CVT online monitoring method based on big data driving

Through the CVT online monitoring method driven by big data, using group algorithms and PCA technology, the problem of untimely detection and misjudgment of CVT online monitoring is solved, and high-precision voltage transformer error detection is realized, which improves the operation of the power system. reliability.

CN119001578APending Publication Date: 2024-11-22STATE GRID HUNAN ELECTRIC POWER COMPANY LIMITED +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410984512.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing online CVT monitoring methods have problems such as untimely detection, long power outage, high cost and inconvenient arrangement, which affects the normal operation of the power system.

Method used

The CVT online monitoring method based on big data is adopted, and the errors of voltage transformers are detected in real time through group algorithms and PCA technology to improve monitoring accuracy. The group algorithm treats multiple sets of voltage transformers as a group, calculates the degree of deviation, and avoids misjudgment based on the error caused by PCA separation voltage imbalance and error caused by abnormal measurement device.

Benefits of technology

It realizes high accuracy of CVT online monitoring, is suitable for a variety of scenarios, reduces misjudgment problems, and improves the normal operation reliability of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119001578A_ABST
    Figure CN119001578A_ABST
Patent Text Reader

Abstract

The invention discloses a CVT online monitoring method based on big data driving, and the method comprises the steps: S1, judging whether there are multiple groups of voltage transformers at the same voltage level in a transformer substation or a power plant, and if there are multiple groups of voltage transformers at the same voltage level, employing a CVT online monitoring method based on a group algorithm; and S2, if only a single group of voltage transformers exist under the same voltage grade, adopting a PCA-based CVT online monitoring method. The method has the advantages of being simple in principle, wide in application range, capable of greatly improving the online monitoring accuracy and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention mainly relates to the technical field of online monitoring of capacitive voltage transformers in power equipment, and in particular to a CVT online monitoring method driven by big data. Background Art

[0002] Accurate measurement of electric energy is the basis for ensuring fair electricity settlement. Every year, the amount of electricity error caused by voltage transformer error is considerable. Compared with electromagnetic voltage transformer (PT), capacitive voltage transformer (CVT) has the advantages of high insulation strength, no ferromagnetic resonance, and relatively low price. CVT is widely used for voltage measurement in gateway metering of 220kV and above voltage level substations (power plants).

[0003] The existing CVT is mainly composed of a capacitor voltage divider, an intermediate transformer and a compensating reactor. Its structural characteristics make its own error characteristics easily affected by the external environment (such as temperature, frequency, and surrounding electric field). At present, the traditional CVT error monitoring method is mainly periodic on-site power outage detection, but there are problems such as "untimely detection, long power outage time, and high detection cost".

[0004] Some practitioners have also proposed a more novel technical means: CVT remote online monitoring device, which can remotely and in real time detect the errors of the metering device, but it has problems such as high cost, inconvenient layout, and potential faults. It may also affect the normal operation of the power system. Therefore, it is urgent to promote the transformation of CVT from "periodic detection" to "online monitoring". Summary of the invention

[0005] The technical problem to be solved by the present invention is: in response to the technical problems existing in the prior art, the present invention provides a CVT online monitoring method based on big data drive which has a simple principle, a wide range of applications and can greatly improve the online monitoring accuracy.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions: A CVT online monitoring method based on big data drive, comprising: Step S1: Determine whether there are multiple sets of voltage transformers at the same voltage level in the substation or power plant. If there are multiple sets of voltage transformers at the same voltage level, use a CVT online monitoring method based on a group algorithm; Step S2: If there is only a single set of voltage transformers at the same voltage level, a PCA-based CVT online monitoring method is used.

[0007] As a further improvement of the method of the present invention: the step S1 comprises: Step S101: select n groups of voltage transformers in the substation and connected to the bus or line of the same voltage level as a group, and the same voltage transformers as a group, respectively. , express , and so on; obtain the three-phase voltage amplitude of 96 points of the voltage transformer group in one day from the acquisition system, The voltage is recorded as , and so on for the other items; Step S102: Calculate the three-phase voltage ratio difference of the transformer group to obtain the ratio difference sequence of the transformer group ; Step S103: Calculate the average value of the mutual inductor group ratio difference sequence , Standard Deviation ; Step S104: Based on the average value Calculate the ratio deviation sequence of the transformer group: ; Step S105: The absolute value of the element in the statistical ratio deviation sequence falls into The number of intervals, the absolute value of the ratio deviation sequence element is considered to be less than No abnormality; greater than Less than 2 It is a warning state; greater than 3 It is an abnormal state; Step S106: According to the normal distribution law, the probability of abnormality or warning of each phase of the transformer group is calculated; with the date as the horizontal axis and the error probability as the vertical axis, the error probability curve of each transformer is obtained; the abnormal transformer is identified according to the trend of the error probability curve, and then the CVT online monitoring based on the group algorithm is realized.

[0008] As a further improvement of the method of the present invention: the step S102 includes the following process: Step S1021: , For transformer The difference sequence of elements; among them, the element Indicates transformer With transformer During the day The ratio difference at the moment; and so on, the ratio difference sequences of the other transformers are elements, thus obtaining Only the mutual inductor ratio difference sequence;

[0009]

[0010] in, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment; Step S1022: By analogy, the ratio difference sequences of the other transformers are elements, thus obtaining Transformer ratio difference matrix : .

[0011] As a further improvement of the method of the present invention: in step S103, according to the ratio difference matrix Calculate the average value of all elements in the day , Standard Deviation :

[0012]

[0013] in, Represents the difference matrix Middle Line Elements of a column.

[0014] As a further improvement of the method of the present invention: the step S104 includes: , For transformer The ratio deviation sequence is elements; and so on, the ratio deviation sequences of the other transformers are elements, thus obtaining Transformer ratio deviation matrix :

[0015]

[0016]

[0017] in, Indicates transformer With transformer During the day The time difference deviation, Indicates transformer With transformer During the day The ratio deviation of time.

[0018] As a further improvement of the method of the present invention: the step S2 includes the following process: Step S201: Acquire 96-point three-phase voltage data for n days (n>30) during the normal operation of the voltage transformer; Step S202: Calculate the voltage imbalance at each moment according to the following formula to obtain sample data: , ;

[0019] in, express The maximum value of the three-phase voltage of A, B, and C at the moment, express The minimum value of the three-phase voltage of A, B, and C at the moment; Step S203: Based on sample data , establish the principal element model of three-phase unbalance of voltage transformer; Step S204: Calculate and determine the statistical threshold ; Step S205: Calculate the voltage transformer operation process according to the established principal element model. Statistics and statistical thresholds obtained under normal operating conditions Comparison is then made to achieve online monitoring of a single CVT.

[0020] As a further improvement of the method of the present invention: the step S203 includes: Step S2031: Sample data After standardization, the data matrix obtained is shown in formula (8);

[0021] in, ; , is the sample data The mean vector of ; , is the covariance matrix of the sample data; Step S2032: Standardize the sample data To break it down:

[0022] in, ,for The principal component space model, is the score matrix on the principal space, and is the loading matrix on the principal space; for The residual space model of is the score matrix in the residual space, is the loading matrix in the residual space ; The principal subspace is the space where the true value is located, and the residual space is the space where the noise is located; The covariance matrix of Perform eigenvalue decomposition to obtain the load matrix , ;

[0023] in, for The characteristic value of is the load vector.

[0024] As a further improvement of the method of the present invention: the step S204 includes: By establishing statistics for hypothesis testing, we can determine whether the measurement data contains abnormal information during operation and whether the process data deviates from the principal component model. We can establish Q statistics in the residual subspace for statistical testing and use Q statistics to evaluate the metering status of CVT. The calculation of is as follows:

[0025] in, , , At the confidence level Critical value of the lower normal distribution.

[0026] As a further improvement of the method of the present invention: the step S205 includes: Step S2051: Calculate the voltage transformer operation process according to the established principal component model. Statistics, the expression is as follows:

[0027] in, It is the standardized voltage imbalance sample data during operation; Step S2052: Comparison and The size of Less than , the voltage transformer is in normal state; if Greater than , an abnormal alarm is generated.

[0028] Compared with the prior art, the advantages of the present invention are: The CVT online monitoring method based on big data driving of the present invention has a simple principle, a wide range of applications, and can greatly improve the accuracy of online monitoring. The present invention is aimed at power plants with multiple groups of voltage transformers, and innovatively proposes a CVT online monitoring method based on a group algorithm. All voltage transformers in the station are regarded as a group, and the degree of deviation of a certain phase in the group is calculated, which can effectively avoid the misjudgment problem caused by primary voltage imbalance. At the same time, the present invention is aimed at power plants with only one group of voltage transformers, and innovatively proposes a CVT online monitoring method based on PCA. This method separates the normal error caused by the primary voltage from the metering error caused by the abnormality of the metering device, which can effectively avoid the misjudgment problem caused by the primary voltage imbalance. Finally, the present invention can be applied to actual needs in a variety of scenarios to improve the accuracy of online monitoring as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the process of the present invention in a specific embodiment. DETAILED DESCRIPTION

[0030] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] As the electricity information collection system has been built for many years, it has covered all electricity users (including power generation users) and various checkpoints, bringing a large amount of electric energy data, which can provide support for the analysis and decision-making of various links in the operation and management of power companies, and provide an information basis for the realization of intelligent two-way interactive services. These massive data resources can be deeply analyzed through appropriate mathematical models, so that the CVT error information contained in the data can be integrated and clearly presented, which is a new way to realize CVT online monitoring.

[0032] For this reason, Figure 1 As shown, the present invention provides a CVT online monitoring method based on big data drive, and the process includes: Step S1: Determine whether there are multiple sets of voltage transformers at the same voltage level in the substation or power plant. If there are multiple sets of voltage transformers at the same voltage level, use a CVT online monitoring method based on a group algorithm; Step S2: If there is only a single set of voltage transformers at the same voltage level, a PCA-based CVT online monitoring method is used.

[0033] In a specific application example, step S1 may include the following process according to actual needs: Step S101: select n groups of voltage transformers in the substation and connected to the bus or line of the same voltage level as a group, and the same voltage transformers as a group, respectively. , express , and so on; obtain the three-phase voltage amplitude of 96 points of the voltage transformer group in one day from the acquisition system, The voltage is recorded as , and so on for the other items; Step S102: Calculate the three-phase voltage ratio difference of the transformer group to obtain the ratio difference sequence of the transformer group ; Step S103: Calculate the average value of the mutual inductor group ratio difference sequence ( ), standard deviation ( ); Step S104: According to the average value ( ) Calculate the ratio deviation sequence of the mutual inductor group:

[0034] Step S105: The absolute value of the element in the statistical ratio deviation sequence falls into The number of intervals, the absolute value of the ratio deviation sequence element is considered to be less than No abnormality; greater than Less than 2 It is a warning state; greater than 3 It is an abnormal state; Step S106: According to the normal distribution law, the probability of abnormality or warning of each phase of the transformer group is calculated; with the date as the horizontal axis and the error probability as the vertical axis, the error probability curve of each transformer is obtained; the abnormal transformer is identified according to the trend of the error probability curve, and then the CVT online monitoring based on the group algorithm is realized.

[0035] As a preferred embodiment, the step S102 may include the following process according to actual needs: Step S1021: , For transformer The difference sequence of elements. Among them, the element Indicates transformer With transformer During the day By analogy, the ratio difference sequences of the other transformers are elements, thus obtaining Only the mutual inductor ratio difference sequence.

[0036] (1) (2) in, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment.

[0037] Step S1022: By analogy, the ratio difference sequences of the other transformers are elements, thus obtaining Transformer ratio difference matrix : .

[0038] As a preferred embodiment, the step S103 may include, according to actual needs: Calculate the average value of all elements in the day , Standard Deviation : (3) (4) in, Represents the difference matrix Middle Line Elements of a column.

[0039] As a preferred embodiment, the step S104 may include, according to actual needs: , For transformer The ratio deviation sequence is elements. By analogy, the ratio deviation sequences of the other transformers are elements, thus obtaining Transformer ratio deviation matrix :

[0040] (5) (6) in, Indicates transformer With transformer During the day The time difference deviation, Indicates transformer With transformer During the day The ratio deviation of time.

[0041] In a specific application example, step S2 may include the following process according to actual needs: Step S201: Acquire 96-point three-phase voltage data for n days (n>30) during the normal operation of the voltage transformer; Step S202: Calculate the voltage imbalance at each moment according to the following formula (7) to obtain sample data , ; (7) in, express The maximum value of the three-phase voltage of A, B, and C at the moment, express The minimum value of the three-phase voltage of A, B, and C at the moment; Step S203: Based on sample data , establish the principal element model of three-phase unbalance of voltage transformer; Step S204: Calculate and determine the statistical threshold ; Step S205: Calculate the voltage transformer operation process according to the established principal element model. Statistics and statistical thresholds obtained under normal operating conditions Comparison is then made to achieve online monitoring of a single CVT.

[0042] As a preferred embodiment, the step S203 may include, according to actual needs: Step S2031: Sample data After standardization, the data matrix obtained is shown in formula (8).

[0043] (8) in, ; , is the sample data The mean vector of ; , is the covariance matrix of the sample data.

[0044] Step S2032: Standardize the sample data To break it down: (9) in, ,for The principal component space model, is the score matrix on the principal space, and is the loading matrix on the principal space; for The residual space model of is the score matrix in the residual space, is the loading matrix in the residual space ; The principal subspace is the space where the true value is located, and the residual space is the space where the noise is located. The covariance matrix of Perform eigenvalue decomposition to obtain the load matrix , .

[0045] (10) in, for The characteristic value of is the load vector.

[0046] As a preferred embodiment, in step S204, whether the measured data contains abnormal information during operation can be tested by establishing statistics to determine whether the process data deviates from the principal component model. A Q statistic is established in the residual subspace for statistical testing. Since the metering error of the three-phase CVT is mainly projected in the residual space, the Q statistic is used to evaluate the metering state of the CVT. Statistical threshold The calculation of is as follows: (11) in, , , At the confidence level Critical value of the lower normal distribution.

[0047] As a preferred embodiment, the step S205 may include the following process according to actual needs: Step S2051: Calculate the voltage transformer operation process according to the established principal component model. Statistics, the expression is as follows: (12) in, It is the standardized voltage imbalance sample data during operation.

[0048] Step S2052: Comparison and The size of Less than , the voltage transformer is in normal state; if Greater than , an abnormal alarm is generated.

[0049] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A CVT online monitoring method based on big data drive, characterized in that: include: Step S1: Determine whether there are multiple sets of voltage transformers at the same voltage level in the substation or power plant. If there are multiple sets of voltage transformers at the same voltage level, use a CVT online monitoring method based on a group algorithm; Step S2: If there is only a single set of voltage transformers at the same voltage level, a PCA-based CVT online monitoring method is used.

2. The CVT online monitoring method based on big data drive according to claim 1 is characterized in that: The step S1 comprises: Step S101: select n groups of voltage transformers in the substation and connected to the bus or line of the same voltage level as a group, and the same voltage transformers as a group, respectively. , express , and so on; obtain the three-phase voltage amplitude of 96 points of the voltage transformer group in one day from the acquisition system, The voltage is recorded as , and so on for the other items; Step S102: Calculate the three-phase voltage difference of the transformer group to obtain the difference sequence of the transformer group. ; Step S103: Calculate the average value of the mutual inductor group ratio difference sequence , Standard Deviation ; Step S104: Based on the average value Calculate the ratio deviation sequence of the transformer group: ; Step S105: The absolute value of the element in the statistical ratio deviation sequence falls into The number of intervals, the absolute value of the ratio deviation sequence element is considered to be less than No abnormality; greater than Less than 2 It is a warning state; greater than 3 It is an abnormal state; Step S106: According to the normal distribution law, the probability of abnormality or warning of each phase of the transformer group is calculated; with the date as the horizontal axis and the error probability as the vertical axis, the error probability curve of each transformer is obtained; the abnormal transformer is identified according to the trend of the error probability curve, and then the CVT online monitoring based on the group algorithm is realized.

3. The CVT online monitoring method based on big data drive according to claim 2 is characterized in that: The step S102 includes the following process: Step S1021: , For transformer The difference sequence of elements; among them, the element Indicates transformer With transformer During the day The ratio difference at the moment; and so on, the ratio difference sequences of the other transformers are elements, thus obtaining Only the mutual inductor ratio difference sequence; in, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment, Indicates transformer In a day The voltage amplitude at the moment; Step S1022: By analogy, the ratio difference sequences of the other transformers are elements, thus obtaining Transformer ratio difference matrix : 。 4. The CVT online monitoring method based on big data drive according to claim 2 is characterized in that: In step S103, according to the ratio difference matrix Calculate the average value of all elements in the day , Standard Deviation : in, Represents the difference matrix Middle Line Elements of a column.

5. The CVT online monitoring method based on big data drive according to claim 2 is characterized in that: The step S104 includes: , For transformer The ratio deviation sequence is elements; and so on, the ratio deviation sequences of the other transformers are elements, thus obtaining Transformer ratio deviation matrix : in, Indicates transformer With transformer During the day The time difference deviation, Indicates transformer With transformer During the day The ratio deviation of time.

6. The CVT online monitoring method based on big data drive according to any one of claims 1 to 5, characterized in that: The step S2 includes the following process: Step S201: Acquire 96-point three-phase voltage data for n days (n>30) during the normal operation of the voltage transformer; Step S202: Calculate the voltage imbalance at each moment according to the following formula to obtain sample data: , ; in, express The maximum value of the three-phase voltage of A, B, and C at the moment, express The minimum value of the three-phase voltage of A, B, and C at the moment; Step S203: Based on sample data , establish the principal element model of three-phase unbalance of voltage transformer; Step S204: Calculate and determine the statistical threshold ; Step S205: Calculate the voltage transformer operation process according to the established principal element model. Statistics and statistical thresholds obtained under normal operating conditions Comparison is then made to achieve online monitoring of a single CVT.

7. The CVT online monitoring method based on big data drive according to claim 6 is characterized in that: The step S203 includes: Step S2031: Sample data After standardization, the data matrix obtained is shown in the following formula: in, ; , is the sample data The mean vector of ; , is the covariance matrix of the sample data; Step S2032: Standardize the sample data To break it down: in, ,for The principal component space model, is the score matrix on the principal space, and is the loading matrix on the principal space; for The residual space model of is the score matrix in the residual space, is the loading matrix in the residual space ; The principal subspace is the space where the true value is located, and the residual space is the space where the noise is located; The covariance matrix of Perform eigenvalue decomposition to obtain the load matrix , ; in, for The characteristic value of is the load vector.

8. The CVT online monitoring method based on big data drive according to claim 6 is characterized in that: The step S204 includes: By establishing statistics for hypothesis testing, we can determine whether the measurement data contains abnormal information during operation and whether the process data deviates from the principal component model. We can establish Q statistics in the residual subspace for statistical testing and use Q statistics to evaluate the metering status of CVT. The calculation of is as follows: in, , , At the confidence level Critical value of the lower normal distribution.

9. The CVT online monitoring method based on big data drive according to claim 6 is characterized in that: The step S205 includes: Step S2051: Calculate the voltage transformer operation process according to the established principal component model. Statistics, the expression is as follows: in, It is the standardized voltage imbalance sample data during operation; Step S2052: Comparison and The size of Less than , the voltage transformer is in normal state; if Greater than , an abnormal alarm is generated.