusters are not changed any more, and S15 is switched; otherwise, turning to S13;
s15, if the objective function converges or meets the terminationConditions to obtainkClustering clusters; otherwise, turning to S12;
s16, countingkThe data in the clusters with the least data form a preliminary singular value data set D1.
10. The intelligent ammeter operation error monitoring system according to claim 8, wherein the singular value detection method based on the local anomaly factor algorithm comprises:
s21, calculating any pointpIs the first of (2)kFirst of the points in the neighborhoodkThe distance can be reached:
in the method, in the process of the invention,
is a neighborhood pointoIs the first of (2)kDistance d (o, p) is the neighborhood pointoTo the pointpIs a distance of (2);
s22, calculating any pointpIs the local first of (2)kLocal reachable density:
in the method, in the process of the invention,
is thatpPoint NokDistance neighborhood->
The number of data points in the memory;
s23, calculating any pointpIs the first of (2)kLocal outlier factor:
s24, selecting data points corresponding to the largest local outliers to form a preliminary singular value data set D2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310408981.2A CN116148753A (en) | 2023-04-18 | 2023-04-18 | Intelligent electric energy meter operation error monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310408981.2A CN116148753A (en) | 2023-04-18 | 2023-04-18 | Intelligent electric energy meter operation error monitoring system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116148753A true CN116148753A (en) | 2023-05-23 |
Family
ID=86360337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310408981.2A Pending CN116148753A (en) | 2023-04-18 | 2023-04-18 | Intelligent electric energy meter operation error monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116148753A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116859321A (en) * | 2023-09-04 | 2023-10-10 | 青岛鼎信通讯科技有限公司 | Electric energy meter metering error monitoring method based on energy controller |
| CN117272033A (en) * | 2023-11-23 | 2023-12-22 | 智联信通科技股份有限公司 | DC shunt current metering abnormity monitoring method |
| CN117630798A (en) * | 2023-11-27 | 2024-03-01 | 国网四川省电力公司营销服务中心 | Error monitoring method, device, equipment and medium for cluster type direct current electric energy meter |
| CN118070197A (en) * | 2024-04-17 | 2024-05-24 | 国网冀北电力有限公司 | Data-driven-based electric energy meter running state online monitoring method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110082699A (en) * | 2019-05-10 | 2019-08-02 | 国网天津市电力公司电力科学研究院 | A kind of low-voltage platform area intelligent electric energy meter kinematic error calculation method and its system |
| CN113126019A (en) * | 2021-04-19 | 2021-07-16 | 广东电网有限责任公司计量中心 | Intelligent ammeter error remote estimation method, system, terminal and storage medium |
| CN113515512A (en) * | 2021-06-22 | 2021-10-19 | 国网辽宁省电力有限公司鞍山供电公司 | Quality control and improvement method for industrial internet platform data |
| CN115079081A (en) * | 2022-05-30 | 2022-09-20 | 广州城市理工学院 | Intelligent electric meter metering abnormity identification method and device based on electricity consumption data |
-
2023
- 2023-04-18 CN CN202310408981.2A patent/CN116148753A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110082699A (en) * | 2019-05-10 | 2019-08-02 | 国网天津市电力公司电力科学研究院 | A kind of low-voltage platform area intelligent electric energy meter kinematic error calculation method and its system |
| CN113126019A (en) * | 2021-04-19 | 2021-07-16 | 广东电网有限责任公司计量中心 | Intelligent ammeter error remote estimation method, system, terminal and storage medium |
| CN113515512A (en) * | 2021-06-22 | 2021-10-19 | 国网辽宁省电力有限公司鞍山供电公司 | Quality control and improvement method for industrial internet platform data |
| CN115079081A (en) * | 2022-05-30 | 2022-09-20 | 广州城市理工学院 | Intelligent electric meter metering abnormity identification method and device based on electricity consumption data |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116859321A (en) * | 2023-09-04 | 2023-10-10 | 青岛鼎信通讯科技有限公司 | Electric energy meter metering error monitoring method based on energy controller |
| CN116859321B (en) * | 2023-09-04 | 2023-12-29 | 青岛鼎信通讯科技有限公司 | Electric energy meter metering error monitoring method based on energy controller |
| CN117272033A (en) * | 2023-11-23 | 2023-12-22 | 智联信通科技股份有限公司 | DC shunt current metering abnormity monitoring method |
| CN117272033B (en) * | 2023-11-23 | 2024-03-01 | 智联信通科技股份有限公司 | DC shunt current metering abnormity monitoring method |
| CN117630798A (en) * | 2023-11-27 | 2024-03-01 | 国网四川省电力公司营销服务中心 | Error monitoring method, device, equipment and medium for cluster type direct current electric energy meter |
| CN117630798B (en) * | 2023-11-27 | 2024-06-11 | 国网四川省电力公司营销服务中心 | Error monitoring method, device, equipment and medium for cluster type direct current electric energy meter |
| CN118070197A (en) * | 2024-04-17 | 2024-05-24 | 国网冀北电力有限公司 | Data-driven-based electric energy meter running state online monitoring method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110082699B (en) | Low-voltage transformer area intelligent electric energy meter operation error calculation method and system | |
| CN105117602B (en) | A kind of metering device running status method for early warning | |
| CN116148753A (en) | Intelligent electric energy meter operation error monitoring system | |
| CN110070282B (en) | Low-voltage transformer area line loss influence factor analysis method based on comprehensive relevance | |
| CN111178611B (en) | Method for predicting daily electric quantity | |
| CN110658487A (en) | Meter box and system capable of achieving intelligent electric meter error online estimation | |
| CN113126019B (en) | Remote estimation method, system, terminal and storage medium for error of intelligent ammeter | |
| WO2022021726A1 (en) | Pmu-based power system state estimation performance evaluation method | |
| CN112149873B (en) | Low-voltage station line loss reasonable interval prediction method based on deep learning | |
| CN111628494B (en) | Low-voltage distribution network topology identification method and system based on logistic regression method | |
| CN110674120A (en) | Wind power plant data cleaning method and device | |
| CN111062620B (en) | Intelligent electric power charging fairness analysis system and method based on hybrid charging data | |
| CN110795690A (en) | Wind power plant operation abnormal data detection method | |
| CN111598165A (en) | Density clustering outlier detection method based on extreme learning machine | |
| CN114611738A (en) | Load prediction method based on user electricity consumption behavior analysis | |
| CN112288157A (en) | Wind power plant power prediction method based on fuzzy clustering and deep reinforcement learning | |
| CN117878979B (en) | Power balance and dynamic compensation system based on electric energy storage | |
| Qi et al. | Load pattern recognition method based on fuzzy clustering and decision tree | |
| CN115204698A (en) | Real-time analysis method for power supply stability of low-voltage transformer area | |
| CN107274025B (en) | System and method for realizing intelligent identification and management of power consumption mode | |
| CN115829418B (en) | Method and system for constructing load characteristic portraits of power consumers suitable for load management | |
| CN115907228A (en) | Short-term power load prediction analysis method based on PSO-LSSVM | |
| CN115908082A (en) | Enterprise pollution discharge monitoring method and device based on electricity utilization characteristic indexes | |
| CN114722098A (en) | Typical load curve identification method based on normal cloud model and density clustering algorithm | |
| CN112241812A (en) | Low-voltage distribution network topology identification method based on unilateral optimization and genetic algorithm cooperation |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230523 |
|
| RJ01 | Rejection of invention patent application after publication |