Online monitoring device for dissolved gas in transformer oil based on remote data analysis

Abstract

The present invention discloses an online monitoring device for dissolved gas in transformer oil based on remote data analysis, which belongs to the technical field of transformer monitoring. The device comprises: a transformer oil cylinder, an oil pump, an oil sample quantitative collection module, an oil-gas separation module, a gas collection module, a first vacuum pump, a gas optical recognition module, a buffer oil return module, a first valve, a second valve, a third valve, a fourth valve, a fifth valve, a communication module and a controller; when the first valve is opened, the second valve is closed, the third valve is opened, the fourth valve is opened, and the fifth valve is opened, the oil pump is opened to clean the oil circuit, and the first vacuum pump is opened to clean the gas circuit; when the first valve is opened, the second valve is opened, the third valve is opened, the fourth valve is closed, and the fifth valve is closed, the dissolved gas in the transformer oil is detected, and the detection result is sent to a server through the communication module, and the server performs remote data analysis to monitor the status of the transformer.

Description

Technical Field

[0001] The present invention belongs to the technical field of transformer monitoring, and in particular relates to an online monitoring device for dissolved gas in transformer oil based on remote data analysis. Background Art

[0002] Transformers are an important link in the power grid system. Once a fault occurs, it may cause major accidents and disasters in the entire power system. In the daily operation of the transformer, its operating status must be detected to provide early warning of transformer faults and timely detect potential transformer faults. This is crucial for the safe and reliable operation of the entire power system, reducing fault repair time and reducing economic losses.

[0003] Most high-voltage transformers, both domestically and internationally, use oil-immersed transformers. Transformer oil is a mixture of long-chain or cyclic hydrocarbon chains extracted from petroleum. When a transformer experiences certain faults, characteristic gases are generated within the oil. For example, electrical and thermal faults produce gases like hydrogen, ethylene, and acetylene. Transformer oil oxidation can lead to the accumulation of carbon monoxide and carbon dioxide. Therefore, analyzing the composition and concentration of gases in transformer oil is one of the key criteria for diagnosing transformer faults. The concentrations of CH4, C2H6, C2H4, C2H2, CO, CO2, and H2 are crucial indicators for determining internal faults in power transformers.

[0004] Currently, the most common method for analyzing dissolved gases in transformer oil is gas chromatography. However, gas chromatography columns are susceptible to contamination and require regular calibration and even column replacement, resulting in high monitoring costs. Furthermore, residual gas from previous tests often remains in the gas chamber, leading to inaccurate monitoring results and prone to false alarms and missed alarms. Summary of the Invention

[0005] In order to solve the technical problems that the existing technology adopts gas chromatography, the gas chromatography column is easily contaminated and needs to be calibrated regularly or even replaced, resulting in excessively high monitoring costs, and the gas chamber often retains gas from the previous test, resulting in inaccurate monitoring results and prone to false alarms and missed alarms, the present invention provides an online monitoring device for dissolved gas in transformer oil based on remote data analysis.

[0006] The present invention provides an online monitoring device for dissolved gas in transformer oil based on remote data analysis, comprising: a transformer oil cylinder, an oil pump, an oil sample quantitative collection module, an oil-gas separation module, a gas collection module, a first vacuum pump, a gas optical identification module, a buffer oil return module, a first valve, a second valve, a third valve, a fourth valve, a fifth valve, a communication module and a controller, wherein the gas optical identification module includes an optical sensor;

[0007] The oil outlet of the transformer oil cylinder is connected to the oil inlet of the oil sample quantitative collection module through the oil pump, the oil outlet of the oil sample quantitative collection module is connected to the oil inlet of the oil and gas separation module through the first valve, the oil outlet of the oil and gas separation module is connected to the oil inlet of the buffer oil return module, and the oil outlet of the buffer oil return module is connected to the oil inlet of the oil cylinder; the gas outlet of the oil and gas separation module is connected to the gas inlet of the gas collection module through the second valve, the gas outlet of the gas collection module is connected to the gas inlet of the gas optical recognition module through the third valve, and the gas outlet of the gas optical recognition module is connected to the outside world through the fourth valve; the first vacuum pump is connected to the gas inlet of the gas collection module through the fifth valve; the communication module establishes a communication connection with the server;

[0008] The transformer oil cylinder, oil pump, first valve, oil sample quantitative collection module, oil and gas separation module, and buffer oil return module constitute the oil circuit; the gas collection module and gas optical recognition module constitute the gas circuit;

[0009] Among them, under the control of the controller, when the first valve is opened, the second valve is closed, the third valve is opened, the fourth valve is opened, and the fifth valve is opened, the oil pump is turned on to clean the oil circuit, and the first vacuum pump is turned on to clean the gas circuit; under the control of the controller, when the first valve is opened, the second valve is opened, the third valve is opened, the fourth valve is closed, and the fifth valve is closed, the optical sensor is used to detect the dissolved gas in the transformer oil, and the detection result is sent to the server through the communication module, and the server performs remote data analysis to monitor the status of the transformer.

[0010] Compared with the prior art, the present invention has at least the following beneficial effects:

[0011] (1) In the present invention, before the formal test begins, an oil pump can be used to clean the oil circuit and a vacuum pump can be used to clean the gas circuit to avoid residual gas from the previous test in the gas chamber, thereby improving the accuracy of dissolved gas detection.

[0012] (2) In the present invention, the gas composition and concentration in the dissolved gas are accurately identified by an optical sensor, and the detection results are sent to the server through the communication module. The server performs remote data analysis to monitor the status of the transformer, avoiding the use of gas chromatography for measurement, eliminating the need for frequent replacement of chromatographic columns, and reducing monitoring costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The preferred embodiments will be described below in a clear and understandable manner with reference to the accompanying drawings to further illustrate the above-mentioned characteristics, technical features, advantages and implementation methods of the present invention.

[0014] Figure 1 This is a schematic structural diagram of an online monitoring device for dissolved gas in transformer oil based on remote data analysis provided by the present invention;

[0015] Figure 2 This is a structural schematic diagram of a magnetic stirrer provided by the present invention;

[0016] Figure 3 This is a structural diagram of a buffer net provided by the present invention;

[0017] Figure 4 It is a structural schematic diagram of a gas optical recognition module provided by the present invention. DETAILED DESCRIPTION

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific embodiments of the present invention will be described below with reference to the accompanying drawings. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings and other embodiments can be obtained based on these drawings without inventive work.

[0019] To simplify the drawings, only portions relevant to the invention are schematically depicted in each figure; they do not represent the actual structure of the product. Furthermore, to simplify the drawings and facilitate understanding, in some figures, only one component with the same structure or function is schematically depicted or labeled. In this document, "one" not only means "only one" but also "more than one."

[0020] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0021] It should be noted that, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be understood broadly. For example, they can refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediary; and internal communication between two components. Those skilled in the art will understand the specific meanings of these terms in the present invention based on the specific circumstances.

[0022] In addition, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0023] In one embodiment, the reference Figure 1 , a structural schematic diagram of an online monitoring device for dissolved gas in transformer oil based on remote data analysis provided by the present invention.

[0024] The present invention provides an online monitoring device for dissolved gas in transformer oil based on remote data analysis, comprising: a transformer oil cylinder 1, an oil pump 2, an oil sample quantitative collection module 3, an oil-gas separation module 4, a gas collection module 5, a first vacuum pump 6, a gas optical recognition module 7, a buffer oil return module 8, a first valve 9, a second valve 10, a third valve 11, a fourth valve 12, a fifth valve 13, a communication module and a controller, wherein the gas optical recognition module 7 includes an optical sensor 72.

[0025] It should be noted that the oil sample quantitative collection module 3 is used to extract a rated volume of oil sample to be tested during a single test. The oil-gas separation module 4 is used to separate the oil and gas in the test sample. The gas collection module 5 is used to collect the separated gas. The gas optical identification module 7 is used to identify the composition and concentration of the test gas. The buffered oil return module 8 is used to return the separated oil to the transformer oil cylinder 1.

[0026] Specifically, the oil outlet of the transformer oil cylinder 1 is connected to the oil inlet of the quantitative oil sample collection module 3 via the oil pump 2. The oil outlet of the quantitative oil sample collection module 3 is connected to the oil inlet of the oil-gas separation module 4 via the first valve 9. The oil outlet of the oil-gas separation module 4 is connected to the oil inlet of the buffer oil return module 8, and the oil outlet of the buffer oil return module 8 is connected to the oil inlet of the oil cylinder. The gas outlet of the oil-gas separation module 4 is connected to the gas inlet of the gas collection module 5 via the second valve 10. The gas outlet of the gas collection module 5 is connected to the gas inlet of the gas optical identification module 7 via the third valve 11. The gas outlet of the gas optical identification module 7 is connected to the outside world via the fourth valve 12. The first vacuum pump 6 is connected to the gas inlet of the gas collection module 5 via the fifth valve 13.

[0027] The oil pump 2, the first vacuum pump 6, the first valve 9, the second valve 10, the fourth valve 12, the fifth valve 13 and the communication module are all electrically connected to the control module, and the control module can control the oil pump 2, the first vacuum pump 6, the first valve 9, the second valve 10, the fourth valve 12, the fifth valve 13 and the communication module.

[0028] The communication module establishes a communication connection with the server, and the communication module can send the detection results to the server so that the server can make maintenance personnel scheduling based on the detection results.

[0029] The transformer oil cylinder 1, oil pump 2, first valve 9, oil sample quantitative collection module 3, oil-gas separation module 4, and buffer oil return module 8 form the oil circuit. The gas collection module 5 and gas optical identification module 7 form the gas circuit. The second valve 10 isolates the oil circuit from the gas circuit.

[0030] Under the control of the controller, the first valve 9 is opened, the second valve 10 is closed, the third valve 11 is opened, the fourth valve 12 is opened, and the fifth valve 13 is opened. The oil pump 2 is turned on to clean the oil circuit, and the first vacuum pump 6 is turned on to clean the gas circuit. Before the formal test begins, the oil pump 2 can be used to clean the oil circuit to prevent residual transformer oil from the previous test in the oil circuit, and the vacuum pump can be used to clean the gas circuit to prevent residual gas from the previous test in the gas chamber, thereby improving the accuracy of dissolved gas detection.

[0031] Under the control of the controller, when the first valve 9, the second valve 10, the third valve 11 are opened, the fourth valve 12, and the fifth valve 13 are closed, the optical sensor 72 detects dissolved gas in the transformer oil. The detection results are sent to the server via the communication module. The server performs remote data analysis to monitor the status of the transformer. This avoids the need for gas chromatography measurements, eliminates the need for frequent replacement of chromatographic columns, and reduces monitoring costs.

[0032] It should be noted that the gas optical identification module 7 can detect the gas composition and concentration in the gas to be tested. The concentration of CH4, C2H6, C2H4, C2H2, CO, CO2, and H2 gases contained in transformer oil is an important basis for judging internal faults of power transformers.

[0033] Common transformer failure types include overheating, partial discharge, spark discharge, and arc discharge. Different types of faults often produce different gases. Therefore, remote server analysis of the composition and concentration of dissolved gases in transformer oil can determine whether a transformer fault has occurred and what type of fault it is. The server can then dispatch maintenance personnel to the transformer for verification and maintenance as needed, significantly reducing transformer failure rates.

[0034] Compared with the prior art, the present invention has at least the following beneficial effects:

[0035] (1) In the present invention, before the formal test begins, the oil circuit can be cleaned using the oil pump 2 and the gas circuit can be cleaned using the vacuum pump to avoid gas from the previous test remaining in the gas chamber, thereby improving the accuracy of dissolved gas detection.

[0036] (2) In the present invention, the gas composition and concentration in the dissolved gas are accurately identified by the optical sensor 72, and the detection results are sent to the server through the communication module. The server performs remote data analysis to monitor the status of the transformer, avoiding the use of gas chromatography for measurement, eliminating the need for frequent replacement of chromatographic columns, and reducing monitoring costs.

[0037] In one possible embodiment, the oil sample quantitative collection module 3 includes a quantitative collector 31 and a liquid level sensor 32. The oil outlet of the transformer oil cylinder 1 is connected to the oil inlet of the quantitative collector 31 via the oil pump 2, and the oil outlet of the quantitative collector 31 is connected to the oil-gas separation module 4 via the first valve 9.

[0038] Among them, the first valve 9 is closed, and oil is supplied to the quantitative collector 31 through the oil pump 2. When the liquid level sensor 32 detects that the liquid level in the quantitative collector 31 has reached a preset liquid level, the first valve 9 is opened to allow the transformer oil in the quantitative collector 31 to be passed into the oil-gas separation module 4.

[0039] The liquid level sensor 32 and the first valve 9 cooperate to ensure that the rated capacity of transformer oil can be accurately extracted in each test.

[0040] In one possible embodiment, the oil-gas separation module 4 includes a degassing tank 41 and a first magnetic stirrer 42. The oil outlet of the oil sample quantitative collection module 3 is connected to the oil inlet of the degassing tank 41 via a first valve 9. The oil outlet of the degassing tank 41 is connected to the oil inlet of the buffer oil return module 8. The gas outlet of the degassing tank 41 is connected to the gas inlet of the gas collection module 5 via a second valve 10. The first magnetic stirrer 42 is located at the bottom of the degassing tank 41 and is used to stir the transformer oil in the degassing tank 41.

[0041] Reference Manual Figure 2 , a structural schematic diagram of a magnetic stirrer provided by the present invention.

[0042] The first magnetic stirrer 42 includes a stator 421 and a stirrer 422. The stator 421 includes a base 4211 and four electromagnetic components 4212. The four electromagnetic components 4212 are arranged in a 2×2 manner on the base 4211. The electromagnetic component 4212 includes a coil 42121, an iron core 42122 and a magnetic pole 42123. The coil 42121, the iron core 42122 and the magnetic pole 42123 constitute a magnetic circuit. The stirrer 422 is arranged on the four magnetic poles 42123. The stirrer 422 is a permanent magnet.

[0043] By changing the magnetic force generated by the four electromagnetic components 4212 and controlling the stirrer 422 to rotate smoothly, the transformer oil in the degassing tank 41 is stirred, so that the gas dissolved in the transformer oil is released as much as possible, thereby improving the accuracy of gas detection.

[0044] It should be noted that the magnetic stirrer does not need to come into contact with the transformer oil when stirring the transformer oil, which can avoid contaminating the transformer oil during the stirring process and causing distortion of the test results.

[0045] In one possible embodiment, the online monitoring device for dissolved gas in transformer oil further includes a sixth valve 14, and the gas collection module 5 includes a gas collecting tank 51 and a variable piston 52. A variable piston 52 is provided in the gas collecting tank 51. The variable piston 52 forms a gas collecting chamber at the top of the gas collecting tank 51, and the gas collecting chamber is provided with an air inlet and an air outlet. The air outlet of the oil-gas separation module 4 is connected to the air inlet of the gas collecting tank 51 through the second valve 10, and the air outlet of the gas collecting tank 51 is connected to the air inlet of the gas optical identification module 7 through the third valve 11. The variable piston 52 forms a vacuum chamber at the bottom of the gas collecting tank 51, and the first vacuum pump 6 is connected to the vacuum chamber through the sixth valve 14.

[0046] To collect gas from the oil-gas separation module 4, the sixth valve 14 is opened, and the first vacuum pump 6 is activated to evacuate the vacuum chamber. The sixth valve 14 is then closed, and the second valve 10 is opened. The variable piston 52 then moves to its bottom, pumping gas from the oil-gas separation module 4 into the gas collection chamber. To transfer gas from the gas collection tank 51 to the optical gas identification module 7, the second valve 10 is closed, and the third valve 11 is opened. The variable piston 52 then moves to its top, transferring gas from the gas collection tank 51 to the optical gas identification module 7.

[0047] It should be noted that collecting gas by drawing vacuum from the gas collecting tank 51 through the first vacuum pump 6 can, on the one hand, avoid the mixing of other gases during the collection process that may cause distortion of the detection results, and on the other hand, ensure that the gas in the oil-gas separation module 4 is completely collected in the gas collecting tank 51.

[0048] In the present invention, the first vacuum pump 6 is used to clean the gas path on the one hand, and to evacuate the vacuum chamber of the gas collecting tank 51 on the other hand, so as to extract the gas separated in the oil-gas separation module 4 into the gas collecting chamber of the gas collecting tank 51.

[0049] In one possible embodiment, the buffer oil return module 8 includes a buffer oil return tank 81, a second vacuum pump 82, a second magnetic stirrer 83, a funnel 84, a buffer screen 85, and a seventh valve 86. A degassing chamber 87 and an oil return chamber 88 are provided within the buffer oil return tank 81. The degassing chamber 87 is connected to the oil return chamber 88 via the seventh valve 86.

[0050] The second vacuum pump 82 is in communication with the degassing chamber 87, and the second magnetic stirrer 83 is disposed at the bottom of the degassing chamber 87. The second vacuum pump 82 is used to further degas the transformer oil in the degassing chamber 87 to exhaust as much gas as possible from the transformer oil.

[0051] The funnel 84 is arranged in the oil return chamber 88 , the oil inlet of the oil return chamber 88 is opposite to the inclined wall of the funnel 84 , the buffer net 85 is arranged at the oil outlet of the funnel 84 , and a plurality of oil return holes 851 are provided on the buffer net 85 .

[0052] Reference Manual Figure 3 , a schematic structural diagram of a buffer net 85 provided by the present invention.

[0053] It should be noted that the transformer oil first flows down along the inclined wall of the funnel 84 to perform the first pressure relief buffering, and then the transformer oil passes through the buffer net 85 for the second pressure relief buffering, which can prevent the transformer oil pressure from being too high and causing damage to the transformer oil cylinder 1 when it is returned.

[0054] In one possible embodiment, the gas optical identification module 7 includes a gas holding tank 71 and an optical sensor 72. The gas outlet of the gas collection module 5 is connected to the gas inlet of the gas holding tank 71 via a third valve 11, and the gas outlet of the gas holding tank 71 is connected to the outside world via a fourth valve 12. The optical sensor 72 is disposed within the gas holding tank 71 and includes a structured light emitter 721 and a structured light receiver 722 positioned oppositely on either side of the gas holding tank 71. The optical sensor 72 is used to determine the composition and concentration of the gas to be measured based on the intensity of the received light.

[0055] Reference Manual Figure 4 , a schematic structural diagram of a gas optical identification module 7 provided by the present invention.

[0056] In a possible implementation, the controller is configured to perform the following steps S101 to S104:

[0057] S101: Detecting the composition and concentration of the gas to be tested in the gas holding tank 71 through a gas detection model.

[0058] S102: When the concentration of the target gas contained in the gas to be tested is greater than the alarm threshold, it is determined that the transformer is in a dangerous state, and the detection result is reported to the server.

[0059] The concentrations of CH4, C2H6, C2H4, C2H2, CO, CO2, and H2 gases are important indicators for determining internal faults in power transformers. The target gas can be any one or more of CH4, C2H6, C2H4, C2H2, CO, CO2, and H2 gases.

[0060] It should be noted that different types of gases with excessive concentrations will result in different causes of internal transformer failures. Therefore, the composition of the gas being tested is also an important measurement that can help diagnose the type of transformer failure.

[0061] Furthermore, the server can perform remote data analysis to determine whether the transformer has failed.

[0062] S103: The server notifies the maintenance personnel to go to the destination to confirm and maintain the status of the transformer.

[0063] S104: Compare the actual result of the transformer status confirmed by the maintenance personnel after arriving at the destination with the detection result of the gas detection model, and modify the alarm threshold.

[0064] In this invention, a gas detection model accurately identifies the gas composition and concentration of dissolved gases, eliminating the need for gas chromatography measurements and frequent column replacement, thus reducing monitoring costs. Furthermore, if a transformer is detected to be in a dangerous state, maintenance personnel can be promptly notified to conduct on-site repairs, enabling real-time monitoring of transformer faults and contributing to grid security.

[0065] In one possible implementation, in order to determine whether the gas to be tested contains corresponding gas components, the controller is specifically configured to:

[0066] A data matrix X of multiple standard samples is obtained, where the multiple standard samples include m types of gases, and each type of gas includes n standard samples. The gas components of the standard samples are known, and the concentration is 100%.

[0067] Summarize to get the total sample matrix P,

[0068] According to the total sample matrix P, calculate the single class sample P i The mean matrix η of the column data i and the mean matrix λ of the total sample matrix P m :

[0069]

[0070]

[0071] in, Represents the eigenvalue of the jth standard sample of the i-th category.

[0072] Calculate the intra-class scatter matrix Q1 and inter-class scatter matrix Q2 of the total sample matrix P:

[0073]

[0074]

[0075] Construct the objective function φ(α) and add constraints to calculate the eigenvalue λ of the total sample matrix P when the intra-class scatter matrix Q1 is minimized and the inter-class scatter matrix Q2 is maximized, so as to distinguish the gas components:

[0076]

[0077] The constraints are:

[0078] α·Q2·α T =1

[0079] φ'(α)=α·Q1·α T -λ(α·Q1·α T -1)=0

[0080] Therefore, the eigenvalue λ of the total sample matrix P is:

[0081]

[0082] Calculate the total sample feature recognition matrix P':

[0083] P'=λ·P

[0084] Take the average of the characteristic values ​​of each type of sample in the total sample feature recognition matrix P' and calculate multiple gas classification matrices Y i :

[0085]

[0086] Based on the gas classification matrix Y i , build a gas detection model.

[0087] The characteristic matrix Z of the gas to be measured in the gas holding tank 71 is extracted.

[0088] Calculate each feature classification matrix Y i The Euclidean distance between the characteristic matrix Z of the gas to be measured is less than the preset distance, and it is determined that the gas to be measured contains the corresponding gas component.

[0089] Among them, those skilled in the art can set the size of the preset distance according to actual conditions so as to accurately identify the components of the gas to be measured. The present invention does not limit the specific value of the preset distance.

[0090] It should be noted that the concentration of each gas component in the gas to be measured can also be obtained in a similar manner.

[0091] In one possible implementation, in order to determine the gas concentration in the gas to be measured, the controller is specifically configured to:

[0092] Construct a gas detection model based on Beer's law:

[0093] I(λ)=I0(λ)e -σ(λ)cL

[0094] Wherein, I(λ) represents the light intensity received by the optical sensor 72, I0(λ) represents the initial light intensity, σ(λ) represents the absorption cross section of the measured substance, λ represents the wavelength, c represents the concentration of the measured substance, and L represents the length of the gas containing cavity.

[0095] By transforming the above formula, we can get:

[0096] cσ(λ)L=ln[I0(λ) / I(λ)]

[0097] Assuming that the gas to be tested contains K gas components and K test points are obtained in the absorption spectrum, then

[0098]

[0099] Where k = 1, 2,…, K.

[0100] Solve K unknown c by K equations k , in order to detect the concentration of each gas component in the gas to be tested.

[0101] In a possible implementation, in order to modify the alarm threshold, the controller is specifically configured to:

[0102] Compare the test results with the actual results and evaluate the test results, wherein the evaluation results include: changing the safety detection to a dangerous result, changing the dangerous detection to a safe result, changing the safety detection to a safe result, and changing the dangerous detection to a dangerous result.

[0103] Let the number of times the safe class result is predicted as safe be TX, the number of times the safe class result is predicted as dangerous be FY, the number of times the dangerous class result is predicted as dangerous be TY, and the number of times the dangerous class result is predicted as safe be FX, then the proportion of safe samples that are predicted incorrectly is FY rate and the proportion of dangerous samples that are predicted correctly FX rate for:

[0104]

[0105]

[0106] Let the cost of predicting a safe result as dangerous be p, and the cost of predicting a dangerous result as safe be q, and modify the alarm threshold so that:

[0107]

[0108] The cost of predicting a safety outcome as a dangerous outcome is p, which can be quantified by the labor cost of the operation and maintenance personnel to confirm the outcome. The cost of predicting a dangerous outcome as a safe outcome is q, which can be quantified by the economic losses caused by the damage to the transformer and the suspension of production and business due to the power outage.

[0109] Balancing the cost of predicting a safe situation as a dangerous situation (p) with the cost of predicting a dangerous situation as a safe situation (q) can keep the alarm threshold at an appropriate value, avoiding both excessive and false warnings.

[0110] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0111] The above embodiments merely illustrate several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. An online monitoring device for dissolved gas in transformer oil based on remote data analysis, characterized in that: include: Transformer oil cylinder, oil pump, oil sample quantitative collection module, oil-gas separation module, gas collection module, first vacuum pump, gas optical identification module, buffer oil return module, first valve, second valve, third valve, fourth valve, fifth valve, communication module and controller, the gas optical identification module includes an optical sensor; The oil outlet of the transformer oil cylinder is connected to the oil inlet of the oil sample quantitative collection module through the oil pump, the oil outlet of the oil sample quantitative collection module is connected to the oil inlet of the oil and gas separation module through the first valve, the oil outlet of the oil and gas separation module is connected to the oil inlet of the buffer oil return module, and the oil outlet of the buffer oil return module is connected to the oil inlet of the oil cylinder; the air outlet of the oil and gas separation module is connected to the air inlet of the gas collection module through the second valve, the air outlet of the gas collection module is connected to the air inlet of the gas optical recognition module through the third valve, and the air outlet of the gas optical recognition module is connected to the outside world through the fourth valve; the first vacuum pump is connected to the air inlet of the gas collection module through the fifth valve; the communication module establishes a communication connection with the server; The transformer oil cylinder, the oil pump, the first valve, the oil sample quantitative collection module, the oil-gas separation module, and the buffer oil return module constitute an oil circuit; the gas collection module and the gas optical recognition module constitute a gas circuit; Wherein, under the control of the controller, when the first valve is opened, the second valve is closed, the third valve is opened, the fourth valve is opened, and the fifth valve is opened, the oil pump is opened to clean the oil circuit, and the first vacuum pump is opened to clean the gas circuit; under the control of the controller, when the first valve is opened, the second valve is opened, the third valve is opened, the fourth valve is closed, and the fifth valve is closed, the optical sensor is used to detect dissolved gas in the transformer oil, and the detection result is sent to the server through the communication module, and the server performs remote data analysis to monitor the status of the transformer; The online monitoring device further includes a sixth valve, and the gas collection module includes a gas collecting tank and a variable-diameter piston; The gas collecting tank is provided with a variable diameter piston; The variable-diameter piston forms a gas collecting chamber at the top of the gas collecting tank, and the gas collecting chamber is provided with an air inlet and an air outlet. The air outlet of the oil-gas separation module is connected to the air inlet of the gas collecting tank through the second valve, and the air outlet of the gas collecting tank is connected to the air inlet of the gas optical identification module through the third valve. The variable-diameter piston forms a vacuum chamber at the bottom of the gas collecting tank, and the first vacuum pump is connected to the vacuum chamber through the sixth valve; Among them, the sixth valve is opened, and the first vacuum pump is started to evacuate the vacuum chamber; the fifth valve is closed, and the second valve is opened, and the variable piston moves to the bottom to extract the gas in the oil-gas separation module into the gas collecting chamber; the second valve is closed, and the third valve is opened, and the variable piston moves to the top to send the gas in the gas collecting tank into the gas optical recognition module.

2. The online monitoring device for dissolved gas in transformer oil according to claim 1, characterized in that: The oil sample quantitative collection module includes a quantitative collector and a liquid level sensor; The oil outlet of the transformer oil cylinder is connected to the oil inlet of the quantitative collector through the oil pump, and the oil outlet of the quantitative collector is connected to the oil-gas separation module through the first valve; Among them, the first valve is closed, and oil is supplied to the quantitative collector through the oil pump. When the liquid level sensor monitors that the liquid level in the quantitative collector reaches a preset liquid level, the first valve is opened to allow the transformer oil in the quantitative collector to flow into the oil-gas separation module.

3. The online monitoring device for dissolved gas in transformer oil according to claim 1, characterized in that: The oil-gas separation module includes a degassing tank and a first magnetic stirrer; The oil outlet of the oil sample quantitative collection module is connected to the oil inlet of the degassing tank through the first valve, the oil outlet of the degassing tank is connected to the oil inlet of the buffer oil return module, and the gas outlet of the degassing tank is connected to the gas inlet of the gas collection module through the second valve; The first magnetic stirrer is arranged at the bottom of the degassing tank, and is used to stir the transformer oil in the degassing tank; The first magnetic stirrer includes a stator and a stirrer. The stator includes a base and four electromagnetic components. The four electromagnetic components are arranged in a 2×2 manner on the base. The electromagnetic components include a coil, an iron core and a magnetic pole. The coil, the iron core and the magnetic pole constitute a magnetic circuit. The stirrer is arranged on the four magnetic poles. The stirrer is a permanent magnet. The stirrer is controlled to rotate smoothly by the four electromagnetic components.

4. The online monitoring device for dissolved gas in transformer oil according to claim 1, characterized in that: The buffer oil return module includes: a buffer oil return tank, a second vacuum pump, a second magnetic stirrer, a funnel, a buffer net and a seventh valve; The buffer oil return tank is provided with a degassing chamber and an oil return chamber inside, and the degassing chamber is connected to the oil return chamber through the seventh valve; The second vacuum pump is connected to the degassing chamber, and the second magnetic stirrer is arranged at the bottom of the degassing chamber; The funnel is arranged in the oil return chamber, the oil inlet of the oil return chamber is opposite to the inclined wall of the funnel, the buffer net is arranged at the oil outlet of the funnel, and the buffer net is provided with a plurality of oil return holes.

5. The online monitoring device for dissolved gas in transformer oil according to claim 1, characterized in that: The gas optical identification module includes: a gas holding tank and an optical sensor; The gas outlet of the gas collection module is connected to the gas inlet of the gas holding tank through the third valve, and the gas outlet of the gas holding tank is connected to the outside world through the fourth valve; The optical sensor is arranged in the gas holding tank, and includes a structured light emitter and a structured light receiver arranged on two sides of the gas holding tank. The optical sensor is used to determine the composition and concentration of the gas to be measured based on the received light intensity.

6. The online monitoring device for dissolved gas in transformer oil according to claim 5, characterized in that: The controller is used to: Detecting the composition and concentration of the gas to be tested in the gas holding tank by a gas detection model; When the concentration of the target gas contained in the gas to be tested is greater than the alarm threshold, determining that the transformer is in a dangerous state and reporting the detection result to the server; The server notifies the maintenance personnel to go to the destination to confirm and maintain the status of the transformer; The actual result of the maintenance personnel confirming the status of the transformer after arriving at the destination is compared with the detection result of the gas detection model, and the alarm threshold is corrected.

7. The online monitoring device for dissolved gas in transformer oil according to claim 6, characterized in that: The controller is specifically used for: Get the data matrix of multiple standard samples X , a plurality of said standard samples include m Types of gases, each type of gas includes n The standard sample, wherein the gas composition of the standard sample is known and the concentration is 100%; Summarize to get the total sample matrix P , , i =1,2,…, m , j =1,2,…, n ; According to the total sample matrix P , calculate the single-class sample P i The mean matrix of column data η i and the total sample matrix P The mean matrix of λ m : in, Indicates the i Class j The characteristic values ​​of the standard samples; Calculate the total sample matrix P The intra-class scatter matrix Q 1 and the inter-class scatter matrix Q 2: ； Constructing the objective function , and add constraints to calculate the intra-class scatter matrix Q 1 is the smallest and the inter-class scatter matrix Q 2 is the largest, the total sample matrix P The eigenvalue of λ , in order to distinguish the gas components: The constraints are: Therefore, the total sample matrix P The eigenvalue of λ for: Calculate the total sample feature recognition matrix P ': The total sample feature recognition matrix P 'Take the mean of the characteristic values ​​of each type of sample and calculate multiple gas classification matrices Y i : Based on the gas classification matrix Y i , constructing the gas detection model; Extract the characteristic matrix of the gas to be tested in the gas holding tank Z ; Calculate each feature classification matrix Y i and the characteristic matrix of the gas to be measured Z When the Euclidean distance is less than the preset distance, it is determined that the gas to be measured contains the corresponding gas component.

8. The online monitoring device for dissolved gas in transformer oil according to claim 6, characterized in that: The controller is specifically used for: The gas detection model is constructed according to Beer's law: in, I ( λ ) represents the light intensity received by the optical sensor, I 0( λ ) represents the initial light intensity, σ ( λ ) represents the absorption cross section of the substance being measured, λ represents the wavelength, c Indicates the concentration of the substance being measured, L Indicates the length of the gas holding tank; By transforming the above formula, we can get: Assume that the gas to be measured includes K Gas components, obtained in absorption spectrum K test points, then there are in, k =1,2,…, K ; pass K Solve the equation K An unknown c k , to detect the concentration of each gas component in the gas to be tested.

9. The online monitoring device for dissolved gas in transformer oil according to claim 6, characterized in that: The controller is specifically used for: Comparing the detection result with the actual result, and evaluating the detection result; wherein the evaluation result includes: evaluating a safe detection result as a dangerous result, evaluating a dangerous detection result as a safe result, evaluating a safe detection result as a safe result, and evaluating a dangerous detection result as a dangerous result; Let the number of times the security is detected as a security result be TX , the number of times the safety detection is regarded as dangerous is FY The number of times the danger is detected as a dangerous result is TY The number of times the danger is detected as a safety result is FX , then the proportion of safety samples that are predicted incorrectly is FY rate and the proportion of dangerous samples predicted correctly FX rate for: The cost of detecting safety as dangerous is p The cost of detecting danger as a safe result is q , the alarm threshold is modified so that: 。

Images