Method, system, storage device and equipment for online monitoring data identification of dissolved gas in oil

Abstract

The present invention discloses a method, system, memory, and device for identifying online dissolved gas in oil monitoring data. This method uses the mean and standard deviation of preprocessed dissolved gas in oil time series samples, the Euler distance and standard deviation of the time series samples, and the difference, mean, and standard deviation between time series samples as statistical features. Based on the Laetitia rule, the upper and lower limits of the statistical feature distribution are calculated to construct a dissolved gas in oil reference sequence. A gray correlation analysis method is used to calculate the correlation between the time series to be tested and the reference sequence. Based on the correlation value, the normality of the tested data is determined. This method can effectively and rationally handle the problem of identifying anomalies in multidimensional data, providing an accurate and reliable basis for operating status evaluation and fault diagnosis of oil-immersed transformers.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer monitoring, and in particular to a method, system, memory and equipment for online monitoring data identification of dissolved gas in oil. Background Art

[0002] Dissolved gas analysis of oil-immersed transformers is a crucial tool for determining operating status and potential fault types. Online dissolved gas monitoring data is a crucial source for monitoring the transformer's real-time operating conditions. Its accuracy and reliability are crucial for ensuring the reliable operation of the transformer and the power grid. In actual operation, issues such as quality issues with online monitoring devices, sensor failures, and abnormalities in data transmission loops can lead to anomalies in online monitoring data. The accuracy and reliability of online dissolved gas monitoring data are crucial for ensuring the stable operation of transformer equipment and power systems.

[0003] Currently, extensive research has been conducted on anomaly detection in online monitoring data for oil-immersed transformers, a crucial component of transformer condition monitoring. While effective methods exist for identifying simple anomalies such as missing data, negative values, and over-range data, there is still room for improvement in holistic identification methods for multidimensional data. Traditional methods such as mathematical statistics, the K-nearest neighbor method, and random forests are ineffective and inadequate for identifying anomalies in multidimensional data. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, system, memory and device for online monitoring data identification of dissolved gas in oil, so as to solve the problem that existing methods cannot effectively and reasonably solve the problem of multi-dimensional data anomaly identification.

[0005] To achieve the above object, the present invention is achieved through the following technical solutions:

[0006] In one aspect, the present invention provides a method for identifying online monitoring data of dissolved gas in oil, comprising:

[0007] Based on the multidimensional statistical characteristics of the time series samples of online monitoring data of dissolved gas in oil-immersed transformers, a reference sequence of dissolved gas in oil is constructed.

[0008] Calculating a correlation value between monitoring data of dissolved gas to be detected in oil-immersed transformer oil and a reference sequence of dissolved gas in the oil;

[0009] The status of the dissolved gas monitoring data to be detected is identified according to the correlation value.

[0010] Furthermore, the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in oil-immersed transformer oil include:

[0011] Based on the sampling period of online monitoring of dissolved gas in oil of oil-immersed transformers, the time window is determined and a time series sample set of online monitoring data of dissolved gas in oil is established in chronological order.

[0012] Applying square root transformation to the time series sample set of online monitoring data of dissolved gas in oil to reconstruct the time series sample set of online monitoring data of dissolved gas in oil;

[0013] Calculating the mean and standard deviation of characteristic gases corresponding to all time series based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0014] Calculating the mean Euler distance and the corresponding standard deviation between each time series sample and the average value of the corresponding characteristic gas of all sequences based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0015] Calculate the mean and corresponding standard deviation of the difference value sequence of adjacent time series samples based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0016] Based on the mean of the characteristic gases corresponding to all sequences, the standard deviation of the characteristic gases corresponding to all sequences, the mean of the Euler distances between the means of the characteristic gases corresponding to all sequences, the standard deviation between the means of the characteristic gases corresponding to all sequences, the mean of the difference sequence of adjacent time series samples and the standard deviation corresponding to the mean of the difference sequence of adjacent time series samples, the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in oil are constructed.

[0017] Furthermore, the construction of a reference sequence of dissolved gas in oil includes:

[0018] Based on Laida's law, the upper and lower limits of normal values ​​for online monitoring data of dissolved gas in oil are established;

[0019] Based on Laida's law, the upper and lower limits of abnormal values ​​of online monitoring data of dissolved gas in oil are established;

[0020] The time series of the calculated mean values, upper and lower limits of normal values, and upper and lower limits of abnormal values ​​are used as the reference series of dissolved gas in oil.

[0021] Furthermore, the calculation of the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil includes:

[0022] The mean, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value of all characteristic gases in the series are used as reference series to establish a grey relational analysis model for the mean of time series.

[0023] Taking the mean of Euler distance, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for Euler distance between time series is established.

[0024] Taking the mean of the difference sequence of adjacent time series samples, the upper limit of normal value, the lower limit of normal value, and the upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for the difference between time series is established;

[0025] Substitute the dissolved gas monitoring data to be detected into the three established grey correlation analysis models in turn, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey correlation sequence and correlation value of each model;

[0026] The dissolved gas monitoring data to be detected is the monitoring data of dissolved gas in oil at a certain moment: hydrogen, methane, ethane, ethylene and acetylene.

[0027] Furthermore, identifying the status of the dissolved gas monitoring data to be detected according to the correlation value includes:

[0028] Determine the state corresponding to the maximum value in each grey relational degree sequence, wherein the state includes normal and abnormal;

[0029] If the states corresponding to at least two maximum values ​​are normal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as normal. If the states corresponding to at least two maximum values ​​are abnormal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as abnormal.

[0030] A second aspect of the present invention provides an abnormality identification system for online monitoring data of dissolved gas in oil of an oil-immersed transformer, comprising:

[0031] A reference sequence construction module is used to construct a reference sequence of dissolved gas in oil based on the multidimensional statistical characteristics of the time series samples of online monitoring data of dissolved gas in oil-immersed transformers;

[0032] A correlation calculation module is used to calculate the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil;

[0033] An identification module is used to identify the status of the dissolved gas monitoring data to be detected according to the correlation value.

[0034] Furthermore, it also includes:

[0035] The raw data acquisition module is used to determine the time window based on the online monitoring sampling period of dissolved gas in oil of the oil-immersed transformer, and to establish a time series sample set of online monitoring data of dissolved gas in oil in chronological order;

[0036] Applying square root transformation to the time series sample set of online monitoring data of dissolved gas in oil to reconstruct the time series sample set of online monitoring data of dissolved gas in oil;

[0037] as well as,

[0038] A feature calculation module is used to calculate the mean and standard deviation of the characteristic gases corresponding to all sequences based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0039] Calculating the mean Euler distance and the corresponding standard deviation between each time series sample and the mean of the corresponding characteristic gas of all sequences based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0040] Calculate the mean and corresponding standard deviation of the difference value sequence of adjacent time series samples based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil;

[0041] Based on the mean of the characteristic gases corresponding to all sequences, the standard deviation of the characteristic gases corresponding to all sequences, the mean of the Euler distances between the means of the characteristic gases corresponding to all sequences, the standard deviation between the means of the characteristic gases corresponding to all sequences, the mean of the difference sequence of adjacent time series samples and the standard deviation corresponding to the mean of the difference sequence of adjacent time series samples, the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in oil are constructed.

[0042] Furthermore, the reference sequence construction module is specifically used to:

[0043] Based on Laida's law, the upper and lower limits of normal values ​​for online monitoring data of dissolved gas in oil are established;

[0044] Based on Laida's law, the upper and lower limits of abnormal values ​​of online monitoring data of dissolved gas in oil are established;

[0045] The time series of the calculated mean values, upper and lower limits of normal values, and upper and lower limits of abnormal values ​​are used as the reference series of dissolved gas in oil.

[0046] Furthermore, the correlation calculation module is specifically used to:

[0047] The mean, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value of all characteristic gases in the series are used as reference series to establish a grey relational analysis model for the mean of time series.

[0048] Taking the mean of Euler distance, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for Euler distance between time series is established.

[0049] Taking the mean of the difference sequence of adjacent time series samples, the upper limit of normal value, the lower limit of normal value, and the upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for the difference between time series is established;

[0050] Substitute the dissolved gas monitoring data to be detected into the three established grey correlation analysis models in turn, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey correlation sequence and correlation value of each model;

[0051] The dissolved gas monitoring data to be detected is the monitoring data of dissolved gas in oil at a certain moment: hydrogen, methane, ethane, ethylene and acetylene.

[0052] Furthermore, the identification module is specifically used to:

[0053] Determine the state corresponding to the maximum value in each grey relational degree sequence, wherein the state includes normal and abnormal;

[0054] If the states corresponding to at least two maximum values ​​are normal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as normal. If the states corresponding to at least two maximum values ​​are abnormal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as abnormal.

[0055] A third aspect of the present invention provides a computer-readable memory storing one or more programs, wherein the one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described above.

[0056] A fourth aspect of the present invention provides a device comprising:

[0057] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the aforementioned methods.

[0058] The present invention has the following beneficial effects:

[0059] The present invention provides a method for identifying anomalies in online monitoring data of dissolved gas in oil-immersed transformers. The method uses a time series of online monitoring data of dissolved gas in oil-immersed transformers as a sample, calculates multidimensional statistical features of the sample, uses the multidimensional statistical features as a reference sequence, calculates the correlation value between the dissolved gas monitoring data to be detected and the reference sequence, and then identifies the state of the dissolved gas monitoring data to be detected. The present invention establishes multidimensional statistical features, uses them as a reference to calculate the correlation value between the data to be detected, and effectively and rationally handles the problem of identifying anomalies in multidimensional data, providing an accurate and reliable basis for evaluating the operating status and diagnosing faults of oil-immersed transformers. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a flow chart of a method for identifying online monitoring data of dissolved gas in oil provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0062] Example 1

[0063] This embodiment provides a method for identifying abnormalities in online monitoring data of dissolved gas in oil-immersed transformer oil, comprising:

[0064] Based on the multidimensional statistical characteristics of the time series samples of online monitoring data of dissolved gas in oil-immersed transformers, a reference sequence of dissolved gas in oil is constructed.

[0065] Calculating a correlation value between monitoring data of dissolved gas to be detected in oil-immersed transformer oil and a reference sequence of dissolved gas in the oil;

[0066] The status of the dissolved gas monitoring data to be detected is identified according to the correlation value.

[0067] In this embodiment, the dissolved gas in the oil-immersed transformer oil is composed of the content characteristics of five characteristic gases, including: hydrogen, methane, ethane, ethylene and acetylene.

[0068] In this embodiment, the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in oil include mean and standard deviation, Euler distance and standard deviation of sample time series, difference between adjacent time series samples, and mean and standard deviation of difference series.

[0069] In this embodiment, the status of the monitoring data of the to-be-detected dissolved gas in oil is identified according to the maximum correlation criterion.

[0070] Example 2

[0071] This embodiment provides a method for identifying online monitoring data of dissolved gas in oil-immersed transformer oil. Figure 1 The specific implementation process is as follows:

[0072] Step 1: Obtain the time series samples of online monitoring data of dissolved gas in pre-treated oil of oil-immersed transformer;

[0073] Step 2: Calculate multidimensional statistical features based on the time series samples of online monitoring data of dissolved gas in oil;

[0074] Step 3: Based on the calculated multidimensional statistical features, a reference sequence of dissolved gas in oil is constructed;

[0075] Step 4: Calculate the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil;

[0076] Step 5: Identify the status of the monitoring data of the dissolved gas to be detected in the oil according to the maximum correlation criterion.

[0077] In this embodiment, a time series sample of online monitoring data of dissolved gas in pre-treated oil of an oil-immersed transformer is obtained. The specific implementation process is as follows:

[0078] Step 11. The time series sample of online monitoring data of dissolved gas in oil-immersed transformer oil is composed of the content characteristics of five characteristic gases; the five characteristic gases include hydrogen, methane, ethane, ethylene and acetylene.

[0079] Step 12: Determine the time window based on the online monitoring sampling period of dissolved gas in oil pretreatment of oil-immersed transformers, and establish a time series sample set T of online monitoring data of dissolved gas in oil in chronological order;

[0080]

[0081] Where T represents the time series sample set of online monitoring data of dissolved gas in oil with a time window length of n, T i , i=1,2,…,n represents the online monitoring data sample of dissolved gas in oil before the i-th sampling interval, H 2,i ,i=1,2,…,n represents the online monitoring data sample of hydrogen in oil before the i-th sampling interval, CH 4,i ,i=1,2,…,n represents the online monitoring data sample of methane in oil before the i-th sampling interval, C2H 6i ,,i=1,2,…,n represents the online monitoring data sample of ethane in oil before the i-th sampling interval, C2H 4i ,,i=1,2,…,n represents the online monitoring data sample of ethylene in oil before the i-th sampling interval, C2H 2,i ,i=1,2,…,n represents the online monitoring data sample of acetylene in oil before the i-th sampling interval;

[0082] Step 13: Use square root transformation on the time series sample set T of online monitoring data of dissolved gas in oil to reconstruct the time series set T of online monitoring data of dissolved gas in oil. SQ :

[0083]

[0084] TSQi ,i=1,2,…,n represents the square root of the online monitoring data sample of dissolved gas in oil before the i-th sampling interval.

[0085] The purpose of this step is to improve the data contained in the time series set T to be close to or consistent with the normal distribution.

[0086] In this embodiment, the multi-dimensional statistical characteristics of the time series samples of dissolved gas in oil are calculated. The specific implementation process is as follows:

[0087] Statistical features include mean and standard deviation, Euler distance and standard deviation of sample time series, difference between adjacent time series samples, mean and standard deviation of difference series, etc.

[0088] Step 21, based on the reconstructed time series sample set T of online monitoring data of dissolved gas in oil SQ Calculate the mean E of the characteristic gas corresponding to all sequences cs and standard deviation σ cs , calculated as follows:

[0089]

[0090] Step 22: Based on the reconstructed time series sample set T of online monitoring data of dissolved gas in oil SQ Calculate the average value E of each time series sample and the corresponding characteristic gas of all sequences cs The mean Euler distance E between OL and the corresponding standard deviation σ OL , calculated as follows:

[0091]

[0092] Step 23, based on the reconstructed time series sample set T of online monitoring data of dissolved gas in oil SQ Calculate the absolute value of the difference between adjacent time series samples and the mean of the difference series E CZ and the corresponding standard deviation σ CZ , the calculation formula is as follows:

[0093] T CZ,i =|T SQi -T SQ(i+1) |;

[0094]

[0095] Among them, T CZ,i It represents the absolute value of the difference between the online monitoring data sample of dissolved gas in oil before the i-th sampling interval and the sample of the adjacent sampling interval.

[0096] In this embodiment, a reference sequence of dissolved gas in oil is constructed based on the multidimensional statistical characteristics of the dissolved gas in oil time series samples. The specific implementation process is as follows:

[0097] Step 31: Based on the Laida rule, establish the upper and lower limits of the normal value of the online monitoring data of dissolved gas in oil, where E csup3 、E OLup3 and E CZup3 They represent the mean E of the sample sequence of dissolved gas in oil within the sequence time window. cs Upper limit of normal value, time series sample and average value E of corresponding characteristic gas of all sequences cs The mean Euler distance E between OL Upper limit of normal value and time series sample set T SQ The mean value of the adjacent time difference series E CZ Upper limit of normal value; E cslow3 、E OLlow3 and E CZlow3 They represent the mean E of the sample sequence of dissolved gas in oil within the sequence time window. cs The lower limit of normal value, the average value E of time series samples and corresponding characteristic gases of all sequences cs The mean Euler distance E between OL Normal value lower limit and time series sample set T SQ The mean value of the adjacent time difference series E CZ lower limit of normal;

[0098] E csup3 =E cs +3×σ cs ;E cslow3 =E cs -3×σ cs ;

[0099] E OLup3 =E OL +3×σ OL ;E OLlow3 =E OL -3×σ OL ;

[0100] E CZup3 =E CZ +3×σ CZ ;E CZlow3 =E CZ -3×σ CZ ;

[0101] Step 32: Based on the Laida rule, establish the upper and lower limits of abnormal values ​​of online monitoring data of dissolved gas in oil: csup4 、E OLup4 and E CZup4They represent the mean E of the sample sequence of dissolved gas in oil within the sequence time window. cs The upper limit of the abnormal value, the mean E of the time series sample and the corresponding characteristic gas of all sequences cs The mean Euler distance E between OL Outlier upper limit and time series sample set T SQ The mean value of the adjacent time difference series E CZ Upper limit of outlier; E cslow4 、E OLlow4 and E CZlow4 They represent the mean E of the sample sequence of dissolved gas in oil within the sequence time window. cs The lower limit of the outlier, the mean E of the time series sample and the corresponding characteristic gas of all sequences cs The mean Euler distance E between OL Outlier lower limit and time series sample set T SQ The mean value of the adjacent time difference series E CZ Outlier lower limit;

[0102] E cscup4 =E cs +4×σ cs ;E csclow4 =E cs -4×σ cs ;

[0103] E OLCup4 =E OL +4×σ OL ;E OLClow4 =E OL -4×σ OL ;

[0104] E CZCup4 =E CZ +4×σ CZ ;E CZClow4 =E CZ -4×σ CZ ;

[0105] Step 33: The solved time series mean, normal value upper and lower limits, and abnormal value upper and lower limits are used as the reference series S of dissolved gas in oil. ref .

[0106] In this embodiment, the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil is calculated. The specific implementation process is as follows:

[0107] Step 41, take the mean E cs , upper limit of normal value E csup3 , lower limit of normal value E cslow3 And the upper limit of outlier E cscup4and the lower limit E csclow4 As a reference sequence, a grey relational analysis model for the mean of time series is established; with the mean E OL , upper limit of normal value E OLup3 , lower limit of normal value E OLlow3 And the upper limit of outlier E OLCcup4 and the lower limit E OLClow4 As a reference sequence, a grey relational analysis model for Euler distance between time series is established; with the mean E CZ , upper limit of normal value E CZup3 , lower limit of normal value E CZlow3 And the upper limit of outlier E CZCcup4 and the lower limit E CZClow4 As a reference sequence, a grey relational analysis model for the difference between time series is established;

[0108] Step 42: Substitute the dissolved gas monitoring data to be detected into the three grey correlation analysis models established above, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey correlation degree sequence of each model:

[0109]

[0110] Among them, D cs 、D csup3 、D cslow3 、D cscup4 and D cslow4 They are respectively the monitoring data of dissolved gas to be detected and E cs 、E csup3 、E cslow3 E cscup4 and E csclow4 The correlation value, D OL 、D OLup3 、D OLlow3 、D OLcup4 and D OLlow4 They are respectively the monitoring data of dissolved gas to be detected and E OL 、E OLup3 、E OLlow3 E OLCcup4 and E OLClow4 The correlation value, D CZ 、D CZup3 、D CZlow3 、D CZcup4 and D CZlow4 They are respectively the monitoring data of dissolved gas to be detected and E CZ 、E CZup3 、E CZlow3 E CZCcup4 and E CZClow4 The correlation value of .

[0111] It should be noted that the dissolved gas monitoring data to be detected is the dissolved gas monitoring data in oil at a certain moment. At the moment t=n+i, its expression is as follows:

[0112] S test,n+i =[H 2(n+i) ,CH 4(n+i) ,C2H 6(n+i) ,C2H 4(n+i) ,C2H 2(n+i) ].

[0113] In this embodiment, the status of the monitoring data of the dissolved gas to be detected in the oil is identified according to the maximum correlation criterion. The specific implementation process is as follows:

[0114] Determine the states corresponding to the maximum values ​​Max(D1), Max(D2) and Max(D3) in each grey relational degree sequence, where D cs 、D OL and E OL The corresponding state is normal; D csup3 、D OLup3 and D CZup3 The corresponding state is normal; D cslow3 、D OLlow3 and D CZlow3 The corresponding state is normal; D cscup4 、D OLcup4 and D CZcup4 The corresponding state is abnormal; D cslow4 、D OLlow4 and D CZlow4 The corresponding status is abnormal;

[0115] If the states corresponding to at least two maximum values ​​are normal, the state of the monitoring data of dissolved gas to be detected in oil is identified as normal. If the states corresponding to at least two maximum values ​​are abnormal, the state of the monitoring data of dissolved gas to be detected in oil is identified as abnormal.

[0116] This embodiment establishes multidimensional statistical features by using the mean, Euler distance, and difference between adjacent time series samples of dissolved gas in oil, and uses the Laetitia rule to determine the upper and lower limits of normal and abnormal values ​​of each dimensional feature to construct a reference sequence. Then, a grey correlation analysis is performed to determine the correlation between the time series to be tested and the reference sequence, thereby effectively and reasonably handling the problem of anomaly identification in multidimensional data, and providing accurate and reliable data for operating status evaluation and fault diagnosis of oil-immersed transformers.

[0117] Example 3

[0118] This embodiment provides an online monitoring data identification system for dissolved gas in oil, comprising:

[0119] A reference sequence construction module is used to construct a reference sequence of dissolved gas in oil based on the multidimensional statistical characteristics of the time series samples of online monitoring data of dissolved gas in oil-immersed transformers;

[0120] A correlation calculation module is used to calculate the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil;

[0121] An identification module is used to identify the status of the dissolved gas monitoring data to be detected according to the correlation value.

[0122] In this embodiment, it also includes:

[0123] The raw data acquisition module is used to determine the time window based on the online monitoring sampling period of dissolved gas in oil of oil-immersed transformers, and to establish a time series sample set of online monitoring data of dissolved gas in oil in chronological order:

[0124]

[0125] Where T represents the time series sample set of online monitoring data of dissolved gas in oil with a time window length of n, T i , i=1,2,…,n represents the online monitoring data sample of dissolved gas in oil before the i-th sampling interval, H 2,i ,i=1,2,…,n represents the online monitoring data sample of hydrogen in oil before the i-th sampling interval, CH 4,i ,i=1,2,…,n represents the online monitoring data sample of methane in oil before the i-th sampling interval, C2H 6i ,,i=1,2,…,n represents the online monitoring data sample of ethane in oil before the i-th sampling interval, C2H 4i, , i=1,2,…,n represents the online monitoring data sample of ethylene in oil before the i-th sampling interval, C2H 2,i ,i=1,2,…,n represents the online monitoring data sample of acetylene in oil before the i-th sampling interval;

[0126] The time series sample set T of the online monitoring data of dissolved gas in oil is transformed by square root transformation to reconstruct the time series sample set of the online monitoring data of dissolved gas in oil as follows:

[0127]

[0128] Among them, T SQ is the reconstructed time series sample set of online monitoring data of dissolved gas in oil, T SQi ,i=1,2,…,n represents the square root of the online monitoring data sample of dissolved gas in oil before the i-th sampling interval;

[0129] as well as,

[0130] The feature calculation module is used to calculate the time series sample set T of the reconstructed online monitoring data of dissolved gas in oil. SQ Calculate the mean E of the characteristic gas corresponding to all sequences cs and standard deviation σ cs ,as follows:

[0131]

[0132] Based on the reconstructed time series sample set T of online monitoring data of dissolved gas in oil SQ Calculate the mean E of each time series sample and the corresponding characteristic gas of all sequences cs The mean Euler distance E between OL and the corresponding standard deviation σ OL ,as follows:

[0133]

[0134] Based on the reconstructed time series sample set T of online monitoring data of dissolved gas in oil SQ Calculate the mean E of the difference sequence of adjacent time series samples CZ and the corresponding standard deviation σ CZ ,as follows:

[0135] T CZ,i =|T SQi -T SQ(i+1) |;

[0136]

[0137] Among them, T CZ,i It represents the absolute value of the difference between the online monitoring data sample of dissolved gas in oil before the i-th sampling interval and the sample of the adjacent sampling interval;

[0138] E cs , σ cs 、E OL , σ OL 、E CZ and σ CZ Multidimensional statistical characteristics of time series samples of online monitoring data of dissolved gas in oil.

[0139] In this embodiment, the reference sequence construction module is specifically used to:

[0140] Based on the Laida law, the upper and lower limits of normal values ​​for online monitoring data of dissolved gas in oil are established as follows:

[0141] E csup3 =E cs +3×σ cs ;E cslow3 =Ecs -3×σ cs ;

[0142] E OLup3 =E OL +3×σ OL ;E OLlow3 =E OL -3×σ OL ;

[0143] E CZup3 =E CZ +3×σ CZ ;E CZlow3 =E CZ -3×σ CZ ;

[0144] Among them, E csup3 、E OLup3 and E CZup3 Represents the mean E cs Upper limit of normal value, mean E OL Upper limit of normal value and mean E CZ Upper limit of normal value; E cslow3 、E OLlow3 and E CZlow3 Represents the mean E cs Lower limit of normal value, mean E OL Lower limit of normal value and mean E CZ lower limit of normal;

[0145] Based on the Laida law, the upper and lower limits of abnormal values ​​of online monitoring data of dissolved gas in oil are established as follows:

[0146] E cscup4 =E cs +4×σ cs ;E csclow4 =E cs -4×σ cs ;

[0147] E OLCup4 =E OL +4×σ OL ;E OLClow4 =E OL -4×σ OL ;

[0148] E CZCup4 =E CZ +4×σ CZ ;E CZClow4 =E CZ -4×σ CZ ;

[0149] Among them, E csup4 、E OLup4and E CZup4 Represents the mean E cs Outlier upper limit, mean E OL Outlier upper limit and mean E CZ Upper limit of outlier; E cslow4 、E OLlow4 and E CZlow4 Represents the mean E cs Outlier lower limit, mean E OL Outlier lower limit and mean E CZ Outlier lower limit;

[0150] The time series of the calculated mean values, upper and lower limits of normal values, and upper and lower limits of abnormal values ​​are used as the reference series of dissolved gas in oil.

[0151] In this embodiment, the correlation calculation module is specifically used to:

[0152] Take the mean E cs , upper limit of normal value E csup3 , lower limit of normal value E cslow3 And the upper limit of outlier E cscup4 and the lower limit E csclow4 As a reference sequence, a grey relational analysis model for the mean of time series is established; with the mean E OL , upper limit of normal value E OLup3 , lower limit of normal value E OLlow3 And the upper limit of outlier E OLCcup4 and the lower limit E OLClow4 As a reference sequence, a grey relational analysis model for Euler distance between time series is established; with the mean E CZ , upper limit of normal value E CZup3 , lower limit of normal value E CZlow3 And the upper limit of outlier E CZCcup4 and the lower limit E CZClow4 As a reference sequence, a grey relational analysis model for the difference between time series is established;

[0153] Substitute the dissolved gas monitoring data to be detected into the three established grey relational analysis models in turn, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey relational degree sequence of each model:

[0154]

[0155] Among them, D is the grey relational value matrix, D1, D2 and D3 are the grey relational value sequences, and D cs 、D csup3 、D cslow3 、D cscup4 and D cslow4 They are respectively the monitoring data of dissolved gas to be detected and E cs、E csup3 、E cslow3 E cscup4 and E csclow4 The correlation value, D OL 、D OLup3 、D OLlow3 、D OLcup4 and D OLlow4 They are respectively the monitoring data of dissolved gas to be detected and E OL 、E OLup3 、E OLlow3 E OLCcup4 and E OLClow4 The correlation value, D CZ 、D CZup3 、D CZlow3 、D CZcup4 and D CZlow4 They are respectively the monitoring data of dissolved gas to be detected and E CZ 、E CZup3 、E CZlow3 E CZCcup4 and E CZClow4 The correlation value of .

[0156] The dissolved gas monitoring data to be detected is the monitoring data of dissolved gas in oil at a certain moment: hydrogen, methane, ethane, ethylene and acetylene.

[0157] In this embodiment, the identification module is specifically used to:

[0158] Determine the states corresponding to the maximum values ​​Max(D1), Max(D2) and Max(D3) in each grey relational degree sequence, where D cs 、D OL and E OL The corresponding state is normal; D csup3 、D OLup3 and D CZup3 The corresponding state is normal; D cslow3 、D OLlow3 and D CZlow3 The corresponding state is normal; D cscup4 、D OLcup4 and D CZcup4 The corresponding state is abnormal; D cslow4 、D OLlow4 and D CZlow4 The corresponding status is abnormal;

[0159] If the states corresponding to at least two maximum values ​​are normal, the state of the monitoring data of dissolved gas to be detected in oil is identified as normal. If the states corresponding to at least two maximum values ​​are abnormal, the state of the monitoring data of dissolved gas to be detected in oil is identified as abnormal.

[0160] Example 4

[0161] This embodiment provides a computer-readable memory storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform any method according to embodiment 1 or embodiment 2.

[0162] Example 5

[0163] This embodiment provides a device, including:

[0164] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing any method according to embodiment 1 or embodiment 2.

[0165] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0166] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0167] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for identifying online monitoring data of dissolved gas in oil, characterized in that: include: Based on the multidimensional statistical characteristics of the time series samples of online monitoring data of dissolved gas in oil-immersed transformers, a reference sequence of dissolved gas in oil is constructed. The online monitoring data time series samples are composed of content characteristics of five characteristic gases, including hydrogen, methane, ethane, ethylene and acetylene; the multidimensional statistical characteristics include: the mean of the characteristic gases corresponding to all time series, the standard deviation of the characteristic gases corresponding to all time series, the mean of the Euler distance between the means of the characteristic gases corresponding to all time series, the standard deviation between the means of the characteristic gases corresponding to all time series, the mean of the difference sequence of adjacent time series samples, and the standard deviation corresponding to the mean of the difference sequence of adjacent time series samples; the construction of the dissolved gas in oil reference sequence includes: establishing the upper and lower limits of the normal value of the online monitoring data of dissolved gas in oil based on Laita's law; establishing the upper and lower limits of the abnormal value of the online monitoring data of dissolved gas in oil based on Laita's law; and using the time series of the calculated means, upper and lower limits of the normal value, and upper and lower limits of the abnormal value as the reference sequence of dissolved gas in oil; Calculating a correlation value between monitoring data of dissolved gas to be detected in oil-immersed transformer oil and a reference sequence of dissolved gas in the oil; The status of the dissolved gas monitoring data to be detected is identified according to the correlation value; the status includes normal and abnormal.

2. The method for identifying online monitoring data of dissolved gas in oil according to claim 1, characterized in that: The method for obtaining the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in the oil-immersed transformer oil includes: Based on the sampling period of online monitoring of dissolved gas in oil of oil-immersed transformers, the time window is determined and a time series sample set of online monitoring data of dissolved gas in oil is established in chronological order. Applying square root transformation to the time series sample set of online monitoring data of dissolved gas in oil to reconstruct the time series sample set of online monitoring data of dissolved gas in oil; Based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil, calculating the mean and standard deviation of the characteristic gases corresponding to all time series; Based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil, calculating the mean Euler distance and the corresponding standard deviation between each time series sample and the average value of the corresponding characteristic gas of all sequences; The mean of the difference sequence of adjacent time series samples and the corresponding standard deviation are calculated based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil.

3. The method for identifying online monitoring data of dissolved gas in oil according to claim 2, characterized in that: The calculating of the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil comprises: The mean, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value of all characteristic gases in the series are used as reference series to establish a grey relational analysis model for the mean of time series. Taking the mean of Euler distance, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for Euler distance between time series is established. Taking the mean of the difference sequence of adjacent time series samples, the upper limit of normal value, the lower limit of normal value, and the upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for the difference between time series is established; Substitute the dissolved gas monitoring data to be detected into the three established grey correlation analysis models in turn, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey correlation sequence and correlation value of each model; The dissolved gas monitoring data to be detected is the monitoring data of dissolved gas in oil at a certain moment: hydrogen, methane, ethane, ethylene and acetylene.

4. The method for identifying online dissolved gas in oil monitoring data according to claim 3, characterized in that: Identifying the status of the dissolved gas monitoring data to be detected according to the correlation value includes: Determine the state corresponding to the maximum value in each gray correlation degree sequence. If the state corresponding to at least two maximum values ​​is normal, then the state of the monitoring data of dissolved gas to be detected in oil is identified as normal. If the state corresponding to at least two maximum values ​​is abnormal, then the state of the monitoring data of dissolved gas to be detected in oil is identified as abnormal.

5. The online monitoring data identification system for dissolved gas in oil is characterized by: The system is used to implement the method for online monitoring data identification of dissolved gas in oil according to any one of claims 1 to 4, comprising: A reference sequence construction module is used to construct a reference sequence of dissolved gas in oil based on the multidimensional statistical characteristics of a time series sample of online monitoring data of dissolved gas in oil of an oil-immersed transformer; the time series sample of the online monitoring data is composed of the content characteristics of five characteristic gases, including hydrogen, methane, ethane, ethylene and acetylene; the multidimensional statistical characteristics include: the mean of the characteristic gases corresponding to all time series, the standard deviation of the characteristic gases corresponding to all time series, the mean of the Euler distance between the means of the characteristic gases corresponding to all time series, the standard deviation between the means of the characteristic gases corresponding to all time series, the mean of the difference sequence of adjacent time series samples, and the standard deviation corresponding to the mean of the difference sequence of adjacent time series samples; the specific method of constructing the reference sequence of dissolved gas in oil is: based on the Laita rule, establishing the upper and lower limits of the normal value of the online monitoring data of dissolved gas in oil; based on the Laita rule, establishing the upper and lower limits of the abnormal value of the online monitoring data of dissolved gas in oil; and using the time series of the calculated means, upper and lower limits of the normal value, and upper and lower limits of the abnormal value as the reference sequence of dissolved gas in oil; A correlation calculation module is used to calculate the correlation value between the monitoring data of the dissolved gas to be detected in the oil-immersed transformer oil and the reference sequence of the dissolved gas in the oil; The identification module is used to identify the status of the dissolved gas monitoring data to be detected according to the correlation value, and the status includes normal and abnormal.

6. The online monitoring data identification system for dissolved gas in oil according to claim 5, characterized in that: Also includes: The raw data acquisition module is used to determine the time window based on the online monitoring sampling period of dissolved gas in oil of the oil-immersed transformer, and to establish a time series sample set of online monitoring data of dissolved gas in oil in chronological order; Applying square root transformation to the time series sample set of online monitoring data of dissolved gas in oil to reconstruct the time series sample set of online monitoring data of dissolved gas in oil; as well as, A feature calculation module is used to calculate the mean and standard deviation of the characteristic gases corresponding to all sequences based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil; Calculating the mean Euler distance and the corresponding standard deviation between each time series sample and the mean of the corresponding characteristic gas of all sequences based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil; Calculate the mean and corresponding standard deviation of the difference value sequence of adjacent time series samples based on the reconstructed time series sample set of the online monitoring data of dissolved gas in oil; Based on the mean of the characteristic gases corresponding to all sequences, the standard deviation of the characteristic gases corresponding to all sequences, the mean of the Euler distances between the means of the characteristic gases corresponding to all sequences, the standard deviation between the means of the characteristic gases corresponding to all sequences, the mean of the difference sequence of adjacent time series samples and the standard deviation corresponding to the mean of the difference sequence of adjacent time series samples, the multidimensional statistical characteristics of the time series samples of the online monitoring data of dissolved gas in oil are constructed.

7. The online monitoring data identification system for dissolved gas in oil according to claim 6, characterized in that: The association calculation module is specifically used to: The mean, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value of all characteristic gases in the series are used as reference series to establish a grey relational analysis model for the mean of time series. Taking the mean of Euler distance, upper limit of normal value, lower limit of normal value, upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for Euler distance between time series is established. Taking the mean of the difference sequence of adjacent time series samples, the upper limit of normal value, the lower limit of normal value, and the upper limit and lower limit of abnormal value as reference sequences, a grey relational analysis model for the difference between time series is established; Substitute the dissolved gas monitoring data to be detected into the three established grey correlation analysis models in turn, calculate the correlation between the dissolved gas monitoring data to be detected and the reference sequence, and obtain the grey correlation sequence and correlation value of each model; The dissolved gas monitoring data to be detected is the monitoring data of dissolved gas in oil at a certain moment: hydrogen, methane, ethane, ethylene and acetylene.

8. The online monitoring data identification system for dissolved gas in oil according to claim 7, characterized in that: The identification module is specifically used to: Determine the state corresponding to the maximum value in each grey relational degree sequence; If the states corresponding to at least two maximum values ​​are normal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as normal. If the states corresponding to at least two maximum values ​​are abnormal, the state of the monitoring data of the dissolved gas to be detected in the oil is identified as abnormal.

9. A computer-readable memory storing one or more programs, characterized in that: The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any one of the methods according to claims 1 to 4 .

10. A device, characterized in that: include, One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for executing any one of the methods according to claims 1 to 4.

Images