Disclosure of Invention
In order to solve the problems, the invention provides a method and a system for detecting the metering performance of a high-voltage transformer, which can master the metering performance state of the high-voltage transformer in real time on the premise of no power outage, have a more targeted guidance on the operation and maintenance work of the high-voltage transformer, and have important significance for ensuring the safe, stable and economic operation of a power system.
The technical scheme adopted for solving the technical problems is as follows:
on one hand, the method for detecting the metering performance of the high-voltage transformer provided by the embodiment of the invention comprises the following steps:
extracting characteristic factors representing the metering performance of the three-phase voltage transformer;
and comparing the deviation of the characteristic factors under the current operation condition and the normal operation condition, and judging the metering performance of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of extracting the characteristic factor representing the metering performance of the three-phase voltage transformer includes:
according to the relevance of the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the characteristics of the three-phase measured secondary output voltage of the high-voltage transformer are decomposed, and the fluctuation caused by the primary voltage and the fluctuation caused by the metering performance of the high-voltage transformer are separated from the measured value.
As a possible implementation manner of this embodiment, the relevance between the three-phase output measurement value of the high-voltage transformer and the network topology of the substation is represented by describing the electrical relevance of the three-phase voltage of the node by using a three-phase voltage imbalance VUF, as follows:
wherein V
A、V
B、V
CIs a three-phase voltage at the primary side,
is the average value of the three-phase voltage.
As a possible implementation manner of this embodiment, the three-phase secondary output measurement voltage of the high-voltage potential transformer has the following decomposition formula:
u is a three-phase output measured value, T reflects the common trend of a three-phase output voltage theoretical value, K represents the linear relation between three-phase output voltage and a trend term, M is a three-phase voltage mean value, and delta is a characteristic factor reflecting the voltage error of a three-phase voltage transformer.
As a possible implementation manner of this embodiment, the process of comparing the deviation between the characteristic factors in the current operating condition and the normal operating condition and determining the metering performance of the high-voltage transformer includes the following steps:
1) calculating a Factor Deviation Matrix FDM (FDM): decomposing three-phase measurement secondary output voltage data of the high-voltage transformer by adopting a factor analysis method, and calculating FDM (frequency division multiplexing) representing the current metering performance and the metering performance of the high-voltage transformer under a normal operation condition;
2) calculating a Factor development Coefficient FDC (FDC): calculating an FDM standard reference value, taking a 2-norm of the FDM distance standard reference value as a detection statistic, and defining the detection statistic as a factor deviation coefficient FDC representing the abnormal degree of the metering performance of the high-voltage transformer;
3) and (3) judging an abnormal state: and comparing the FDC with the abnormal state control limit, and judging whether the metering performance of the current high-voltage transformer is in an abnormal state.
As a possible implementation manner of this embodiment, in step 1), the process of calculating the factor deviation matrix FDM is:
selecting a three-phase voltage data set X of the high-voltage transformer under the normal operation conditionNExtracting characteristic factor delta of metering performance of high-voltage transformer by adopting factor analysisN,NThe following formula:
ΔN,N=FN(XN)
wherein, FNRepresents the mapping relation between the three-phase output voltage and the metering performance under the normal condition of the high-voltage transformer, deltaN,NThe first N in the data is used for indicating that the factor is extracted from normal data, and the second N is used for indicating that the mapping relation is obtained by training the normal data;
selecting a three-phase measurement secondary output voltage data set Xt of the high-voltage transformer under the current operation condition, extracting characteristic factors of a test set by using factor analysis, and defining a factor deviation matrix FDM as follows:
FDM=Δt,t-Δt,N=Ft(Xt)-FN(Xt)
and Ft represents the mapping relation between the three-phase output voltage and the metering performance of the high-voltage transformer under the current condition, and the FDM represents the deviation between the metering performance of the high-voltage transformer under the current condition and the normal condition of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of determining the abnormal state specifically includes: judging whether the current FDC exceeds a control limit, if so, judging that the FDC is abnormal and needing maintenance work; if the abnormal state control limit is not exceeded, the current data is supplemented to the historical data, and the control limit is recalculated.
As a possible implementation manner of this embodiment, the process of acquiring the abnormal state control limit includes: and converting the FDC of the skewed distribution into a normal distribution form by using quantile transformation, solving an upper limit value of the normal distribution by using a Lauda criterion, and further performing quantile inverse transformation on the upper limit value to obtain an abnormal state control limit.
On the other hand, the system for detecting the metering performance of the high-voltage transformer provided by the embodiment of the invention comprises:
the characteristic factor extraction module is used for extracting characteristic factors representing the metering performance of the three-phase voltage transformer;
and the performance judging module is used for comparing the deviation of the characteristic factors under the current operation condition and the normal operation condition and judging the metering performance of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of extracting the characteristic factor representing the metering performance of the three-phase voltage transformer includes:
according to the relevance of the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the characteristics of the three-phase measured secondary output voltage of the high-voltage transformer are decomposed, and the fluctuation caused by the primary voltage and the fluctuation caused by the metering performance of the high-voltage transformer are separated from the measured value.
As a possible implementation manner of this embodiment, the relevance of the three-phase output measurement value of the high-voltage transformer to the topological structure of the transformer substation network is represented by the electrical relevance of the three-phase voltage of the node described by the three-phase voltage imbalance VUF, which is as follows:
wherein V
A、V
B、V
CIs a three-phase voltage at the primary side,
is the average value of the three-phase voltage.
As a possible implementation manner of this embodiment, the three-phase secondary output measurement voltage of the high-voltage potential transformer is expressed by the following decomposition formula:
u is a three-phase output measured value, T reflects the common trend of a three-phase output voltage theoretical value, K represents the linear relation between three-phase output voltage and a trend term, M is a three-phase voltage mean value, and delta is a characteristic factor reflecting the voltage error of a three-phase voltage transformer.
As a possible implementation manner of this embodiment, the performance determination module includes:
the FDM calculation module is used for decomposing the three-phase measurement secondary output voltage data of the high-voltage transformer by adopting a factor analysis method and calculating FDM representing the current metering performance and the metering performance of the high-voltage transformer under a normal operation condition;
the FDC calculation module is used for calculating the FDM standard reference value, taking the 2-norm of the FDM distance standard reference value as detection statistic and defining the detection statistic as a factor deviation coefficient FDC representing the abnormal degree of the metering performance of the high-voltage transformer;
and the abnormal state judging module is used for comparing the FDC with the abnormal state control limit and judging whether the metering performance of the current high-voltage transformer is in an abnormal state or not.
As a possible implementation manner of this embodiment, the FDM calculation module is further configured to:
selecting a three-phase voltage data set X of a high-voltage transformer under a normal operation conditionNExtracting characteristic factor delta of metering performance of high-voltage transformer by adopting factor analysisN,NThe following formula:
ΔN,N=FN(XN)
wherein, FNRepresents the mapping relation between the three-phase output voltage and the metering performance under the normal condition of the high-voltage transformer, deltaN,NThe first N in the data is used for indicating that the factor is extracted from normal data, and the second N is used for indicating that the mapping relation is obtained by training the normal data;
selecting a three-phase measurement secondary output voltage data set Xt of the high-voltage transformer under the current operation condition, extracting characteristic factors of a test set by utilizing factor analysis, and defining a factor deviation matrix FDM as follows:
FDM=Δt,t-Δt,N=Ft(Xt)-FN(Xt)
and Ft represents the mapping relation between the three-phase output voltage and the metering performance of the high-voltage transformer under the current condition, and the FDM represents the deviation between the metering performance of the high-voltage transformer under the current condition and the normal condition of the high-voltage transformer.
As a possible implementation manner of this embodiment, the abnormal state determination module is specifically configured to determine whether the current FDC exceeds a control limit, and if the current FDC exceeds the control limit of the abnormal state, it is determined that the current FDC is abnormal, and maintenance work needs to be performed; if the abnormal state control limit is not exceeded, the current data is supplemented to the historical data, and the control limit is recalculated.
As a possible implementation manner of this embodiment, the process of acquiring the abnormal state control limit includes: and converting the FDC of the skewed distribution into a normal distribution form by using quantile transformation, solving an upper limit value of the normal distribution by using a Lauda criterion, and further performing quantile inverse transformation on the upper limit value to obtain an abnormal state control limit.
The technical scheme of the embodiment of the invention has the following beneficial effects:
the invention can master the metering performance state of the high-voltage transformer in real time on the premise of no power failure, has more targeted guidance on the operation and maintenance work of the high-voltage transformer, and has important significance for ensuring the safe, stable and economic operation of a power system. Meanwhile, the related technical route and research method for detecting the metering performance of the high-voltage transformer can be popularized to the research of other types of power transformers, and the method has important reference value for promoting the technical development of the industry.
According to the invention, the high-voltage transmission line does not need to be subjected to power failure maintenance, so that the labor and time cost brought by regular maintenance and the economic loss in the power failure process are reduced; the running state of the high-voltage transformer can be mastered in real time so as to adapt to the measurement performance detection of the high-voltage transformer in different running environments at different periods, guidance is provided for the monitoring and overhauling work of the high-voltage transformer, and the risk of running of an abnormal high-voltage transformer is greatly reduced.
The method can master the metering performance state of the high-voltage transformer in real time so as to adapt to the metering performance detection of the high-voltage transformer in different operating environments at different periods, provide guidance for the monitoring and maintenance work of the high-voltage transformer and greatly reduce the risk of the abnormal high-voltage transformer in operation.
Detailed Description
The invention is further illustrated by the following examples in conjunction with the accompanying drawings:
in order to clearly explain the technical features of the present invention, the following detailed description of the present invention is provided with reference to the accompanying drawings. The following disclosure provides many different embodiments, or examples, for implementing different features of the invention. To simplify the disclosure of the present invention, specific example components and arrangements are described below. Furthermore, the present invention may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and procedures are omitted so as to not unnecessarily limit the invention.
Fig. 1 is a flow chart illustrating a method of power distribution network master device status monitoring in accordance with an exemplary embodiment. As shown in fig. 1, a method for detecting metering performance of a high-voltage transformer provided by an embodiment of the present invention includes the following steps:
extracting characteristic factors representing the metering performance of the three-phase voltage transformer;
and comparing the deviation of the characteristic factors under the current operation condition and the normal operation condition, and judging the metering performance of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of extracting the characteristic factor characterizing the metering performance of the three-phase voltage transformer is as follows:
according to the relevance of the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the characteristics of the three-phase measured secondary output voltage of the high-voltage transformer are decomposed, and the fluctuation caused by the primary voltage and the fluctuation caused by the metering performance of the high-voltage transformer are separated from the measured value.
As a possible implementation manner of this embodiment, the relevance between the three-phase output measurement value of the high-voltage transformer and the network topology of the substation is represented by describing the electrical relevance of the three-phase voltage of the node by using a three-phase voltage imbalance VUF, as follows:
wherein V
A、V
B、V
CIs a three-phase voltage at the primary side,
is the average value of the three-phase voltage.
As a possible implementation manner of this embodiment, the three-phase secondary output measurement voltage of the high-voltage potential transformer has the following decomposition formula:
u is a three-phase output measured value, T reflects the common trend of a three-phase output voltage theoretical value, K represents the linear relation between three-phase output voltage and a trend term, M is a three-phase voltage mean value, and delta is a characteristic factor reflecting the voltage error of a three-phase voltage transformer.
As a possible implementation manner of this embodiment, the process of comparing the deviation between the characteristic factors in the current operating condition and the normal operating condition and determining the metering performance of the high-voltage transformer includes the following steps:
1) calculating a Factor Deviation Matrix FDM (FDM): decomposing three-phase measurement secondary output voltage data of the high-voltage transformer by adopting a Factor Analysis (FA) method, and calculating FDM (frequency division multiplexing) representing the current metering performance and the metering performance under a normal operation condition of the high-voltage transformer;
2) calculating a Factor development Coefficient FDC (FDC): calculating a standard reference value of the FDM, taking a 2-norm of the FDM from the standard reference value as detection statistic, and defining the statistic as a factor deviation coefficient FDC representing the abnormal degree of the metering performance of the high-voltage transformer;
3) and (3) judging an abnormal state: and comparing the FDC with the abnormal state control limit, and judging whether the metering performance of the current high-voltage transformer is in an abnormal state.
As a possible implementation manner of this embodiment, in step 1), the process of calculating the factor deviation matrix FDM is:
selecting a three-phase voltage data set X of the high-voltage transformer under the normal operation conditionNExtracting characteristic factor delta of metering performance of high-voltage transformer by adopting factor analysisN,NThe following formula:
ΔN,N=FN(XN)
wherein, FNRepresents the mapping relation between the three-phase output voltage and the metering performance under the normal condition of the high-voltage transformer, deltaN,NThe first N in the data is used for indicating that the factor is extracted from normal data, and the second N is used for indicating that the mapping relation is obtained by training the normal data;
selecting a three-phase measurement secondary output voltage data set Xt of the high-voltage transformer under the current operation condition, extracting characteristic factors of a test set by utilizing factor analysis, and defining a factor deviation matrix FDM as follows:
FDM=Δt,t-Δt,N=Ft(Xt)-FN(Xt)
and Ft represents the mapping relation between the three-phase output voltage and the metering performance of the high-voltage transformer under the current condition, and the FDM represents the deviation between the metering performance of the high-voltage transformer under the current condition and the normal condition of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of determining the abnormal state specifically includes: judging whether the current FDC exceeds a control limit, if so, judging that the FDC is abnormal, and carrying out maintenance work; if the abnormal state control limit is not exceeded, the current data is supplemented to the historical data, and the control limit is recalculated.
As a possible implementation manner of this embodiment, the process of acquiring the abnormal state control limit includes: and converting the FDC of the skewed distribution into a normal distribution form by using quantile transformation, solving an upper limit value of the normal distribution by using a Lauda criterion, and further performing quantile inverse transformation on the upper limit value to obtain an abnormal state control limit.
The invention can detect the metering performance of the high-voltage transformer by monitoring and analyzing the three-phase measurement secondary output voltage of the high-voltage transformer in real time on the premise of no power failure. The invention relates to a high-voltage transformer, which is characterized in that a secondary voltage measurement output value of the high-voltage transformer is in linear proportional relation with a primary voltage, the secondary voltage measurement output value contains fluctuation caused by load change, active power grid voltage regulation control and other factors and measurement errors of the high-voltage transformer, how to extract the fluctuation caused by the measurement errors of the high-voltage transformer from the fluctuation of a measured value, formulate a reasonable criterion and judge whether the current metering performance of the high-voltage transformer is in a normal state, and is a key problem to be solved by the invention.
According to the relevance between the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the three-phase measurement secondary output voltage of the high-voltage transformer is selected for decomposition, characteristic factors representing the metering performance of the high-voltage transformer are extracted as detection basis, a factor deviation matrix FDM representing the metering performance fluctuation of the high-voltage transformer is decomposed from the voltage signal of the high-voltage transformer, a factor deviation coefficient FDC of the current metering performance of the high-voltage transformer and the metering performance deviation under the normal operation condition is calculated, an abnormal state control limit is established based on the operation data of the high-voltage transformer under the normal operation condition, the metering performance of the current high-voltage transformer is detected, and guidance is provided for monitoring and maintenance work of the high-voltage transformer during operation.
The embodiment of the invention provides a system for detecting the metering performance of a high-voltage transformer, which comprises:
the characteristic factor extraction module is used for extracting characteristic factors representing the metering performance of the three-phase voltage transformer;
and the performance judging module is used for comparing the deviation of the characteristic factors under the current operation condition and the normal operation condition and judging the metering performance of the high-voltage transformer.
As a possible implementation manner of this embodiment, the process of extracting the characteristic factor representing the metering performance of the three-phase voltage transformer includes:
according to the relevance of the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the characteristics of the three-phase measured secondary output voltage of the high-voltage transformer are decomposed, and the fluctuation caused by the primary voltage and the fluctuation caused by the metering performance of the high-voltage transformer are separated from the measured value.
As a possible implementation manner of this embodiment, the relevance of the three-phase output measurement value of the high-voltage transformer to the topological structure of the transformer substation network is represented by the electrical relevance of the three-phase voltage of the node described by the three-phase voltage imbalance VUF, which is as follows:
wherein V
A、V
B、V
CIs a three-phase voltage at the primary side,
is the average value of the three-phase voltage.
As a possible implementation manner of this embodiment, the three-phase secondary output measurement voltage of the high-voltage potential transformer has the following decomposition formula:
u is a three-phase output measured value, T reflects the common trend of a three-phase output voltage theoretical value, K represents the linear relation between three-phase output voltage and a trend term, M is a three-phase voltage mean value, and delta is a characteristic factor reflecting the voltage error of a three-phase voltage transformer.
As a possible implementation manner of this embodiment, the performance determination module includes:
the FDM calculation module is used for decomposing three-phase measurement secondary output voltage data of the high-voltage transformer by adopting a factor analysis method and calculating FDM representing the current metering performance and the metering performance under a normal operation condition of the high-voltage transformer;
the FDC calculation module is used for calculating the FDM standard reference value, taking the 2-norm of the FDM distance standard reference value as detection statistic and defining the detection statistic as a factor deviation coefficient FDC representing the abnormal degree of the metering performance of the high-voltage transformer;
and the abnormal state judging module is used for comparing the FDC with the abnormal state control limit and judging whether the metering performance of the current high-voltage transformer is in an abnormal state or not.
As a possible implementation manner of this embodiment, the FDM calculation module is further configured to:
selecting a three-phase voltage data set X of the high-voltage transformer under the normal operation conditionNExtracting characteristic factor delta of metering performance of high-voltage transformer by adopting factor analysisN,NThe following formula:
ΔN,N=FN(XN)
wherein, FNRepresents the mapping relation between the three-phase output voltage and the metering performance under the normal condition of the high-voltage transformer, deltaN,NThe first N in the data is used for representing that the factor is extracted from normal data, and the second N represents that the mapping relation is obtained by training the normal data;
selecting a three-phase measurement secondary output voltage data set Xt of the high-voltage transformer under the current operation condition, extracting characteristic factors of a test set by utilizing factor analysis, and defining a factor deviation matrix FDM as follows:
FDM=Δt,t-Δt,N=Ft(Xt)-FN(Xt)
and Ft represents the mapping relation between the three-phase output voltage and the metering performance of the high-voltage transformer under the current condition, and the FDM represents the deviation between the metering performance of the high-voltage transformer under the current condition and the normal condition of the high-voltage transformer.
As a possible implementation manner of this embodiment, the abnormal state determination module is specifically configured to determine whether the current FDC exceeds a control limit, and if the current FDC exceeds the control limit of the abnormal state, it is determined that the current FDC is abnormal, and maintenance work needs to be performed; if the abnormal state control limit is not exceeded, the current data is supplemented to the historical data, and the control limit is recalculated.
As a possible implementation manner of this embodiment, the process of acquiring the abnormal state control limit includes: and converting the FDC of the skewed distribution into a normal distribution form by using quantile transformation, solving an upper limit value of the normal distribution by using a Lauda criterion, and further performing quantile inverse transformation on the upper limit value to obtain an abnormal state control limit.
According to the relevance between the measured value of the high-voltage transformer and the network topological structure of the transformer substation, the three-phase measured secondary output voltage of the high-voltage transformer is decomposed, and the characteristic factor representing the metering performance of the high-voltage transformer is extracted to serve as the detection basis.
The relevance of a three-phase output measured value of the high-voltage transformer and a network topological structure of a transformer substation is analyzed, and the electrical and physical relevance of the primary side three-phase voltage is described by adopting three-phase voltage unbalance VUF, wherein the relevance is as follows:
wherein V
A、V
B、V
CIs a three-phase voltage of a primary side,
is the average value of the three-phase voltage.
For a high-voltage system in normal operation, when the unbalance degree of the three-phase voltage is small and constant, the variation trend among the three-phase voltage is kept consistent;
the high-voltage transformer outputs a primary side high-voltage signal as a secondary side small signal, and the output measurement value of the high-voltage transformer comprises a theoretical value fluctuating along with the primary voltage signal and an error value caused by the difference of the metering performance of the high-voltage transformer;
on the premise that the unbalance degree of the three-phase voltage of the high-voltage system is stable, the decomposition formula of the three-phase output measurement voltage of the high-voltage transformer is as follows:
u is a three-phase output measured value, T reflects the common trend of a theoretical value of three-phase output voltage, K represents the linear relation between the three-phase output voltage and a trend item, M is a three-phase voltage mean value, and delta is a characteristic factor reflecting the voltage error of the three-phase voltage transformer, namely the detection basis of the metering performance of the voltage transformer.
The detection flow of the metering performance of the high-voltage transformer is shown in fig. 2, three-phase output voltage of the high-voltage transformer under a normal operation condition is selected as historical data, a factor deviation matrix and corresponding detection statistics are calculated, and a control limit of the detection statistics is formed; and calculating a factor deviation matrix and corresponding detection statistics of the real-time input data, if the detection statistics of the real-time input data exceed the control limit, alarming for an abnormal condition, and if the detection statistics of the real-time input data do not exceed the control limit, supplementing the real-time data to a historical data set to correct the control limit.
The voltage amplitude data of a group of high-voltage transformers (A, B, C three phases) of a certain transformer substation is selected, the total time is about two months, and 5760 data points are selected in each 15 minutes of a group of sampling data. In order to verify the abnormal detection effect of the model, from 2501 points, one voltage data of the B-phase high-voltage transformer is addedStep error of 0.2% and a slope of 10-6The fade error of (2). The method for detecting the metering performance of the high-voltage transformer is adopted for analysis, and the specific implementation steps are described as follows:
1) calculating a factor deviation matrix FDM:
the factor analysis method is adopted to decompose the three-phase measured secondary output voltage data of the high-voltage transformer, and the factor analysis model is shown as the following formula:
Ψ-M=WH+E
where Ψ is the data to be analyzed, H is the latent variable factor, W is the factor loading matrix, and M and E represent the offset and residual.
And taking historical data of the three-phase measured secondary output voltage of the high-voltage transformer as a training set, and recording as X. The factor analysis is used for a three-phase voltage data set X of the high-voltage transformer to obtain a common trend KT (corresponding to WH) representing the voltage fluctuation on the primary side and a characteristic factor delta (corresponding to E) representing the metering performance of the voltage transformer, namely the following formula is established:
Δ=F(X)
selecting a three-phase voltage data set X of the high-voltage transformer under the normal operation conditionNExtracting characteristic factor delta of metering performance of high-voltage transformer by adopting factor analysisN,NThe following formula:
ΔN,N=FN(XN)
wherein, FNThe mapping relation, delta, of the three-phase output voltage and the metering performance under the normal condition of the high-voltage transformer is shownN,NThe first N in (a) indicates that the factor is extracted from normal data, and the second N indicates that the mapping relationship is trained for normal data.
Selecting a three-phase measurement secondary output voltage data set Xt of the high-voltage transformer under the current operation condition, extracting characteristic factors of a test set by utilizing factor analysis, and defining a factor deviation matrix FDM as follows:
FDM=Δt,t-Δt,N=Ft(Xt)-FN(Xt)
and Ft represents the mapping relation between the three-phase output voltage and the metering performance of the high-voltage transformer under the current condition, and the FDM represents the deviation between the metering performance of the high-voltage transformer under the current condition and the normal condition of the high-voltage transformer.
And (3) performing factor analysis on the voltage amplitude of the three-phase high-voltage transformer in the first 20 days, wherein the three-phase voltage of the original sequence is shown in figure 3, and the FDM obtained by the factor analysis is shown in figure 4.
2) Calculating a factor deviation coefficient FDC:
obtaining the FDM standard reference value in the step 1) by adopting cluster analysis, calculating a 2-norm of the FDM distance standard reference value, taking the 2-norm as a detection statistic for further anomaly detection, marking the detection statistic as a factor deviation coefficient FDC, representing the anomaly degree of the metering performance of the high-voltage transformer, and a curve chart of the anomaly degree is shown in fig. 5.
3) And (3) judging an abnormal state:
and comparing the FDC with the abnormal state control limit, and judging whether the metering performance of the current high-voltage transformer is in an abnormal state.
The abnormal state control limit is established as follows: firstly, according to a historical data set X of secondary output voltage of a high-voltage transformer under a normal operation conditionNTraining to obtain the mapping relation F between the output voltage and the characteristic factor under normal conditionNAnd characteristic factor ΔN,N. Historical data set X ofNEqually divided into n subsets, denoted xi(i-1, 2, …, n), then there is a relationship XN={x1,x2,…,xn}. Extracting each subset xiCharacteristic factors of (i ═ 1,2, …, n) are denoted as Δ' ═ { δ1,1,δ2,2,…,δn,nThen the following holds:
from the above equation, the FDM under normal operating conditions can be obtained as follows:
FDM=Δ′-ΔN,N
and (3) repeating the step 2) to calculate the FDC corresponding to the FDM under the normal operation condition, and converting the FDC with the skewed distribution into a normal distribution form by using Quantile Transform (Quantile Transform), wherein the FDC under the normal operation condition is distributed as shown in FIG. 6, and the FDC subjected to the Quantile Transform is distributed as shown in FIG. 7. The upper limit value of the normal distribution is obtained by using the raydeta Criterion (Pauta Criterion), and the upper limit value is further subjected to quantile inverse transformation to obtain the abnormal state control limit, as shown in fig. 8, which is an FDC graph added with the control limit.
Further, the method for correcting the abnormal state control limit comprises the following steps: judging whether the current FDC exceeds a control limit, if so, judging that the FDC is abnormal, and carrying out maintenance work; if the control limit is not exceeded, the current data is supplemented to the historical data and the control limit is recalculated.
Over time, when a step error of 0.2% is added, the model can immediately detect an anomaly, as shown in fig. 9, beginning at 2501 points, with FDC exceeding the abnormal state control limit; when the addition slope is 10-6When the error is gradual, the model detects an abnormality at about 3500, as shown in fig. 10.
According to the invention, the detection statistic is established through the data of the high-voltage transformer under the normal operation condition, the metering performance of the high-voltage transformer is detected, the power-off maintenance of the high-voltage transmission line is not needed, and the manpower and time cost and the economic loss in the power-off process caused by the regular maintenance are reduced.
The above description is only a preferred embodiment of the present invention, and it will be apparent to those skilled in the art that various modifications and improvements can be made without departing from the principle of the present invention, and these modifications and improvements are also considered to be within the scope of the present invention.