KR20220011047A - Method for determination of status of oil immersed type transformer and apparatus for using the method - Google Patents

Abstract

The present invention relates to a method for determining a transformer state based on data correction of gas in oil, and a device using the same. The method comprises: a step of enabling a transformer state determination device to receive data of gas in oil; a step of enabling the transformer state determination device to determine a measurement error value which is a correction target on the data of the gas in oil; a step of enabling the transformer state determination device to correct the measurement error value; and a step of enabling the transformer state determination device to determine a transformer state based on the data of the gas in oil including the corrected measurement error value.

Description

Transformer status determination method based on gas-in-oil data correction and apparatus using such a method

The present invention relates to a method for determining the state of a transformer based on gas-in-oil data correction and to an apparatus using the method. More particularly, it relates to a method and an apparatus for more accurately determining the state of a transformer by performing correction for a measurement error value among gas-in-oil data.

The use of power transformers has increased due to a surge in demand for electrical energy due to rapid industrial development. As a result, many of the transformers currently installed are aging, and unexpected facility accidents often occur. As power transformers become large-capacity and power systems become more complex, if an accident occurs due to facility failure, extensive power outages are accompanied, and economic losses due to restoration and supply disruptions increase.

To minimize these losses, the current state of the transformer must be diagnosed as accurately as possible. It is necessary to carry out the necessary operational maintenance to minimize the unexpected accidents of the transformer.

The case that accounts for the largest proportion of accidents in transformers is deterioration of dielectric strength. Transformer insulation breakdown can be accompanied by explosion in nature. Dissolved Gas Analysis (DGA) is mainly used as the most effective method for analyzing insulation degradation characteristics. Organic insulating materials such as insulating oil and insulating paper used in transformers cause temperature rise and local overheating due to operation.

In addition, degradation products including various gases are formed by thermal decomposition by electric discharge or the like. Gas among the degradation products is dissolved in insulating oil (oil). Therefore, if the transformer insulating oil is regularly sampled during operation and the dissolved gas concentration is analyzed, it can be estimated whether there is an abnormality inside the transformer. However, if the condition of the transformer is simply determined by whether or not the standard value for a specific gas is exceeded or a pattern, it is difficult to accurately diagnose the operation, maintenance, or replacement of the transformer.

Therefore, there is a need for a method for diagnosing the condition more reliably than the present along with the cause of the abnormal occurrence of the transformer.

An object of the present invention is to solve all of the above problems.

In addition, an object of the present invention is to more accurately diagnose the state of the transformer by correcting the gas-in-oil data.

In addition, an object of the present invention is to more accurately diagnose the state of the transformer by determining whether the transformer has oil filtration.

A representative configuration of the present invention for achieving the above object is as follows.

According to an embodiment of the present invention, the method for determining the state of a transformer based on gas-in-oil data correction comprises the steps of: a transformer state determination device receiving gas-in-oil data; Determining a value, the transformer state determining device correcting the measurement error value and the transformer state determining device determining the transformer state based on the gas-in-oil data including the corrected measurement error value can do.

Meanwhile, the determining of the measurement error value may include determining, by the transformer state determination device, the measurement error value in consideration of whether oil is filtered based on the gas-in-oil data.

In addition, the step of correcting the measurement error value may include correcting the measurement error value by the transformer state determination device based on a similar data set similar to a data set including the measurement error value.

According to another embodiment of the present invention, a transformer state determination device for determining a transformer state based on gas-in-oil data correction is operable with a gas-in-oil data input unit and the gas-in-oil data input unit implemented to receive gas-in-oil data ( operatively) a coupled processor, wherein the processor receives gas-in-oil data, determines a measurement error value to be corrected on the gas-in-oil data, corrects the measurement error value, and includes a corrected measurement error value. The state of the transformer may be determined based on the data.

Meanwhile, the processor may be implemented to determine the measurement error value in consideration of whether oil is filtered based on the gas-in-oil data.

Also, the processor may correct the measurement error value based on a similar data set similar to the data set including the measurement error value.

According to the present invention, the state of the transformer can be more accurately diagnosed by correcting the gas-in-oil data.

In addition, according to the present invention, the state of the transformer can be more accurately diagnosed based on the determination of whether the transformer is oil filtered.

1 is a conceptual diagram illustrating a measurement error during analysis of a conventional gas-in-oil.
2 is a conceptual diagram illustrating a transformer state determination device for determining a transformer state based on correction for gas-in-oil data according to an embodiment of the present invention.
3 is a conceptual diagram illustrating a method for determining whether oil is filtered according to an embodiment of the present invention.
4 is a conceptual diagram illustrating a method for performing correction on gas-in-oil measurement data according to an embodiment of the present invention.
5 is a conceptual diagram illustrating a method of determining a measurement error value (correction target) based on whether oil is filtered according to an embodiment of the present invention.
6 is a conceptual diagram illustrating a gas-in-oil correction method according to an embodiment of the present invention.
7 is a conceptual diagram illustrating a second correction procedure for performing statistical-based measurement error correction according to an embodiment of the present invention.
8 is a conceptual diagram illustrating a method for performing correction on a measurement error value (correction target) according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0012] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] Reference is made to the accompanying drawings, which show by way of illustration specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different but need not be mutually exclusive. For example, certain shapes, structures, and characteristics described herein may be implemented with changes from one embodiment to another without departing from the spirit and scope of the present invention. In addition, it should be understood that the location or arrangement of individual components within each embodiment may be changed without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention should be taken as encompassing the scope of the claims and all equivalents thereto. In the drawings, like reference numerals refer to the same or similar elements throughout the various aspects.

Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings in order to enable those of ordinary skill in the art to easily practice the present invention.

The state of the existing transformer is determined through periodic oil (or insulating oil) sampling and Dissovled Gas Analysis (DGA) analysis in oil (or insulating oil). Specifically, the current state of the transformer is diagnosed by comparing the gas-in-oil data collected and analyzed from the transformer with the rule-based international standard IEEE/IEC/JAPAN ETRA/CIGRE.

Specifically, a transformer is a static type device that transfers electrical energy between two or more circuits through an inductive electrical conductor, and serves to change the voltage of an alternating current for the purpose of application. In particular, transformers are essential for production and development facilities that require constant, specific power. Failure of a transformer causes enormous financial and personal injury. Accordingly, the development of diagnostic technology for securing reliability and preventive maintenance of transformers is being actively carried out.

Transformers mainly used in transformers have advantages such as excellent insulation and cooling performance, low manufacturing cost, etc. by filling the inside with oil (or insulating oil), which is an insulating medium. Transformers generate abnormal gases and degradation products due to overheating due to electrical and/or mechanical abnormalities, and deterioration of oil or solid insulation materials. The generated abnormal gas and deterioration product may be dissolved in the oil to change the gas-in-oil component and the amount of gas in the oil. Gas-in-oil components and gas quantities in oil can be analyzed using gas chromatography.

For example, the condition diagnosis (or fault diagnosis) of the transformer based on the IEEE Std C57.104-2008 standard for diagnosing the condition of a transformer using the amount of gas in oil, the standard for diagnosing the condition of a transformer using the gas-in-oil composition ratio of IEEE Std C57.104-2008, etc. ) can be done.

However, depending on circumstances, data on gas in oil may not be accurately measured, and in this case, it is difficult to accurately diagnose the condition of the transformer.

1 is a conceptual diagram illustrating a measurement error during analysis of a conventional gas-in-oil.

In FIG. 1, an error in measuring the amount of gas in oil and H _{2 gas of oil periodically measured is disclosed.}

Referring to FIG. 1 , oil is filled in an enclosed space due to the nature of the structure of the transformer, and the generated gas in oil gradually increases in the oil over time. However, when the insulating oil is exposed to air or gas chromatography is not performed properly in the process of collecting the insulating oil of the transformer, a measurement error may be issued as shown in FIG. 1 .

In other words, since the insulating oil exists in a closed space, only a specific gas cannot rapidly drop from the unfiltered insulating oil over time. Referring to the measurement result of FIG. 1 , among the periodically measured gas-in-oil data, H ₂ gas was measured as a value that suddenly decreased at a specific point in time. When the amount of gas in the insulating oil is rapidly reduced compared to the existing measurement value, it may be determined as a measurement error.

Measurement errors cause errors in judgment about the state of the transformer. Therefore, a technique capable of correcting a measurement error is required, but such a technique has not been developed at present. Therefore, in the method for determining the state of a transformer based on correction of gas-in-oil data according to an embodiment of the present invention, a measurement error value for a specific gas is determined from periodically measured gas-in-oil data, and based on a statistical-based missing value correction model Disclosed is a method for reducing erroneous diagnosis of the state of a transformer by compensating for a measurement error value.

2 is a conceptual diagram illustrating a transformer state determination device for determining a transformer state based on correction for gas-in-oil data according to an embodiment of the present invention.

2, the transformer state determination device is a gas-in-oil data input unit 200, an oil filtration determination unit 210, a measurement error value determination unit 220, a gas-in-oil data correction unit 230, a transformer condition determination unit ( 240 ) and a processor 250 .

The gas-in-oil data input unit 200 may be implemented to receive gas-in-oil data. The gas-in-oil data may include information about the 6-type gas-in-oil and/or the 6-type gas-in-oil composition ratio information measured by the transformer.

Information on 6 types of gas in oil includes information on hydrogen H ₂ , methane CH ₄ , ethylene C ₂ H ₄ , ethane C ₂ H ₆ , acetylene C ₂ H ₂ , and carbon monoxide CO, and information on the composition ratio of 6 types of gas in oil is specific It can include information about gas-in-oil values versus the sum of six gas-in-oils.

The gas-in-oil data corrected in the present invention may be information on the amount of six types of gas-in-oil included in the insulating oil (or oil).

The oil filtration determination unit 210 may be implemented to determine whether filtration of the oil of the transformer has been performed based on the gas-in-oil data. The oil filtration determining unit 210 may determine whether to filter the oil in order to determine whether to correct the gas-in-oil measurement data.

When a specific gas-in-oil deviates from the normal range (for example, a sudden rise) or as a periodic maintenance, oil filtration can be performed to remove foreign substances, moisture, gas, etc. from the oil inside the entire transformer. Oil filtration can improve insulation performance and improve the condition of the transformer to some extent.

Specifically, when an internal failure occurs in a transformer, heat is generated, and the insulating oil in contact with heat is thermally decomposed, and hydrogen (H ₂ ), methane (CH ₄ ), acetylene (C ₂ H ₂ ), ethylene (C ₂ H ₄ ), A gas such as ethane (C ₂ H _{6 ) is generated.} And in the cellulosic insulating paper, methane (CH ₄ ), hydrogen (H ₂ ), carbon monoxide (CO), carbon dioxide (CO ₂ ), etc. are generated. Therefore, a specific gas is generated pattern according to the failure type inside the transformer, and the determination of the failure type of the transformer can be performed through chromatographic analysis of these main gases.

When the oil filtration operation is performed, the continuous condition deterioration pattern of the transformer is interrupted in terms of the condition prediction of the transformer. That is, the state prediction result is different according to the determination of whether oil is filtered in the transformer. If it is not determined whether the oil is filtered, it becomes a fatal problem for accurate judgment of the state of the transformer. Therefore, it is necessary to determine whether the oil filtration operation has been performed in the transformer to determine the state of the transformer.

The gas-in-oil for determining whether to perform such an oil filtration operation may be defined as an oil filtration determination gas. For example, the oil filtration determination gas may be three types of gas in oil (methane CH ₄ , ethylene C ₂ H ₄ , ethane C ₂ H ₆ ). The oil filtration determining unit may determine whether to filter the oil in consideration of a change in the oil filtration determination gas. For example, if an _{oil filtration judge gas such as methane (CH 4} ), ethylene (C ₂ H ₄ ), or ethane (C ₂ H ₆ ) tri-gas decreases to a critical percentage (60% or less) after filtration, the oil It can be judged that filtration has been performed.

The measurement error value determination unit 220 may be implemented to determine a measurement error value among the gas-in-oil data. The measurement error value may include non-measured values such as missing values, and may include abnormal range values such as values that deviate from the normal range in comparison with the existing measurement values for each gas. For convenience of explanation, a missing value is described as an example of a measurement error value, but may be a value outside the normal range determined based on a learning model for each gas as well as a missing value. For example, if there is no oil filtration, the gas in oil in the insulating oil gradually increases over time due to the structural characteristics of the transformer. have.

The gas-in-oil data correction unit 230 may be implemented to correct the gas-in-oil data. The gas-in-oil data correction unit 230 may perform correction on the measurement error value determined by the measurement error value determination unit.

The gas-in-oil data correction unit 230 may determine whether the measurement error value can be corrected, and correct the measurement error value in consideration of whether oil is filtered, whether there is a measurement error value around the measurement error value, and the like. A measurement error value that can be corrected among measurement error values may be expressed in terms of a measurement error value (correction target).

First, a measurement error value corresponding to a time of oil filtration at which oil filtration is performed is not corrected based on whether oil filtration is performed, and a measurement error value corresponding to a time point other than the time of oil filtration may be corrected. Next, the gas-in-oil data correction unit 230 may correct the measurement error value in consideration of whether a previous measurement value and a subsequent measurement value (or a subsequent measurement value) exist based on the measurement error value.

The gas-in-oil data correction unit 230 may perform a correction procedure based on the surrounding measurement value for the correctable measurement error value (correction target). The gas-in-oil data correction unit 230 may perform a first correction procedure and/or a second correction procedure, which will be described in detail later.

Transformer state determination unit 240 may determine the state of the transformer based on the gas-in-oil data corrected by the gas-in-oil data correction unit 230 . For example, the transformer state may be determined as one of Normal, Warning, Critical, and Fault.

The processor 250 operates the gas-in-oil data input unit 200 , the oil filtration determination unit 210 , the measurement error value determination unit 220 , the gas-in-oil data correction unit 230 and/or the transformer state determination unit 240 . can be implemented to control

The processor 250 operates with the gas-in-oil data input unit 200 , the oil filtration determination unit 210 , the measurement error value determination unit 220 , the gas-in-oil data correction unit 230 and/or the transformer state determination unit 240 . may be operatively linked.

In the present invention, the transformer state determination method based on gas-in-oil data correction includes the steps of: the transformer state determination device receives gas-in-oil data; the transformer state determination device determines the measurement error value to be corrected on the gas-in-oil data; The determination device may include correcting the measurement error value and the transformer condition determination device determining the transformer state based on gas-in-oil data including the corrected measurement error value.

The determining of the measurement error value may include determining, by the transformer state determination device, the measurement error value in consideration of whether oil is filtered based on the gas-in-oil data. In addition, the step of correcting the measurement error value may include the step of correcting the measurement error value based on a similar data set similar to the data set including the measurement error value by the transformer state determination device.

Hereinafter, a method for determining the state of a transformer based on detailed gas-in-oil data correction is disclosed.

3 is a conceptual diagram illustrating a method for determining whether oil is filtered according to an embodiment of the present invention.

In FIG. 3 , a method for determining an oil filtration determination gas for determining whether or not oil is filtered by the oil filtration determination unit is disclosed.

Referring to FIG. 3 , the determination of whether to filter the oil may be performed based on the oil filtration determination gas 350 capable of increasing the accuracy of determination of whether oil is filtered among the six types of gas-in-oil 300 .

Specifically, the oil filtration determination gas 350 may be a gas that is reduced by a critical percentage or more as a result of measurement when the oil filtration operation is performed. In an embodiment of the present invention, methane (CH ₄ ), ethylene (C ₂ H ₄ ), and ethane (C ₂ H ₆ ) three gases among the six gas-in-oil 300 are simultaneously critical percent (for example, , at 60% drop), it can be determined that oil filtration has proceeded.

The gas-in-oil that cannot be selected as the oil filtration determination gas 350 may be a gas with a loss of more than a critical percentage into the air during oil extraction, a gas with a data distribution above a threshold, and a gas with a data change/occurrence below a threshold. have.

For example, in the present invention, H ₂ of the six types of gases is often lost to the air during oil extraction, so it corresponds to a gas with low data accuracy, CO corresponds to a gas with too large data distribution, and C ₂ H ₂ is As a gas that does not occur in most cases other than arc failure, it may not be selected as the oil filtration determination gas 350 .

Table 1 below is data on actual transformer gas-in-oil that has undergone oil filtration, and data on 6 types of gas-in-oil (300) where 2015 is the time of oil filtration. The unit may be parts per million (ppm).

<Table 1>

It can be seen that the methane (CH ₄ ), ethylene (C ₂ H ₄ ), and ethane (C ₂ H ₆ ) three gases are reduced to 60% or less after oil filtration, and the three gases (or at least one of the three gases) One gas) is used as the oil filtration determination gas 350 and may be used to determine whether the oil is filtered or not.

The oil filtration determination gas 300 is an example, and other oil filtration determination gases may be utilized to determine whether oil is filtered, and this embodiment may also be included in the scope of the present invention.

4 is a conceptual diagram illustrating a method for performing correction on gas-in-oil measurement data according to an embodiment of the present invention.

In FIG. 4, a measurement error value correction model based on statistics for correction of gas-in-oil measurement data is disclosed.

Referring to FIG. 4 , the statistical-based measurement error value correction may correct the measurement error value based on a similar data set similar to the target data set 420 including the measurement error value (correction target) 400 . As described above, the gas-in-oil data correction unit performs correction of the measurement error value (correction target) that is the target of the correction among the measurement error values by considering whether there is oil filtration and whether there is a subsequent measurement value based on the measurement error value. can be done

Hereinafter, the measurement error value (correction target) 400 and the measurement error value (correction target) 400 are predicted based on k neighboring data based on the measurement error value (correction target) 400 Thus, the measurement error value (correction target) 400 may be corrected. For example, the target data set 420 including the surrounding data of the measurement error value (correction target) 400 and the measurement error value (correction target) 400 and the data set including the previously measured measurement value ( A comparison between measurements) 440 may be performed. Among the data sets (measurements) 440 , a data set (measurement) 440 having a high similarity to the target data set 420 may be determined as the similar data set 460 . In other words, the similar data set 460 may be a data set adjacent to the target data set 420 . A threshold number of similar data sets may be determined, and a measurement error value (correction target) may be corrected based on values included in the threshold number of similar data sets.

Specifically, in order to find a similar data set 460 similar to the target data set 420 including the measurement error value (correction target), a similarity analysis based on other measurement values included in the target data set 420 will be performed. can

The size of the window (or the size of the data set) for the similarity analysis is determined, the number of data to be included in the target data set 420 is determined, and based on the measured values 1 to n included in the target data set 420 . A similar data set 460 may be searched for.

The similar data set 460 is a difference value between each of the measured values 1 to n included in the target data set 420 and the corresponding measured values 1' to n' located in the data set (measurement) 440 . Based on the (or distance value), the smaller the difference value (or the distance value), the higher the similarity data set 460 may be determined.

In order to determine the similarity data set 460, a distance calculation technique for calculating a distance between measured values includes Euclidean Distance, Manhattan Distance, Mahalanobis Distance, and the like, and may be selectively used depending on the case of data.

A similar data set 460 may be determined considering each of n years of data for each gas-in-oil. The size of the window for determining the similar data set 460 may be set differently for each gas-in-oil, and may be set differently depending on the distribution (or amount of data) of accumulated data.

Alternatively, according to an embodiment of the present invention, a similar data set 460 may be determined by defining separate data sets for each gas when oil filtration is performed and when oil filtration is not performed. Specifically, the first data set (measurement) includes only measurement data when oil filtration is not performed within the critical time, and the second data set (measurement) includes measurement data when oil filtration is performed within the critical time can do

In correcting the measurement error value (correction target) after the oil filtration time point using these separate first data set (measurement) and second data set (measurement), considering whether oil filtration was performed within the critical time point, each other Determination of similar data sets may be made based on other data sets. For example, if the peripheral measurement value for correcting the measurement error value (correction target) includes a measurement time point within a critical period after the oil filtration time point, a similar data set is determined on the second data set (measurement) to determine the measurement error value (correction target) can be corrected.

5 is a conceptual diagram illustrating a method of determining a measurement error value (correction target) based on whether oil is filtered according to an embodiment of the present invention.

In FIG. 5 , a method for determining whether to filter oil based on the oil filtration determination gas 500 and whether to perform a measurement error value (correction target) based on a measurement error value correction model is disclosed.

Referring to FIG. 5 , three types of gases (CH ₄ , C ₂ H ₄ , C ₂ H ₆ ) among the six types of gas in oil may be the oil filtration determination gas 500 . When the oil filtration determination gas 500 is reduced by a critical percentage (eg, 60%) at the same time at the next measurement time point, it may be determined that the oil filtration has progressed.

As described above, in order to determine whether the oil is filtered _{, it may be determined whether a phenomenon in which the CH 4} , C ₂ H ₄ , C ₂ H ₆ gas is simultaneously reduced by more than a critical percentage occurs within a certain change range.

As shown in FIG. 5 , a time point at which the value of the oil filtration determination gas 500 is rapidly reduced or measured to 0 may be determined as a time point at which oil filtration is performed. A correction procedure for the measurement error value may not be performed on the measurement error value in the section including the oil filtration time point 520 . Conversely, the correction procedure for the measurement error value may not be performed for the measurement error value in the section not including the oil filtration time point 520 . That is, the measurement error value of the section including the oil filtration time point 520 is not set as the measurement error value (correction target) and is not subject to the correction procedure.

_{For example, since the value of H 2} measured in July 1997 corresponding to the oil filtration time 520 has a value reduced from 26 ppm to 0 ppm, it may be determined as a measurement error value, but July 1997 As the point at which filtration was performed, the calibration algorithm may not be performed. Conversely, a measurement error value among measurement values after November 1997 that does not correspond to the oil filtration time point 520 may be a measurement error value (correction target) to which a correction algorithm is applied. For example, the measurement error value among the values measured at 1997.11, 1998.3, 1998.8, and 1998.12, which is a time point after the oil filtration time, is determined as a measurement error value (correction target), and a correction algorithm may be applied. Specifically, if H _{2 measured in August 1998 and C 2} H ₆ measured in March 1998 are measurement error values (correction target), based on the correction algorithm for these measurement error values (correction target), A calibration procedure may proceed. A correction algorithm for the measurement error value (correction target) will be described later.

6 is a conceptual diagram illustrating a gas-in-oil correction method according to an embodiment of the present invention.

A first calibration procedure and a second calibration procedure for gas-in-oil calibration are disclosed in FIG. 6 .

Referring to FIG. 6 , a measurement error value may be determined from gas-in-oil measurement data (step S600 ).

As described above, the measurement error value may include a value for which measurement is not performed, such as a missing value, and may include an abnormal range value, which is a value out of a normal range in comparison with an existing measurement value for each gas.

A measurement error value (correction target) that can be corrected among measurement error values may be determined (step S610).

Among the measurement error values, the measurement error value corresponding to the oil filtration time may be excluded from the correction target. In addition, when there is no previous measurement value and/or a subsequent measurement value based on the measurement error value among the measurement error values, it may be excluded from the correction target. After the determination of the measurement error value, if the previous measurement value and the subsequent measurement value do not exist based on the measurement error value, the additional correction procedure may not be performed because the accuracy of the correction is low.

Correction is performed on the measurement error value (correction target) (step S620).

Correction for the measurement error value (correction target) may be performed when there is a subsequent measurement value as data of a section in which oil filtration is not performed.

It is determined whether to perform the first calibration procedure or the second calibration procedure (step S630), and a first calibration procedure (step S640) or a second calibration procedure (step S650) may be performed based on the determination. A first calibration procedure (step S640) or a second calibration procedure (step S650) may be selectively performed on a plurality of measurement error values included in the target data set.

The first correction procedure (step S640 ) may be a correction procedure for determining a measurement error value (correction target) based on a previous measurement value in the target data set.

The second correction procedure (step S650 ) may be a correction procedure based on the statistical-based measurement error value correction described above with reference to FIG. 4 .

Any target data set set based on all possible windows (hereinafter, available windows) including the measurement error value (correction target) does not contain more than a threshold number of data to have accuracy by performing statistical-based measurement error value correction. In this case, the first correction procedure (step S640) may be performed.

Specifically, the first correction procedure (step S640) may be performed when the following conditions are satisfied.

Condition (first calibration procedure): When no target data set set based on the available window containing measurement error values (calibration target) contains measurements as first and last values

Conversely, the second calibration procedure may be performed when the first calibration procedure is not satisfied. Specifically, it may be performed when the following conditions (second correction procedure) are satisfied.

Condition (Second Calibration Procedure): When the target data set set based on the available window containing the measurement error value (calibration target) contains the measured values as the first and last values.

The measurement value for determining whether the condition (the first calibration procedure) and the condition (the second calibration procedure) is satisfied may include an already corrected measurement error value (correction target). That is, after the measurement error value (correction target) is also corrected by the first or second correction procedure, it is determined as a measured value to determine whether the condition (the first correction procedure) and the condition (the second correction procedure) are satisfied. Judgment may be performed.

A first calibration procedure and a second calibration procedure may be performed based on the above conditions.

The reason for separately performing the first and second calibration procedures as described above is the statistical-based measurement error correction procedure when a specific measurement value (eg, the first value and/or the last value in the data set) does not exist. This is because the accuracy of the correction decreases when the second correction procedure is performed.

The size of the window used to determine the data set for the first calibration procedure and the second calibration procedure may be set differently for each gas-in-oil, and may differ from each other depending on the distribution (or amount of data) of the accumulated data. It can be set differently.

7 is a conceptual diagram illustrating a second correction procedure for performing statistical-based measurement error correction according to an embodiment of the present invention.

7 discloses a second correction procedure, which is a statistical-based measurement error value correction procedure for correcting a measurement error value (correction target) in the target data set based on the target data set and the data set (measurement).

Referring to FIG. 7 , a KNN (k-nearest neighbor)-based ranking may be determined between a data set (measurement) 720 and a target data set 700 including a measurement error value (correction target). A data set (measurement) having a high similarity to the set 700 may have a higher ranking.

When the threshold number of data sets (measurements) 720 having a relatively high ranking among the target data sets 700 is determined as the similar data set 740 , the measured values 755 , included in the similar data set 740 , Based on 765 , a corresponding measurement value in the target data set 700 may be determined.

For example, in the case of the first measurement error value (correction target) 750 , the second is included in the target data set 700 , and the first measurement error value (correction target) 750 includes the similar data set 740 . ) may be determined as an average value of the second measured values 755 included in the . In the case of the second measurement error value (correction target) 760, the third is included in the target data set 700, and the second measurement error value (correction target) 760 is included in the similar data set 740 It may be determined as an average value of the third measured value 765 .

8 is a conceptual diagram illustrating a method for performing correction on a measurement error value (correction target) according to an embodiment of the present invention.

8 discloses a method of determining a measurement error value (correction target) that can be corrected among measurement error values and performing correction on the measurement error value (correction target).

Referring to FIG. 8 , when there is no measurement value after the measurement error value as in the data set A 800 , the measurement error value may not be corrected and may be determined to be a non-correctable value.

Conversely, when there is a measurement value measured after the measurement error value as in the data set B 820 , the measurement error value is determined as the measurement error value (correction target) and correction may be performed.

If it is assumed that the target data set includes 4 data by setting the window size to 4, the following correction procedure may be performed.

1) With respect to the first measurement error value (correction target) 801 , since the above-described condition (first correction procedure) is satisfied, the first correction procedure 850 may be performed. That is, the first measurement error value (correction target) 801 may be corrected to the previous measurement value of 7.8.

2) With respect to the second measurement error value (correction target) 802 , since the above-described condition (first correction procedure) is satisfied, the first correction procedure 850 may be performed. That is, the second measurement error value (correction target) 802 may be corrected to the previous measurement value of 7.8.

3) With respect to the third measurement error value (correction target) 803 , since the above-described condition (first correction procedure) is satisfied, the first correction procedure 850 may be performed. That is, the third measurement error value (correction target) 803 may be corrected to the previous measurement value of 7.8.

4) For the fourth measurement error value (correction target) 804 and the fifth measurement error value (correction target) 805 , the above-described condition (second correction procedure) is satisfied, so the second correction procedure 860 is performed can be

In this way, the measurement error value (correction target) included in the data set B 820 may be determined through the first correction procedure 850 and the second correction procedure 860 .

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the computer software field. Examples of computer-readable recording media include hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks. medium), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. A hardware device may be converted into one or more software modules to perform processing in accordance with the present invention, and vice versa.

In the above, the present invention has been described with reference to specific matters such as specific components and limited embodiments and drawings, but these are only provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments, Those of ordinary skill in the art to which the invention pertains can make various modifications and changes from these descriptions.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and the scope of the spirit of the present invention is not limited to the scope of the scope of the present invention. will be said to belong to

Claims (6)

The method for determining the condition of a transformer based on gas-in-oil data calibration is:
Transformer state determination device receiving the gas-in-oil data;
determining, by the transformer state determination device, a measurement error value to be corrected on the gas-in-oil data;
correcting, by the transformer state determination device, the measurement error value; and
Method comprising the step of determining the state of the transformer based on the gas-in-oil data including the corrected measurement error value by the transformer state determination device. The method of claim 1, wherein determining the measurement error value comprises:
and determining, by the transformer state determination device, the measurement error value in consideration of whether oil is filtered based on the gas-in-oil data. The method of claim 2, wherein correcting the measurement error value comprises:
and correcting, by the transformer state determination device, the measurement error value based on a similar data set similar to a data set including the measurement error value. Transformer condition determination device that determines the condition of the transformer based on gas-in-oil data correction,
a gas-in-oil data input configured to receive gas-in-oil data; and
a processor operatively connected to the gas-in-oil data input;
The processor receives gas-in-oil data,
Determining a measurement error value to be corrected on the gas-in-oil data,
correct the measurement error value,
Transformer state determination device, characterized in that for determining the state of the transformer based on the gas-in-oil data including the corrected measurement error value. 5. The method of claim 4,
Transformer state determination device, characterized in that the processor is implemented to determine the measurement error value in consideration of whether oil filtering based on the gas-in-oil data. 6. The method of claim 5,
and the processor corrects the measurement error value based on a similar data set similar to a data set including the measurement error value.

Images