An online fault combined diagnosis method and system for power transformer

Abstract

The application discloses an online fault combined diagnosis method and system of a power transformer, and the method comprises the following steps: initially determining a transformer monitoring state according to oil chromatogram data; constructing a transformer fault case library according to transformer fault characteristic data; calculating a case correlation degree to determine a fault type; and further positioning a fault according to other online monitoring state quantities. The application constructs the transformer fault case library, calculates the case correlation degree to determine the fault type, and further positions the fault according to the other online monitoring state quantities, so that the transformer fault diagnosis precision is effectively improved, and the application has important theoretical and practical significance for guaranteeing equipment safety and reducing accident loss.

Description

TECHNICAL FIELD

[0001] The present application relates to power equipment state analysis technology, and particularly relates to an online fault joint diagnosis method and system for a power transformer. BACKGROUND

[0002] The power transformer is an important device in the power system, and undertakes the important work of power grid interaction and energy transmission. The device state of the power transformer has a huge impact on system power supply reliability, system stable operation and normal production and life of users. Therefore, research on power transformer device state evaluation and fault diagnosis is of great significance to maintain the safe and stable operation of the power transformer and normal power supply of the system. At present, the online fault diagnosis method for the power transformer is still mainly based on the judgment method of the dissolved gas in the oil. However, this method does not fully consider the existing related online monitoring data, resulting in inaccurate diagnosis results. SUMMARY

[0003] The present application aims to provide an online fault joint diagnosis method and system for a power transformer, which can improve the fault diagnosis accuracy.

[0004] The technical scheme is as follows: the online fault joint diagnosis method for a power transformer comprises the following steps:

[0005] (1) determining the monitoring state of the transformer according to the oil chromatographic data;

[0006] (2) constructing a transformer fault case library according to the transformer fault feature data;

[0007] (3) calculating the case correlation degree to determine the fault type;

[0008] (4) further fault positioning according to other online monitoring state quantities.

[0009] Preferably, the step (1) comprises: according to the oil chromatographic data, if at least the first condition and the second condition are met, the transformer is monitored and the fault diagnosis process is entered;

[0010] if any one of the first condition and the second condition and at least one of the third condition and the fourth condition are met, the transformer is continuously monitored and the fault diagnosis process is entered;

[0011] if only any one of the first condition, the second condition, the third condition and the fourth condition is met, or the third condition and the fourth condition are met at the same time, the transformer is in normal operation and is monitored according to the period;

[0012] the first condition is that the content of the characteristic gas exceeds a first threshold value;

[0013] the second condition is that the gas production rate of the characteristic gas exceeds a second threshold value;

[0014] The third condition is that there is acetylene characteristic gas generation;

[0015] The fourth condition is that the characteristic gas ratio code is upgraded.

[0016] Preferably, the fault case library in step (2) is stored in a matrix form as:

[0017]

[0018] Wherein, X is a fault characteristic matrix, x i,j is the i-th fault characteristic data of the j-th fault case, n is the number of fault characteristics, m is the number of fault cases, Y is a fault result matrix, y j is the fault result data of the j-th fault case.

[0019] Preferably, the step (3) comprises:

[0020] (3.1) calculating the absolute distance between the data to be diagnosed and the transformer fault case library data, and the calculation result is stored in a matrix form as:

[0021] D = [d i,j ] n×m

[0022] Wherein, D is an absolute distance matrix, d i,j is the absolute distance between the data to be diagnosed and the fault characteristic data, d i,j = |x i,j -z i |, x i,j is the corresponding fault characteristic data, and z i is the corresponding data to be diagnosed.

[0023] (3.2) calculating the correlation coefficient between the data to be diagnosed and the transformer fault case library data, and obtaining a correlation matrix G

[0024] G = [g j ] 1×m

[0025] Wherein, G is a correlation matrix, g j is the corresponding correlation;

[0026] The corresponding correlation g j is:

[0027]

[0028] Wherein, r i,j is the corresponding correlation coefficient;

[0029] The corresponding correlation coefficient r i,jFor:

[0030]

[0031] wherein, ρ is a resolution coefficient, D max is a maximum distance, D min is a minimum distance;

[0032] The maximum distance and the minimum distance are respectively:

[0033] wherein, d i,j is an absolute distance between the data to be diagnosed and the fault feature data, MAX{} is a maximum value operator, and MIN{} is a minimum value operator.

[0034] Preferably, the resolution coefficient is:

[0035]

[0036] wherein, ρ is a resolution coefficient, α1 is a lower limit weight, α2 is an upper limit weight, D m is a distance mean, D max is a maximum distance, D min is a minimum distance;

[0037] wherein, the distance mean D m is:

[0038] Preferably, the step (4) further fault positioning comprises:

[0039] If the diagnosis result is a discharge type fault, for an arc discharge or arc discharge and overheat type fault, the transformer needs to be stopped for inspection; for other discharge type faults, the transformer needs to be continuously monitored;

[0040] If the diagnosis result is an overheat type fault, it is further judged whether the characteristic gas change is related to the load, if the characteristic gas change is related to the load, it is positioned as a circuit overheat type fault; if the characteristic gas change is not related to the load, it is positioned as a magnetic circuit overheat type fault.

[0041] Preferably, the method for judging whether the characteristic gas change is related to the load is:

[0042] The characteristic gas and load correlation coefficient is calculated:

[0043]

[0044] wherein, μ AB is a characteristic gas and load correlation coefficient, A k is a kth characteristic gas sequence, B k Y iN is the sequence length for the kth load current sequence;

[0045] When the absolute value of the characteristic gas-load correlation coefficient is greater than the third threshold μ0, the overheat type fault is considered to be related to the load; when the absolute value of the characteristic gas-load correlation coefficient is less than the third threshold μ0, the overheat type fault is considered to be unrelated to the load.

[0046] Preferably, for the circuit overheat type fault, if the CO and / or CO2 characteristic gas grows too fast, it is located as overheat inside the winding, otherwise, it is located as bare metal heating;

[0047] For the magnetic circuit overheat type fault, if the core grounding current is higher than the attention value, it is located as the core multi-point grounding fault; otherwise, it is located as the core internal circulating current or magnetic shielding fault.

[0048] Preferably, the overheat area is located by infrared online monitoring data, if the sleeve capacitance and dielectric loss online monitoring data are abnormal, it is located as the sleeve fault; if the submersible pump state online monitoring data is abnormal, it is located as the submersible pump fault.

[0049] The power transformer online fault joint diagnosis system provided by the application comprises:

[0050] The transformer monitoring state judgment module is used for determining the transformer monitoring state according to the oil chromatographic data according to the corresponding criterion.

[0051] The transformer fault case library construction module is used for constructing the transformer fault case library according to the transformer fault characteristic data.

[0052] The case correlation degree calculation module is used for calculating the correlation degree of the to-be-diagnosed data and the related fault cases according to the to-be-diagnosed data, and determining the transformer fault.

[0053] The fault positioning module is used for further fault positioning in combination with other online monitoring state quantities.

[0054] The electronic device provided by the application comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the computer program realizes the power transformer online fault joint diagnosis method described above when loaded to the processor.

[0055] The computer readable storage medium provided by the application stores a computer program, and the computer program realizes the power transformer online fault joint diagnosis method described above when executed by the processor.

[0056] Beneficial effects: compared with the prior art, the present application has the following remarkable advantages: by constructing a transformer fault case library, calculating the case correlation degree to determine the fault type, and further fault positioning according to other online monitoring state quantities, the transformer fault diagnosis precision is effectively improved, which has important theoretical and practical significance for ensuring equipment safety and reducing accident loss. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 The flow chart of the transformer fault diagnosis method of the present application;

[0058] Figure 2 The flow chart of the transformer fault positioning of the present application. DETAILED DESCRIPTION

[0059] The technical solutions of the present application are further described below in combination with the drawings.

[0060] As shown in the drawings, the power transformer online fault joint diagnosis method of the present application comprises the following steps: Figure 1 (1) determining the transformer monitoring state according to the oil chromatogram data.

[0061] The transformer monitoring state is determined according to the following four criteria:

[0062] 1) judging whether the content of characteristic gas exceeds the first threshold value according to the current oil chromatogram data.

[0063] 2) judging whether the gas production rate of characteristic gas exceeds the second threshold value according to the historical oil chromatogram data, wherein the gas production rate is calculated according to the following formula:

[0064]

[0065]

[0066] Wherein, γ is the relative gas production rate, C i,1 is the concentration of a certain gas in the oil measured by the first sampling, C i,2 is the concentration of a certain gas in the oil measured by the second sampling, and Δt is the actual running time interval between the two samplings.

[0067] 3) judging whether acetylene characteristic gas is generated according to the current oil chromatogram data.

[0068] 4) judging whether the characteristic gas ratio code is upgraded according to the historical oil chromatogram data. Wherein, the characteristic gas ratios are C2H2 / C2H4, CH4 / H2 and C2H4 / C2H6, and the coding is according to the following rules.

[0069] For C2H2 / C2H4, if the ratio is less than 0.1, the code is 0, if the ratio is between 0.1 and 3, the code is 1, and if the ratio is greater than 3, the code is 2.​

[0070] For CH4 / H2, if the ratio is less than 0.1, the code is 1, if the ratio is between 0.1 and 1, the code is 0, and if the ratio is greater than 1, the code is 2.

[0071] For C2H4 / C2H6, if the ratio is less than 1, the code is 0, if the ratio is between 1 and 3, the code is 1, and if the ratio is greater than 3, the code is 2.

[0072] According to the above chromatographic data of each oil, the transformer state is classified and processed, which can be divided into the following three cases:

[0073] Case 1, meet 1), 2) two criteria, 1), 2), 3) and 1), 2), 4) three criteria, 1), 2), 3), 4) four criteria, transformer monitoring, into the fault diagnosis process.

[0074] Case 2, meet 1), 3), 1), 4), 2), 3) and 2), 4) two criteria, 1), 3), 4) and 2), 3), 4) three criteria, transformer continues to monitor, into the fault diagnosis process.

[0075] Case 3, meet 1), 2), 3), 4) one of the criteria, and 3), 4) two criteria, transformer normal operation, monitoring according to the period.

[0076] (2) According to the transformer fault feature data, the transformer fault case library is constructed, and the fault case library and the to-be-diagnosed data are stored in matrix form,

[0077] The specific structure of the fault case library is:

[0078]

[0079] Among them, X is the fault feature matrix, x i,j is the corresponding fault feature data, n is the number of fault features, m is the number of fault cases, Y is the fault result matrix, y j is the corresponding fault result data.

[0080] The specific structure of the to-be-diagnosed data can be expressed as: Z=[z i ] n×1

[0081] Among them, Z is the to-be-diagnosed data matrix, z i is the corresponding to-be-diagnosed data.

[0082] (3) Calculate the case correlation degree to determine the fault type.

[0083] Firstly, the absolute distance between the data to be diagnosed and the transformer fault case library data is calculated, and the calculation result is stored in a matrix form, and the specific structure can be represented as:

[0084] D=[d i,j ] n×m

[0085] wherein, D is the absolute distance matrix, d i,j is the absolute distance between the data to be diagnosed and the fault feature data, d i,j =|x i,j -z i |.

[0086] The maximum and minimum distances are represented as:

[0087]

[0088] wherein, D max is the maximum distance, MAX{} is the maximum value operator, D min is the minimum distance, and MIN{} is the minimum value operator.

[0089] The corresponding distance mean D m is represented as:

[0090]

[0091] The corresponding resolution coefficient p is represented as:

[0092]

[0093] wherein, a1 is the lower limit weight, and a2 is the upper limit weight.

[0094] Secondly, the correlation coefficient between the data to be diagnosed and the transformer fault case library data is calculated, and the calculation result is stored in a matrix form, and the specific structure is represented as:

[0095] R=[r i,j ] n×m

[0096] wherein, R is the correlation coefficient matrix, and r i,j is the corresponding correlation coefficient.

[0097] The corresponding correlation coefficient is represented as:

[0098]

[0099] Finally, the correlation degree between the data to be diagnosed and the transformer fault case library data is calculated, and the calculation result is stored in a matrix form, and the specific structure is represented as:

[0100] G=[g j ]1×m

[0101] wherein G is a correlation matrix; g j is the corresponding correlation degree, and

[0102] (4) Further fault location according to other online monitoring state quantities.

[0103] As Figure 2 shown, if the diagnosis result is a discharge type fault, the transformer needs to be shut down for inspection for arc discharge or arc discharge and overheating type faults; and the transformer needs to be continuously monitored for other discharge type faults.

[0104] If the diagnosis result is an overheating type fault, it is further judged whether the characteristic gas change is related to the load. If the characteristic gas change is related to the load, it is located as a circuit overheating type fault; and if the characteristic gas change is not related to the load, it is located as a magnetic circuit overheating type fault.

[0105] The method for judging whether the characteristic gas change is related to the load is:

[0106] Calculate the characteristic gas and load correlation coefficient:

[0107]

[0108] wherein μ AB is the characteristic gas and load correlation coefficient, A k is the kth characteristic gas sequence, B k Y i is the kth load current sequence, and N is the sequence length.

[0109] When the absolute value of the characteristic gas and load correlation coefficient is greater than a threshold μ0, the overheating type fault is considered to be related to the load; and when the absolute value of the characteristic gas and load correlation coefficient is less than the threshold μ0, the overheating type fault is considered to be not related to the load.

[0110] For the circuit overheating type fault, if the CO and / or CO2 characteristic gas grows too fast, it is located as a winding internal overheating; and otherwise, it is located as a bare metal heating.

[0111] For the magnetic circuit overheating type fault, if the core grounding current is higher than a value of attention, it is located as a core multi-point grounding fault; and otherwise, it is located as a core internal circulating current or magnetic shielding fault.

[0112] In addition, the overheating region is located through infrared online monitoring data. If the sleeve capacitance and dielectric loss online monitoring data are abnormal, it is located as a sleeve fault; and if the submersible pump state online monitoring data is abnormal, it is located as a submersible pump fault.

[0113] The application further discloses an online fault combined diagnosis system of a power transformer, which comprises:

[0114] A transformer monitoring state distinguishing module is used for determining the monitoring state of the transformer according to the oil chromatographic data and corresponding criteria.

[0115] A transformer fault case library constructing module is used for constructing a transformer fault case library according to transformer fault feature data.

[0116] A case correlation degree calculating module is used for calculating the correlation degree between the to-be-diagnosed data and the related fault cases according to the to-be-diagnosed data, and determining the transformer fault.

[0117] A fault positioning module is used for further fault positioning in combination with other online monitoring state quantities.

[0118] The application further discloses an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the computer program realizes the online fault combined diagnosis method of the power transformer when being loaded to the processor.

[0119] The application further discloses a computer readable storage medium, which stores a computer program, and the computer program realizes the online fault combined diagnosis method of the power transformer when being executed by a processor.

Claims

1. A method for joint online fault diagnosis of power transformers, characterized in that, include: (1) The transformer monitoring status was initially determined based on oil chromatography data; (2) Construct a transformer fault case library based on transformer fault characteristic data; (3) Calculate the case correlation degree to determine the fault type; (4) Further fault location is performed based on the online monitoring status data; The further fault location in step (4) includes: If the diagnosis is a discharge-type fault, the transformer needs to be shut down for inspection if it is an arc discharge or an arc discharge combined with overheating fault; for other discharge-type faults, the transformer needs to be continuously monitored. If the diagnosis result is an overheating fault, further determine whether the characteristic gas change is related to the load. If the characteristic gas change is related to the load, then locate it as an overheating fault in the circuit; if the characteristic gas change is not related to the load, then locate it as an overheating fault in the magnetic circuit.

2. The online fault joint diagnosis method for power transformers according to claim 1, characterized in that, The step (1) includes: based on the oil chromatography data, if at least the first and second conditions are met, the transformer is monitored in a key manner and the fault diagnosis process is initiated. If either the first or second condition, or at least one of the third or fourth conditions is met, the transformer will be continuously monitored and the fault diagnosis process will begin. If any one of the first, second, third, and fourth conditions is met, or if the third and fourth conditions are met simultaneously, the transformer will operate normally and be monitored periodically. The first condition is: the content of the characteristic gas exceeds a first threshold; The second condition is: the characteristic gas production rate exceeds the second threshold. The third condition is: the generation of acetylene characteristic gas; The fourth condition is: upgrade of characteristic gas ratio encoding.

3. The online fault joint diagnosis method for power transformers according to claim 1, characterized in that, The fault case library in step (2) is stored in a matrix format, as follows: Where X is the fault feature matrix, x i,j Let y be the fault feature data of the j-th fault case, n be the number of fault features, m be the number of fault cases, and Y be the fault result matrix. j This represents the failure result data for the j-th failure case.

4. The online fault joint diagnosis method for power transformers according to claim 1, characterized in that, Step (3) includes: (3.1) Calculate the absolute distance between the data to be diagnosed and the data in the transformer fault case database. The calculation results are stored in a matrix format as follows: D=[d i,j ] n×m Where D is the absolute distance matrix, d i,j d represents the absolute distance between the data to be diagnosed and the fault characteristic data. i,j =|x i,j -z i |,x i,j For the corresponding fault characteristic data, z i For the corresponding diagnostic data; (3.2) Calculate the correlation coefficient between the data to be diagnosed and the data in the transformer fault case database to obtain the correlation matrix. G=[g j ] 1×m Where G is the correlation matrix, g j For the corresponding degree of correlation; Corresponding correlation degree g j for: Where, r i,j This corresponds to the correlation coefficient; The corresponding correlation coefficient r i,j for: Where ρ is the resolution coefficient, and D max For the maximum distance, D min Minimum distance; The maximum distance and minimum distance are respectively: Where, d i,j This represents the absolute distance between the data to be diagnosed and the fault characteristic data. MAX{} is the maximum value operator, and MIN{} is the minimum value operator.

5. The online fault joint diagnosis method for power transformers according to claim 4, characterized in that, The resolution coefficient is: Where ρ is the resolution coefficient, α1 is the lower limit weight, α2 is the upper limit weight, and D... m D is the distance mean. max For the maximum distance, D min Minimum distance; Wherein, the distance mean D m for:

6. The online fault joint diagnosis method for power transformers according to claim 1, characterized in that, The method for determining whether changes in characteristic gases are related to load is as follows: Calculate the correlation coefficient between characteristic gas and load: Where, μ AB A is the correlation coefficient between the characteristic gas and the load. k For the k-th characteristic gas sequence, B k Y i Let N be the k-th load current sequence, and N be the sequence length. When the absolute value of the correlation coefficient between the characteristic gas and the load is greater than the third threshold μ0, the overheating fault is considered to be related to the load; when the absolute value of the correlation coefficient between the characteristic gas and the load is less than the third threshold μ0, the overheating fault is considered to be unrelated to the load.

7. The online fault joint diagnosis method for power transformers according to claim 6, characterized in that, For circuit overheating faults, if the growth rate of CO and / or CO2 characteristic gases exceeds the specified threshold, it is identified as overheating inside the winding; otherwise, it is identified as overheating of bare metal. For magnetic circuit overheating faults, if the core grounding current is higher than the warning value, it is identified as a core multi-point grounding fault; otherwise, it is identified as an internal circulating current or magnetic shielding fault.

8. The online fault joint diagnosis method for power transformers according to claim 7, characterized in that, The overheated area is located by infrared online monitoring data. If the online monitoring data of the bushing capacitance and dielectric loss are abnormal, the bushing is identified as having a fault. If the online monitoring data of the submersible pump status is abnormal, the submersible pump is identified as having a fault.

9. A power transformer online fault joint diagnosis system capable of implementing the power transformer online fault joint diagnosis method according to any one of claims 1-8, characterized in that, include: The transformer monitoring status determination module is used to determine the transformer monitoring status based on oil chromatography data and corresponding criteria. The transformer fault case library construction module is used to build a transformer fault case library based on transformer fault characteristic data. The case correlation calculation module is used to calculate the correlation between the data to be diagnosed and related fault cases based on the data to be diagnosed, and to determine the transformer fault. The fault location module is used to combine other online monitoring status variables for further fault location.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is loaded into the processor, it implements the online fault joint diagnosis method for power transformers according to any one of claims 1-8.

11. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the online fault joint diagnosis method for power transformers according to any one of claims 1-8.

Images