CN107727955B - Transformer loss analysis and control method based on power grid line operation error remote calibration - Google Patents
Transformer loss analysis and control method based on power grid line operation error remote calibration Download PDFInfo
- Publication number
- CN107727955B CN107727955B CN201710788741.4A CN201710788741A CN107727955B CN 107727955 B CN107727955 B CN 107727955B CN 201710788741 A CN201710788741 A CN 201710788741A CN 107727955 B CN107727955 B CN 107727955B
- Authority
- CN
- China
- Prior art keywords
- electric energy
- data
- error
- meter
- energy meter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention relates to a transformer loss analysis and control method based on power grid line operation error remote calibration, which is a mathematical method based on a set of new electric energy metering theory and measures the real error of an electric energy metering device by detecting the electric energy data error. The characteristic solves the technical pain point of poor practicability of the metering function of the electric power automation equipment. According to the transformer loss analysis and control method, big data calculation and analysis of electric energy data are utilized, the transformer loss calculation of the transformer is completed rapidly under the condition of no power failure, and the transformer loss analysis and control method based on power grid line operation error remote calibration is realized.
Description
Technical Field
The invention belongs to the field of electric energy metering, and particularly relates to a transformer loss analysis and control method based on power grid line operation error remote calibration.
Background
At present, the structure of a general electric energy meter and a mutual inductor determines that an error-undetectable 'dead end' exists in an electric energy metering device: the error of the electric energy meter is a scalar quantity, and the error of the current-voltage transformer is a vector synthesized by a ratio difference and an angle difference. Since no operation can be done between vector and scalar. The error of the electric energy metering device cannot be obtained by respectively detecting the errors of the electric energy meter and the mutual inductor. Moreover, the number of household electric energy meters is huge, the workload of detection is overlarge, and the detection must be powered off, so that the problem that the error of the electric energy metering device cannot be detected is solved.
The loss detection is very beneficial to the prediction of the service life of the transformer and is also a necessary technical work for reducing the loss, saving energy and reducing consumption. The actual measurement of the variable loss is a technical problem in the world, and the new technology fully utilizes calibration electric power automation equipment to obtain error-free electric energy data to detect the variable loss of the transformer.
Through search, published patent documents of similar technologies are not found.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a transformer loss analysis and control method based on power grid line operation error remote calibration.
The technical problem to be solved by the invention is realized by adopting the following technical scheme:
a transformer loss analysis and control method based on power grid line operation error remote calibration is characterized in that: connecting the power grid central processing system with a whole-network electric energy metering device by using an energy internet, and processing electric energy data by using a big data system;
⑴, dividing detection subareas, taking a local area power grid as one detection subarea, and installing electric energy metering devices on the high-voltage side and the low-voltage side of a transformer of each transformer substation;
⑵ each detection partition is provided with 2-3 detachable single-phase electric energy meters as error standard devices for error detection, and the known errors of the error standard devices are used for correcting the overall data deviation of the error calculation result;
⑶, acquiring all the electric energy data of the detection subarea, substituting the data into an electric energy data error calculation model, and calculating transformer loss;
the integral real errors of the electric energy metering devices on the high-voltage side and the low-voltage side of the transformer are respectively calculated, and the transformer loss can be calculated after the electric energy data are corrected; it is noted that the reactive current also causes copper loss;
⑷, after data acquisition is completed, 2-3 detachable single-phase electric energy meters installed in the detection subarea are detached to the laboratory, the source is traced to the standard current, voltage transformer and standard electric energy meter, and the accuracy of the calculation result is checked;
⑸, managing and controlling the specific detection subareas according to the obtained transformer loss and variation calculation results.
Moreover, the calculation steps of the electric energy data error calculation model are as follows:
① the intelligent electric energy meter cluster is installed to form a tree topology structure, and the general meter and each sub-meter flow data in the intelligent electric energy meter cluster can be remotely obtained;
②, obtaining a flow conservation algorithm model of the intelligent electric energy meter cluster according to a tree topology structure formed by the intelligent electric energy meter cluster;
③, introducing a virtual branch to correct the intelligent electric energy meter cluster topology model, and correcting the flow conservation algorithm model;
④, calculating relative errors, acquiring daily incremental data of a plurality of groups of intelligent electric energy meter clusters, substituting the acquired daily incremental data of the intelligent electric energy meter clusters into the corrected flow conservation algorithm model, and calculating errors;
⑤ correcting the calculation result and evaluating the uncertainty;
⑥ to obtain the error calculation result.
And the flow conservation algorithm model is that Y epsilon is- η, wherein,
ε＝(ε1ε2… εn-1)T，η＝(y1,0ε0y2,0ε0… ym,0ε0)T；
relative error delta input to any instrument0It is then possible to deduce the relative error delta of the other meters from the sets of measurements of all metersj,j＝1,2,…,n-1。
And the virtual branch circuit replaces the sum of loss values of line loss, and comprises a virtual intelligent electric energy meter and a virtual load.
And, the outflow of the virtual branch is a cluster summary table M0Increment of reading minus each sub-table Mj(j ═ 1, …, n) and this value is multiplied by a loss factor determined from empirical values; after adding the virtual branch, the corrected flow conservation algorithm model is
α is a correction experience.
Moreover, the method for solving the corrected flow conservation algorithm model comprises the following steps:
acquiring daily increment data of the intelligent electric energy meter cluster for n-1 times, substituting the daily increment data into a flow conservation algorithm model:
y ∈ - η, wherein,
wherein
yiIs the ith measurement result vector; e ═ e (e)1ε2… εn-1)T；η＝(y1,0ε0-α(y1,0x1,n) y2,0ε0-α(y2,0x2,n) … yn-1,0ε0-α(y1,0xn-1,n))T；
In order to solve the equation set, matrix Y is divided into LU, Y is LU, z is Uepsilon, and the equation set Lz is- η, wherein Z is easy to solve because L is a lower triangular matrix, and epsilon is easy to solve because U is an upper triangular matrix, and then relative error delta is obtainedj。
And, the step ⑷ is to obtain the daily increment data of the multiple groups of intelligent electric energy meter clusters, before calculation, check the reasonability and integrity of the data, perform data preprocessing on the metering database, and screen out the data with strong independence.
Moreover, the steps of the data preprocessing method are as follows in sequence: real-time inspection, data integrity inspection, data integration, missing item processing, abnormal data discovery by data analysis and orthogonality inspection.
And in the calculating method of the uncertainty evaluation, under the condition that the given initial empirical value α is 0, the clustering error condition of the electric energy meters is calculated, if the calculation result meets the reliability requirement, namely the calculated out-of-tolerance electric energy meter quantity is less than the specified range, the algorithm is ended, otherwise, the value interval of α is adjusted upwards, and the algorithm is returned to solve.
The invention has the advantages and positive effects that:
1. the method is based on a new electric energy metering theory, is a mathematical method, and measures the real error of the electric energy metering device by detecting the electric energy data error. The characteristic solves the technical pain point of poor practicability of the metering function of the electric power automation equipment.
2. The method obtains the error of the electric energy data by calculating and analyzing the electric energy data. The characteristic can solve the technical pain problem that the error detection of the electric energy metering device must be powered off.
3. The method is an application of big data and machine learning technology in the technical field of electric energy metering. The process of calculation is also the process of machine learning. The more times of calculation, the more accurate the calculation result will be.
4. The method comprises a quantity value transmission path: the error of at least one power data in the system is true and credible. The method solves the problem that the error of the structure of the electric energy meter and the mutual inductor cannot be measured.
5. The method can detect the errors of all the electric energy metering devices of one power grid at one time, and can greatly improve the working efficiency. The pain point of the electric energy meter which cannot be detected due to too large quantity can be solved, so that the problem that the pain point of the household electric energy meter which cannot be detected on site due to too large quantity can be solved.
6. According to the transformer loss analysis and control method, big data calculation and analysis of electric energy data are utilized, the transformer loss calculation of the transformer is completed rapidly under the condition of no power failure, and the transformer loss analysis and control method based on power grid line operation error remote calibration is realized.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a schematic diagram of a cluster of intelligent electric energy meters under a tree topology;
FIG. 3 is a schematic diagram of a power meter cluster incorporating a virtual branch;
FIG. 4 is a diagram illustrating variation of the measurement reliability R (t);
FIG. 5 is a diagram of an electric energy meter and a power grid field simulation operation;
FIG. 6 is a block diagram of the operating principle of the load simulation unit;
FIG. 7 is a PWM rectifying and boosting circuit;
fig. 8 is a resistance discharge circuit.
Detailed Description
The present invention will be described in further detail with reference to the following embodiments, which are illustrative only and not limiting, and the scope of the present invention is not limited thereby.
A transformer loss analysis and control method based on power grid line operation error remote calibration is characterized in that a power grid central processing system is connected with a full-grid electric energy metering device through an energy internet, and electric energy data are processed through a big data system;
⑴, dividing detection partitions, and constructing a virtual intelligent electric energy meter cluster tree topology structure in a power grid central processing system;
for convenience of management and calculation, a local area network is taken as a detection partition, for example, 30 substations are taken as a local area network partition; the high-voltage side and the low-voltage side of the transformer of each transformer substation are respectively provided with an electric energy metering device;
⑵ each detection partition is provided with 2-3 detachable single-phase electric energy meters as error standard devices for error detection, and the known error of the error standard devices is used for correcting the overall data deviation of the error calculation result;
the error detection of the electric energy metering device of the power grid is greatly influenced by the error standard device. In the error detection process, the high-precision high-voltage electric energy meter is used for the traditional electric energy metering device with good calibration error stability, the electric energy data precision of the latter is improved, and the high-precision high-voltage electric energy meter is used as 'error standard electric energy data' and is very helpful for detecting the electric energy metering error of a power grid.
⑶, acquiring all the electric energy data of the detection subarea, substituting the data into an electric energy data error calculation model, and calculating transformer loss;
the integral real errors of the electric energy metering devices at the high side and the middle side and the low side of the transformer are respectively calculated, and the transformer loss can be calculated after the electric energy data are corrected; it is noted that the reactive current also causes copper loss;
⑷, after data acquisition is completed, 2-3 detachable single-phase electric energy meters are installed in the detection subarea, the single-phase electric energy meters are detached to the laboratory, the sources of the single-phase electric energy meters are traced to the standard current, the voltage transformer and the standard electric energy meters, and the accuracy of the calculation result is checked;
⑸, managing and controlling the specific detection subareas according to the transformer loss calculation result.
Under the conventional technology, a recognition error area of line loss measurement exists, and it is considered that electric energy data measured by power system automation equipment (such as an integrated automation unit, a DTU and the like) cannot be used for detecting line loss. However, through experimental verification, when the method is used for remotely verifying the electric energy data errors of the automation equipment on line, the line loss is calculated by using the electric energy data errors, so that the value of the stock equipment can be improved.
The calculation steps of the electric energy data error calculation model are as follows:
① the intelligent electric energy meter cluster is installed to form a tree topology structure, and the general meter and each sub-meter flow data in the intelligent electric energy meter cluster can be remotely obtained;
②, obtaining a flow conservation algorithm model of the intelligent electric energy meter cluster according to a tree topology structure formed by the intelligent electric energy meter cluster;
③, introducing a virtual branch to correct the intelligent electric energy meter cluster topology model, and correcting the flow conservation algorithm model;
④, calculating relative errors, acquiring daily incremental data of a plurality of groups of intelligent electric energy meter clusters, substituting the acquired daily incremental data of the intelligent electric energy meter clusters into the corrected flow conservation algorithm model, and calculating errors;
⑤ correcting the calculation result and evaluating the uncertainty;
⑥ to obtain the error calculation result.
The embodiment mainly takes an intelligent electric meter as an example, and illustrates the implementation steps of the method: the method calculates the error of the electric energy meter by an algorithm according to the daily frozen electric quantity, and realizes real-time remote monitoring on the error of the electric energy meter. And carrying out remote calibration on the electric energy meter by using meter reading data of the electric energy meter acquisition system.
The cluster formed after the intelligent electric energy meter is similar to the intelligent electric energy meter has tree topology, so that the actual flow increment of the total meter in the same period is equal to the sum of the actual flow increments of the sub-meters under the constraint of flow conservation. Since the actual flow increment can be expressed in reading increments and relative errors, an equation can be derived that contains all the meter reading increments and relative errors. If the relative error of any one meter in the cluster is known and the relative error of the other meters is taken as an unknown quantity, and it is noted that the equations can be made to reach or exceed the number of unknowns by increasing the measurement period, then the relative error of the other meters can be determined by solving the system of equations. The autonomous algorithm calculates errors by mutual comparison of the cluster internal instruments without the help of external standard instruments.
When the tree topology flow tube has leakage, the equation system must be corrected. In electric energy metering, leakage manifests as line leakage, resistive losses, and meter power consumption. In the method, the loss is regarded as the load of the virtual branch and a virtual electric energy meter is introduced. And determining a loss correction term by utilizing the electrical parameter measuring function of the intelligent electric energy meter. And giving out the corrected algorithm flow and the simulation result.
The method comprises the following specific decomposition processes:
for the sake of clarity of the description of the method, the following definitions are first given:
flow: a scalar quantity (shown by an arrow in fig. 2) flowing into or out of a certain closed curve S. Inflow is positive and outflow is negative. Convention is made for the flow as follows: the flows do not pile up and obey the law of conservation at any time. I.e. the algebraic sum of the flows through any closed surface S is equal to zero. Streams are marked with integers starting from 0.
Generalized flow rate: flow, or flow integrated over time. Flow conservation can be inferred from flow conservation. I.e. the algebraic sum of the flow increments through the closed surface S equals zero in any time period.
The intelligent electric energy meter comprises: abbreviated as flow meter or flow meter (M in figure 2)0～M4) The instantaneous flow of the corresponding stream is recorded. Flow meter with MjThe label, j, is the number of the corresponding stream.
Clustering: the set of flow meters for all flows through a closed curve S.
Meter reading: readings are taken of all meters within the cluster at a time. Logically, the meter reading action should return the readings of all the meters at the same time. In practice, however, the same time may be considered if the time interval between the return of each meter reading is sufficiently short.
Measurement: one measurement is taken by two meters spaced apart by a period of time for the cluster, resulting in an increment of two readings before and after.
First, flow conservation in an ideal case:
relative error δ:
where x is the actual increase in flow through a meter over a period of time and y is the increase in the meter reading over the same period of time.
Certain closed surface S-defined flow meter cluster AS＝{MjThe | flow j passes through the closed curved surface S }, and the number N of the flow meters in the clusterSN. With a flow meter MjThe ith measurement is yijThe actual value corresponding thereto is xij. Where i is 1,2, …, m, m is the number of measurements. According to the flow conservation convention have
If deltajIndicating a flow meter MjThe relative error of (2) can be obtained by defining the relative error of the section
Substituting it into formula (2) with
This is true. Without loss of generality, assume δ0Known, then there are
Order to
Substituted into formula (3) with
(5) The formula can be written in matrix form
Yε＝-η (6)
Wherein
ε＝(ε1ε2… εn-1)T，η＝(y1,0ε0y2,0ε0… ym,0ε0)T。
(6) The formula shows that if the relative error delta of a certain meter in the cluster is known0Then the relative error delta of other meters can be calculated by multiple groups of measurement results of all metersj,j＝1,2,…,n-1。
Solving of a system of equations
Cluster capable of measuring current
The measurement times m should not be less than n-1, otherwise the formula (13) is an indefinite equation set; when m is greater than n-1, the redundancy equation should be deleted. In the discussion below, m-n-1 is assumed. Writing formula (13) in matrix form
Yε＝-η (7)
Wherein
yiIs the ith measurement result vector; e ═ e (e)1ε2… εn-1)T；η＝(y1,0ε0-α(y1,0x1,n) y2,0ε0-α(y2,0x2,n) … yn-1,0ε0-α(y1,0xn-1,n))T。
To solve the equation set of formula (14), matrix Y is decomposed into LU, Y equals LU, z equals U ∈, and equation set Lz equals- η, where L isThe lower triangular array is easy to solve out z; since U is an upper triangular matrix, epsilon is easy to be solved, and then relative error delta is obtainedj。
The remote calibration of the error of the electric energy meter is to calculate the error of the electric energy meter by using the acquired data, so the requirement on the data is high, the accuracy of the measurement result of the measured meter seriously influences the calculation result, and a complete and reasonable data screening and pretreatment are necessary. The step removes noise in the data, fills in null values, missing values and processes inconsistent data by filling in missing data, eliminating abnormal data, smoothing noisy data, and correcting inconsistent data. As shown in the figure, firstly, measurement results of all meters in the whole tree structure system need to be obtained, and under the condition that measurement together has uncertainty, data can be obtained as much as possible so as to ensure the accuracy and completeness requirements of the results under reasonable calculation errors.
And (3) real-time checking: we check if the same set of meter values were read at the same time point. If this requirement is not met, the data cannot be used.
Data integration: the data integration is that the meter reading data of the electric energy meter is added into the historical meter reading data of the meter on the basis of the original meter reading data.
Integrity checking and missing value processing: first we need to perform an integrity check on the data. The main task of integrity checking is missing value handling. For missing value processing, the traditional method is to directly delete or replace with average value, median, quantile, mode, random value, etc. This works generally because it is equivalent to a person making a modification to the data. The method is characterized in that a missing value is calculated by taking a historical error and other error quantities as a prediction model, the missing value is calculated and then added into an algorithm for secondary verification, and whether the condition of the missing value can meet the missing requirement or not is judged. Specifically, if there are many missing values and the missing amount is too large in a group of data, we will directly discard the data, because if the missing value is introduced, large noise will be generated, and the result will be adversely affected. According to the judgment of an empirical value, if the missing value is more than 10%, the meter reading data is more prone to be waited for. If the requirement can be met, firstly, the reading of the electric energy meter with the missing collected data is added into the electricity consumption of the virtual branch, and the error condition of the electric energy meter is calculated by using an original model. If the calculated result is kept within a reasonable range after the data is added to the virtual branch circuit for updating, the line loss is deducted from the electric quantity of the virtual branch circuit according to the average line loss ratio of the distribution room to fill the data. After filling, recalculating the error condition by using the filled table, and if the conditions that the other meters do not change much and the missing meter does not deviate much from the historical calculation value twice are met, determining that the filling of the result is correct. Without loss of generality, the electric energy meter with the missing value is taken as a table 1, and under the condition that the measurement times m are not less than n-1, the formula (7) is written into a matrix form
Yε＝-η' (8)
Wherein
yiIs the ith measurement result vector; e ═ e (e)2ε3…εn-1)T；η'＝η-y·1＝(y1,0ε0-α(y1,0x1,n)-y11y2,0ε0-α(y2,0x2,n)-y12… yn-1,0ε0-α(y1,0xn-1,n)-y1n)T. From the above, the same equation (15) can be solved. If ε ═ ε2ε3…εn-1)TAnd if the requirement is met, the equation is reasonable, and the result is recovered according to the virtual branch value.
Data analysis (with the aid of visualization tools) finds anomalous data: the daily meter reading data of the electric energy meter is used as a vector, and a vector group can be formed by reading for many times. In the multidimensional data analysis technology, each vector is visually represented, so that a special vector with abnormal variation is found out. By using a CrystalAnalysis visualization tool method, a special outlier value can be found, special analysis is carried out on the outlier value, data monitoring is further enhanced, and core data are optimized. In addition, the vector is clustered and analyzed by using a clustering algorithm, and the vector is used as a special discovery method based on the abnormal data discovered by visual representation.
And (3) testing orthogonality: the error analysis algorithm requires weak correlation among equations in the equation set, otherwise, the equation set is ill-conditioned, so that the error of the calculation result exceeds an acceptable range. The equations can only be calculated when sufficient requirements are met.
Secondly, adding a virtual branch to calculate loss:
the loss (electric energy meter loss, leakage loss, line resistance loss and the like) exists in an actual electric energy meter cluster, and the concept of a virtual branch is introduced to replace the sum of loss values with the line loss as a theme in the problem. The branch circuit includes a virtual power meter and a virtual load. The cluster total loss may be equivalent to the energy consumption of the virtual load. In FIG. 3, the imaginary enclosed curved surface S defines a cluster of meters, which are denoted by the symbol MjAnd (j ═ 0,1, …, n). MnIs a virtual electric energy meter. The energy input into S is specified to be positive and the output is specified to be negative.
Set at the ith measurement period TiInner flow through MjElectric energy of xi,j(i ═ 1,2, …, m, which is a number identifying each measurement period, and m is the number of measurements), the following equation holds according to the law of conservation of energy.
Generally, M is considered to be within the ith measurement periodjIncrement of reading yi,jAnd xi,jHas the following relationship.
yi,j＝(1+δj)xi,j(10)
(10) In the formula, deltajIs MjIs measured (averaged) relative error. If order
Then formula (11) can be changed into
Without loss of generality, assume that M is known0Relative error delta of0Thus, let us know ∈0. Specifying virtual electric energy meters MnRelative error delta ofn0, thus having epsilon n1 and xi,n＝yi,nThis is true. Thus (2-22) is formulated as
Expression of losses in a three, virtual branch
(13) Third term x on the left in the formulai,nRepresenting a measurement period TiThe energy consumption (total cluster loss) of the internal virtual load, and the equations of the equations (17) and (19) are energy conservation equations. They show the relative error delta for a particular meter in a cluster if known, given conservation of energy0It is then possible to pass through multiple sets of measurements yi,jDeducing epsilon of all other electric energy metersjAnd then the relative error delta is obtainedj(j ═ 1,2, …, n-1). In actual computation, the outflow of the virtual branch is a cluster summary table M0Increment of reading minus each sub-table Mj(j-1, …, n) and multiplying this value by a loss factor determined from empirical values. The loss coefficient is used as the energy loss of the electric energy meter cluster and is obtained by the following loop iteration method. To this end, after adding the virtual branch, the equation changes accordingly to
α is a correction experience and is a random value within a certain range.
And (3) screening data:
the method comprises the steps of firstly acquiring relevant meter reading data of the electric energy meter from an electricity utilization information acquisition system, and removing noise, filling null values, lost values and processing inconsistent data in the data by filling missing data, eliminating abnormal data, smoothing noise data and correcting inconsistent data so as to obtain effective data.
Firstly, measurement results of all meters in the whole tree structure system need to be obtained, and under the condition that measurement is uncertain, data can be obtained as much as possible so as to guarantee the requirements of accuracy and completeness of results under reasonable calculation errors. The minimum data quantity is not less than the number of electric meters plus one. The data screening comprises the following steps: 1. and (5) checking real-time performance. 2. Integrity checking and missing value handling. 3. Visualization method finding abnormal data 4. orthogonality test
And (3) real-time checking: we check if the same set of meter values were read at the same time point. If this requirement is not met, the data cannot be used.
Data integration: the data integration is that the meter reading data of the electric energy meter is added into the historical meter reading data of the meter on the basis of the original meter reading data.
Integrity checking and missing value processing: first we need to perform an integrity check on the data. The main task of integrity checking is missing value handling. For missing value processing, the traditional method is to directly delete or replace with average value, median, quantile, mode, random value, etc. This works generally because it is equivalent to a person making a modification to the data. The method is characterized in that a missing value is calculated by taking a historical error and other error quantities as a prediction model, the missing value is calculated and then added into an algorithm for secondary verification, and whether the condition of the missing value can meet the missing requirement or not is judged. Specifically, if there are many missing values and the missing amount is too large in a group of data, we will directly discard the data, because if the missing value is introduced, large noise will be generated, and the result will be adversely affected. According to the judgment of an empirical value, if the missing value is more than 10%, the meter reading data is more prone to be waited for. If the requirement can be met, firstly, the reading of the electric energy meter with the missing collected data is added into the electricity consumption of the virtual branch, and the error condition of the electric energy meter is calculated by using an original model. If the calculated result is kept within a reasonable range after the data is added to the virtual branch circuit for updating, the line loss is deducted from the electric quantity of the virtual branch circuit according to the average line loss ratio of the distribution room to fill the data. After filling, recalculating the error condition by using the filled table, and if the conditions that the other meters do not change much and the missing meter does not deviate much from the historical calculation value twice are met, determining that the filling of the result is correct. Without loss of generality, the electric energy meter with the missing value is taken as a table 1, and the formula (14) is written into a matrix form under the condition that the measurement times m are not less than n-1
Yε＝-η' (15)
Wherein
yiIs the ith measurement result vector; e ═ e (e)2ε3…εn-1)T；η'＝η-y·1＝(y1,0ε0-α(y1,0x1,n)-y11y2,0ε0-α(y2,0x2,n)-y12… yn-1,0ε0-α(y1,0xn-1,n)-y1n)T. From the above, the same equation (15) can be solved. If ε ═ ε2ε3… εn-1)TAnd if the requirement is met, the equation is reasonable, and the result is recovered according to the virtual branch value.
Data analysis (with the aid of visualization tools) finds anomalous data: the daily meter reading data of the electric energy meter is used as a vector, and then a plurality of sets of reading can form a vector set. In the multidimensional data analysis technology, each vector is visually represented, so that a special vector with abnormal variation is found out. By using a CrystalAnalysis visualization tool method, a special outlier value can be found, special analysis is carried out on the outlier value, data monitoring is further enhanced, and core data are optimized. In addition, the vector is clustered and analyzed by using a clustering algorithm, and the vector is used as a special discovery method based on the abnormal data discovered by visual representation.
And (3) testing orthogonality: the error analysis algorithm requires weak correlation among equations in the equation set, otherwise, the equation set is ill-conditioned, so that the error of the calculation result exceeds an acceptable range. The equations can only be calculated when sufficient requirements are met.
The calculation method comprises the following steps: electric energy meter error iterative calculation method based on measurement uncertainty
Since the method measures the average error of the electric energy meter in a period of time, the measured result is only the measured estimated value, and the uncertainty of the measurement can be caused by uncertain factors in the measuring process. The measurement result obtained after correcting the known system effect is still only the measured estimation value, and an uncertainty evaluation method is considered, so that the measurement can still reflect the real situation under the conditions of incomplete ideal reproduction or insufficient sample representativeness and the like.
The basic idea is to derive a new value of a variable from its original value in our method, iteration is applied to the estimation of the virtual circuit modification experience α.
Here, first, the assumption is given: the number of error over-tolerance tables in one station area is less than 5% of the total number of tables, namely that the reliability of the measuring instrument is considered to be 95%.
In general, the reliability of the metrology tool will decrease with time, so we perform a floating certification reliability target range for this ratio at different stations. Fig. 4 is a diagram illustrating the variation of measurement reliability with time. According to the average service time of the table meter in the station area, the value is corrected.
The method is briefly described as follows, under the condition that a given initial empirical value α is 0, the clustering error condition of the electric energy meters is calculated, if the calculation result meets the reliability requirement, namely the calculated out-of-tolerance electric energy meter quantity is less than a specified range, the algorithm is ended, otherwise, the value interval of α is adjusted upwards, and the algorithm is returned to be solved.
The joint uncertainty evaluation method comprises the following steps:
the mathematical model is that delta is 1/(Y)-1η)-1
then we have a joint extension uncertainty U-kucTaking k as 2, the following components are:
if the error of the solved electric energy meter is in accordance with the actual expectation, then the measured value is independently and repeatedly observed, the experimental standard deviation s (x) is obtained by a statistical analysis method through a series of obtained measured values, and when the arithmetic mean value is used as the estimated value of the measured value, the uncertainty of the estimated value is calculated as follows:
independently and repeatedly observing the same measured quantity for n times under the repetitive condition by using a Bessel formula method to obtain n measured values xiThe best estimate of the measured X is the arithmetic mean of n independently measured values ( i 1, 2.. times.n)
Calculated according to the following formula:
single measured value xkExperimental variance s of2(x)：
Single measured value xkExperimental standard deviation of (a), (b):
measured estimated value
Uncertainty of
District platform body model structure that this embodiment was built
In order to promote the reproduction of field problems and better analyze and solve the field problems, the project simulates and establishes the technical scheme of the field operation states of the electric energy meter and the power grid, and the working principle block diagram of the project is shown in fig. 5. The functional block diagram mainly comprises a PC unit, a multi-channel communication service unit, a high-power program-controlled alternating-current voltage source, a gateway electric energy meter, a line loss simulation unit, a tested electric energy meter, a corresponding load simulation unit and the like.
The working principle is as follows: the PC controls all the devices through the multi-channel communication service unit, the amplitude, the frequency and the harmonic content of the high-power program-controlled alternating-current voltage source can be set randomly so as to simulate the influence of the change of the power quality of a power grid on metering, and meanwhile, the line loss between a gateway electric energy meter and a common electric energy meter is simulated by using the line loss simulation unit. Considering that the line loss unit usually has active loss, the line loss simulation unit adopts the material object resistive load equivalent with adjustable impedance, and the range of the line loss simulation unit is continuously adjustable from 0 omega to 10 omega. For the load simulation unit, the current household load characteristics are considered to be more changeable and more complex, and the load simulation unit is realized by adopting an electronic load mode in the scheme. Through the parameter setting of the PC software, the power factor and the power of the electronic load can be set at will, so that the actual metering operation load of an electric energy field can be simulated better. The following description focuses on the operation principle of the load simulation unit.
The load simulation unit is designed and realized by comprehensively utilizing a power electronic technology, an embedded software technology and an electric power system automation technology, and can simulate and adjust the power factor and the power magnitude of the load. The operation principle block diagram of the analog load is shown in fig. 6, and the basic operation principle is as follows: firstly, the input voltage of the test power supply is sampled, the input voltage signal is analyzed, and the voltage signal is used as the reference signal of the current reference signal Iref. And generating a corresponding current reference signal Iref by the singlechip according to the set working mode (constant voltage, constant current, constant power and constant resistance) and power factor. The reference signal and the actual input current Is are subjected to differential amplification to obtain an Ierr signal, the Ierr signal Is compared with a triangular wave of 23kHz to generate an SPWM pulse signal, the signal Is used for controlling the on and off of an H bridge IGBT through a driving isolation circuit, and Boost Is boosted to a direct-current voltage Udc. The rear-stage discharge circuit generates PWM waves with corresponding duty ratios by the singlechip according to the sampled direct-current voltage Udc, and controls the on-off time of the IGBT by driving the isolation circuit, thereby controlling the connection and disconnection of a nickel-chromium wire resistor (a small resistor and a large resistor), and equivalently forming a required resistance value. The duty cycle of the PWM is adjusted according to the amount of change in Udc to maintain Udc at the designed value.
The simulation load designed by the scheme meets the requirement that the test power supply performs tests in different states. The designed analog load is of a resistance consumption type, namely, the generated energy of the analog load is consumed by the resistance in the test process. According to the working principle of the analog load circuit, the implementation scheme of the analog load circuit is mainly divided into two stages. The front-stage PWM rectifying and boosting circuit (as shown in figure 7) adopts a fully-controlled device and works in a high-frequency state. The input alternating voltage is boosted to a designed value Udc by adopting a proper SPWM pulse wave to control the on and off of the power switch device.
The latter stage is a resistive discharge circuit, see fig. 8. According to the sampled direct current voltage Udc, the single chip microcomputer generates a proper PWM wave to drive the IGBT to be switched on and off, so that the purpose of different resistance values is achieved. The part of the circuit completes the energy consumption of the test power supply in the test process.
Experimental verification of the algorithm:
we select 19 electric energy meter clusters in a certain area in Tianjin to obtain day freezing data of 150 working days of the clusters. For this data, we perform the algorithm described above for this. We found that after 8 iterations, taking the virtual branch empirical value in the interval [0.8,0.9], the calculation of the error of the electric energy meter is greatly improved and has good results:
table 1 initial state (α ═ 0) error calculation amount
| Reference numerals | Error calculation value | Reference numerals | |
| 1 | 1.5590％ | 10 | -14.5640％ |
| 2 | -2.7938％ | 11 | -9.0017％ |
| 3 | -5.1089％ | 12 | 5.5293％ |
| 4 | -0.2388％ | 13 | 0.0485％ |
| 5 | -3.1209％ | 14 | -0.4204％ |
| 6 | 3767.6317％ | 15 | -3.8429％ |
| 7 | -0.3551％ | 16 | -5.7363％ |
| 8 | -4.0477％ | 17 | -3.1323％ |
| 9 | -2.9027％ | 18 | -2.2543％ |
Table 2 iterative Final (α ∈ [0.8,0.9]) computation results
In the interval, the error of the electric energy meter is found to be in a reasonable level, 10 times of electric energy meter error data are calculated in a slip method according to a measurement uncertainty evaluation method, and a calculation result and a half-width interval under standard inaccuracy evaluation are given:
TABLE 3 calculation results and half-Width intervals thereof
| Reference numerals | Error calculation value | Half-width interval | Reference numerals | Error calculation value | Half- |
| 1 | 0.09％ | 0.02％ | 10 | 0.13％ | 0.09％ |
| 2 | 0.35％ | 0.02％ | 11 | 0.04％ | 0.02％ |
| 3 | -0.21％ | 0.02％ | 12 | -0.11％ | 0.09％ |
| 4 | 0.54％ | 0.02％ | 13 | 0.00％ | 0.00％ |
| 5 | 0.60％ | 0.02％ | 14 | -0.50％ | 0.03％ |
| 6 | -1.90％ | 0.09％ | 15 | 0.43％ | 0.02％ |
| 7 | -0.56％ | 0.02％ | 16 | -0.26％ | 0.07％ |
| 8 | 0.08％ | 0.01％ | 17 | 0.10％ | 0.02％ |
| 9 | -0.09％ | 0.01％ | 18 | 0.06％ | 0.01％ |
After the experimental result is obtained, after the error of the electric energy meter is analyzed, the error of the meter 6 is found to be in a higher condition, namely, the error is about to exceed the tolerance, and the meter is randomly checked by field workers.
Experimental verification shows that by using the algorithm, errors of all other electric energy meters can be calculated only by analyzing the reading of the electric energy meter cluster in the tree topology on the premise that the actual relative error of any electric energy meter is known without an external standard instrument. The autonomous error calculation method can be executed on line, so that the maintenance cost of the electric energy meter in use is expected to be remarkably reduced. When the actual accuracy level of a certain meter is known, but not the actual relative error, the estimation of the error range can be realized although the exact relative error of other electric energy meters cannot be calculated. The method can effectively lock the suspected out-of-tolerance electric energy meter, so that the calibration and calibration are purposeful.
Although the embodiments of the present invention and the accompanying drawings are disclosed for illustrative purposes, those skilled in the art will appreciate that: various substitutions, changes and modifications are possible without departing from the spirit and scope of the invention and the appended claims, and therefore the scope of the invention is not limited to the disclosure of the embodiments and the accompanying drawings.
Claims (4)
1. A transformer loss analysis and control method based on power grid line operation error remote calibration is characterized in that: connecting the power grid central processing system with a whole-network electric energy metering device by using an energy internet, and processing electric energy data by using a big data system;
⑴, dividing detection subareas, taking a local area power grid as one detection subarea, and installing electric energy metering devices on the high-voltage side and the low-voltage side of a transformer of each transformer substation;
⑵ each detection partition is provided with 2-3 detachable single-phase electric energy meters as error standard devices for error detection, and the known errors of the error standard devices are used for correcting the overall data deviation of the error calculation result;
⑶, acquiring all the electric energy data of the detection subarea, substituting the data into an electric energy data error calculation model, and calculating transformer loss;
the method specifically comprises the following steps: the virtual branch circuit replaces the sum of loss values of line loss, and comprises a virtual intelligent electric energy meter and a virtual load;
the outflow of the virtual branch is a cluster summary table M0Increment of reading minus each sub-table Mj(j ═ 1, …, n) and multiplying the virtual leg outflow by a loss factor determined empirically; after adding the virtual branch, the corrected flow conservation algorithm model is
α is correction experience;
the integral real errors of the electric energy metering devices on the high-voltage side and the low-voltage side of the transformer are respectively calculated, and the transformer loss can be calculated after the electric energy data are corrected; it is noted that the reactive current also causes copper loss;
the calculation steps of the electric energy data error calculation model are as follows:
① the intelligent electric energy meter cluster is installed to form a tree topology structure, and the general meter and each sub-meter flow data in the intelligent electric energy meter cluster can be remotely obtained;
②, obtaining a flow conservation algorithm model of the intelligent electric energy meter cluster according to a tree topology structure formed by the intelligent electric energy meter cluster;
③, introducing a virtual branch to correct the intelligent electric energy meter cluster topology model, and correcting the flow conservation algorithm model;
④, calculating relative errors, acquiring daily incremental data of a plurality of groups of intelligent electric energy meter clusters, substituting the acquired daily incremental data of the intelligent electric energy meter clusters into the corrected flow conservation algorithm model, and calculating errors;
before calculation, the rationality and integrity of data need to be checked, data preprocessing is carried out on a metering database, and data with strong independence are screened out; the data preprocessing method comprises the following steps in sequence: real-time inspection, data integrity inspection, data integration, missing item processing, abnormal data discovery by data analysis and orthogonality inspection;
⑤ correcting the calculation result and evaluating the uncertainty;
⑥ obtaining an error calculation result;
⑷, after data acquisition is completed, 2-3 detachable single-phase electric energy meters installed in the detection subarea are detached to the laboratory, the source is traced to the standard current and voltage transformers and the standard electric energy meters, and the accuracy of the calculation result is checked;
⑸, managing and controlling the specific detection subareas according to the obtained transformer loss and variation calculation results.
2. The transformer loss analysis and control method based on power grid line operation error remote calibration according to claim 1, wherein the flow conservation algorithm model is Y epsilon- η, wherein,
yifor instrument M1～Mn-1The vector of the result of the ith measurement, wherein i is between 1 and m;
ε＝(ε1ε2εn-1)T，η＝(y1,0ε0y2,0ε0ym,0ε0)T；
where Y is the resulting vector matrix and ε is the error parameter
Wherein if deltajIndicating a flow meter MjRelative error of (2); relative error delta input to any instrument0Then, the relative error delta of other meters is calculated by the multiple groups of measurement results of all the metersjJ ═ 1,2, …, n-1; number N of flow meters in clusterS＝n。
3. The transformer loss analysis and control method based on power grid line operation error remote calibration according to claim 1, characterized in that: the solving method of the corrected flow conservation algorithm model comprises the following steps:
acquiring daily increment data of the intelligent electric energy meter cluster for n-1 times, substituting the daily increment data into a flow conservation algorithm model:
y ∈ - η, wherein,
wherein
yiIs the ith measurement result vector; e ═ e (e)1ε2εn-1)T；η＝(y1，0ε0-a(y1，0x1，n) y2，0ε0-α(y2，0x2，x) … yn-1，0ε0-α(yn-1，0xn-1，n))T；
In order to solve the equation set, matrix Y is divided into LU, Y is LU, z is Uepsilon, and the equation set Lz is- η, wherein Z is easy to solve because L is a lower triangular matrix, and epsilon is easy to solve because U is an upper triangular matrix, and then relative error delta is obtainedj。
4. The transformer loss analysis and control method based on power grid line operation error remote calibration according to claim 1 is characterized in that the uncertainty evaluation calculation method is used for calculating the cluster error condition of the electric energy meters under the condition that a given initial empirical value α is 0, if the calculation result meets the reliability requirement, namely the calculated out-of-tolerance electric energy meter quantity is smaller than a specified range, the algorithm is ended, otherwise, the value interval of α is adjusted upwards, and the algorithm is returned to solve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710788741.4A CN107727955B (en) | 2017-09-05 | 2017-09-05 | Transformer loss analysis and control method based on power grid line operation error remote calibration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710788741.4A CN107727955B (en) | 2017-09-05 | 2017-09-05 | Transformer loss analysis and control method based on power grid line operation error remote calibration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107727955A CN107727955A (en) | 2018-02-23 |
| CN107727955B true CN107727955B (en) | 2020-05-01 |
Family
ID=61204886
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710788741.4A Active CN107727955B (en) | 2017-09-05 | 2017-09-05 | Transformer loss analysis and control method based on power grid line operation error remote calibration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107727955B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109683118A (en) * | 2018-12-10 | 2019-04-26 | 国网冀北电力有限公司电力科学研究院 | Based on the improved electric energy meter protection feature optimization method and device of transformer |
| CN112305492B (en) * | 2020-10-30 | 2021-08-31 | 国网山东省电力公司营销服务中心(计量中心) | Intelligent electric energy meter online error check accurate control system and method thereof |
| CN112684399B (en) * | 2020-11-23 | 2022-04-22 | 国网江苏省电力有限公司营销服务中心 | Electric energy meter operation error monitoring data fitting method and system based on artificial intelligence |
| CN113077020B (en) * | 2021-06-07 | 2021-08-24 | 广东电网有限责任公司湛江供电局 | Transformer cluster management method and system |
| CN114264865B (en) * | 2022-03-03 | 2022-05-24 | 武汉格蓝若智能技术有限公司 | Online self-calibration method for current collection device |
| CN115575884B (en) * | 2022-11-08 | 2023-03-10 | 国网湖北省电力有限公司营销服务中心(计量中心) | Method for aligning charging quantity of charging piles in charging station and calculating metering error |
| CN115524657B (en) * | 2022-11-24 | 2023-03-10 | 中国电力科学研究院有限公司 | Metering error analysis method, metering device and self-calibration method of electric energy meter |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102749541A (en) * | 2012-07-09 | 2012-10-24 | 云南电力试验研究院（集团）有限公司电力研究院 | Energy conservation-based real-time bus and metering equipment state on-line monitoring system |
| CN106026086A (en) * | 2016-07-08 | 2016-10-12 | 国网江苏省电力公司电力科学研究院 | Power grid operation state dynamic estimation method |
| CN106338706A (en) * | 2015-07-10 | 2017-01-18 | 侯飞 | Electric energy metering device overall error detecting method, device and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9390388B2 (en) * | 2012-05-31 | 2016-07-12 | Johnson Controls Technology Company | Systems and methods for measuring and verifying energy usage in a building |
-
2017
- 2017-09-05 CN CN201710788741.4A patent/CN107727955B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102749541A (en) * | 2012-07-09 | 2012-10-24 | 云南电力试验研究院（集团）有限公司电力研究院 | Energy conservation-based real-time bus and metering equipment state on-line monitoring system |
| CN106338706A (en) * | 2015-07-10 | 2017-01-18 | 侯飞 | Electric energy metering device overall error detecting method, device and system |
| CN106026086A (en) * | 2016-07-08 | 2016-10-12 | 国网江苏省电力公司电力科学研究院 | Power grid operation state dynamic estimation method |
Non-Patent Citations (1)
| Title |
|---|
| 面向智能电网AMI的网络计量关键技术与用户用电数据挖掘研究";郭景涛;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20120715(第07期);第C042-69页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107727955A (en) | 2018-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107643507B (en) | Lean line loss analysis and control method based on power grid line operation error remote calibration | |
| CN107817395B (en) | Electricity stealing troubleshooting method based on power grid line operation error remote calibration | |
| CN107727955B (en) | Transformer loss analysis and control method based on power grid line operation error remote calibration | |
| CN111398885B (en) | Intelligent electric meter operation error monitoring method combining line loss analysis | |
| CN110471024B (en) | Intelligent electric meter online remote calibration method based on measurement data analysis | |
| CN107621621A (en) | A kind of intelligent electric energy meter is in Line synthesis error calibration method | |
| CN112698261A (en) | Fault early warning method and device for intelligent electric energy meter | |
| CN104134999B (en) | Distribution network based on multi-data source measures the practical method of calculation of efficiency analysis | |
| CN107818395B (en) | Electric energy meter error iterative calculation method based on measurement uncertainty | |
| CN107870312A (en) | A kind of intelligent electric energy meter kinematic error remote calibration method | |
| CN107145707B (en) | Distribution network transformer planning method considering photovoltaic output uncertainty and life cycle cost | |
| CN107741578B (en) | Original meter reading data processing method for remote calibration of running error of intelligent electric energy meter | |
| CN111162533B (en) | Smart power grid hidden topology structure identification method based on convex optimization | |
| CN107944652A (en) | A kind of broad sense flow instrument kinematic error remote calibration method | |
| CN113078630B (en) | Low-voltage distribution network topology identification method based on real-time measurement data | |
| CN104836223A (en) | Power grid parameter error and bad data coordinated identification and estimation method | |
| CN110299762B (en) | PMU (phasor measurement Unit) quasi-real-time data-based active power distribution network robust estimation method | |
| CN108445435A (en) | A kind of electric energy meter calibrating apparatus error online evaluation method | |
| CN103632235A (en) | Improved forward-backward sweep algorithm based power-grid low-voltage platform area line loss information processing method | |
| CN110363334B (en) | Grid line loss prediction method of photovoltaic grid connection based on gray neural network model | |
| CN113655425B (en) | Metering point operation error monitoring method and system suitable for 10KV wiring line | |
| CN112130109A (en) | Method for detecting metering performance abnormity of intelligent electric energy meter | |
| CN115508770B (en) | KL-NB algorithm-based electric energy meter operation state online evaluation method | |
| CN107257130A (en) | The low-voltage network loss computing method of decoupling is measured based on region | |
| CN109494726A (en) | Stability of power system online evaluation method based on DLRNN neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |