A method and device for rapidly diagnosing gas content of insulating oil of an oil-filled equipment in a substation

Abstract

The application discloses a kind of transformer substation oil-filled equipment insulating oil gas content rapid diagnosis method and device, it is related to electric power facility safety monitoring technical field, the method includes four steps of online sampling and pretreatment, multimodal synchronous perception, three-dimensional dynamic feature matrix fusion diagnosis, oil sample purification and closed loop return;Subsequently, three-dimensional feature matrix is constructed, weight coefficient is adjusted dynamically in combination with equipment working condition, and hierarchical diagnosis is quickly completed;Finally, after temperature adjustment, insulating oil returns to the body of equipment to form closed loop;The device is closed loop online monitoring system, including online sampling pretreatment unit, multimodal synchronous perception unit, data diagnosis unit and oil sample purification closed loop return unit;Real-time continuous monitoring of insulating oil gas content is realized, diagnosis accuracy is high, without manual intervention, can effectively identify early latent fault such as medium-low temperature overheating, partial discharge, and provide reliable technical support for safe and stable operation of oil-filled equipment.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition monitoring and fault diagnosis technology, specifically a method and device for rapid diagnosis of gas content in insulating oil of oil-filled equipment in substations, used for real-time, continuous, online rapid diagnosis of gas content in insulating oil of oil-filled equipment in substations. Background Technology

[0002] The gas content (i.e., the total amount of dissolved gases and the concentration of key components) in the insulating oil of oil-filled electrical equipment in substations (such as transformers, reactors, bushings, etc.) is a key indicator reflecting the internal insulation condition and latent faults of the equipment. The gases in the oil originate from the thermal decomposition or corona discharge of the insulating materials. Changes in their content and composition can provide early warnings of defects such as partial discharge and low-temperature overheating within the equipment, which is crucial for ensuring the safe operation of the power grid.

[0003] Currently, the monitoring of gas content in insulating oil mainly relies on traditional offline gas chromatography analysis and online monitoring devices based on a single principle. These methods have significant limitations in terms of real-time performance, diagnostic accuracy, and intelligence level, specifically:

[0004] 1. Offline chromatographic analysis suffers from poor timeliness and significant human interference: This method requires manual sampling periodically from the equipment and delivery to the laboratory for analysis, a process that typically takes more than 4 hours and cannot achieve real-time status monitoring. Errors are easily introduced during sampling, transportation, and degassing, leading to gas loss or contamination. Furthermore, it cannot capture the dynamic gas evolution characteristics during the development of a fault, resulting in delayed warnings and the risk of misjudgment.

[0005] 2. Existing online monitoring devices suffer from limited sensing dimensions and insufficient information fusion: Currently deployed online monitoring devices mostly employ single-sensor technology, such as monitoring only the total amount of hydrogen (H2) or a few characteristic gases. This approach has significant drawbacks: (a) It cannot simultaneously acquire multi-dimensional physical signals, providing only static data on gas components, lacking synchronous sensing of related characteristics such as acoustics and dielectrics, resulting in severe information silos; (b) It has weak anti-interference capabilities, as single sensor signals are easily affected by oil temperature fluctuations, flow rate changes, oil deterioration, and the on-site electromagnetic environment, leading to drift or distortion in measurement results; (c) It has low sensitivity for early fault identification, making it difficult to reliably detect minute changes in single gas concentrations in latent faults such as slow-producing low-temperature overheating, easily resulting in missed reports.

[0006] 3. Existing diagnostic models have low intelligence and poor adaptability: Whether it is the "three ratio method" or "David's triangle method" based on offline chromatographic data, or online alarms that simply set fixed thresholds, their diagnostic models all have obvious defects: (a) The models are static and rigid, and are mostly based on historical experience to set fixed thresholds or ratios. They cannot adapt to changes in different equipment models, years of operation, loads and ambient temperatures, resulting in a high false alarm rate; (b) They lack the ability to fuse multi-source information and fail to effectively utilize the potential coupling relationship and spatiotemporal correlation between multi-sensor (sound, light, and electricity) data, resulting in a one-sided diagnostic basis; (c) They cannot achieve dynamic weight adjustment. Under complex operating conditions, the reliability of different sensor signals will change. Existing models lack a dynamic evaluation and weighted fusion mechanism for this, resulting in unreliable diagnostic conclusions under critical operating conditions.

[0007] In summary, existing technologies suffer from systemic shortcomings in three areas: real-time continuous monitoring, multimodal sensing fusion, and intelligent adaptive diagnosis. Therefore, there is an urgent need for a method and device for diagnosing the gas content of insulating oil that can integrate simultaneous multi-physics field measurements, possess online self-calibration and anti-interference capabilities, and achieve rapid, accurate, and early fault warnings through dynamic information fusion. This would address the core pain points of traditional methods, such as long cycles, poor accuracy, and low levels of intelligence, and meet the urgent need of modern intelligent substations for real-time and accurate perception of equipment status. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to address the shortcomings of the prior art by disclosing a rapid diagnostic method and device for the gas content of insulating oil in substation oil-filled equipment, thereby solving the technical problems of long detection time, large human influence, low accuracy and single diagnostic means in traditional insulating oil gas content detection methods.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method and apparatus for rapid diagnosis of gas content in insulating oil of substation oil-filled equipment, comprising the following steps:

[0010] S1, Online Sampling and Pretreatment: Insulating oil is extracted online from the sampling valve at the bottom of the oil filling equipment body, and its flow rate is controlled within the range of 0.5-2L / min by the first oil pump. The oil is then passed through the first and second filter modules connected in series to remove particulate impurities.

[0011] S2, Multimodal synchronous sensing: The pre-treated insulating oil is diverted through the first four-way valve and simultaneously flows through the parallel ultrasonic measurement module, laser spectral measurement module and broadband dielectric measurement module to acquire the sound velocity change signal, characteristic gas absorption spectrum signal and broadband dielectric spectrum signal characterizing the gas state in the oil at the same time and from the same source.

[0012] S3, Three-dimensional dynamic feature matrix fusion diagnosis: The sound speed change signal, the characteristic gas absorption spectrum signal, and the broadband dielectric spectrum signal are fused to construct and calculate the three-dimensional dynamic feature matrix G value: ;

[0013] in, The normalized rate of change of sound speed, Δv is the difference between the current measured sound speed v and the reference sound speed v0;

[0014] The normalized characteristic gas concentration change rate is A, where A is the characteristic gas concentration value measured by the laser spectroscopy measurement module, and A0 is the reference concentration threshold of the characteristic gas.

[0015] ( ) represents the normalized rate of change of dielectric properties, and ε represents the dielectric constant. The dielectric loss factor of insulating oil;

[0016] α is ( The weighting coefficients and β are () The weighting coefficients and γ are () The dynamic weighting coefficients of )

[0017] Based on the G value, the gas content of insulating oil is used for condition diagnosis and graded early warning.

[0018] S4, Oil Sample Purification and Closed-Loop Return: The tested insulating oil is collected through the second four-way valve and sent to the vacuum oil filtration module for degassing and purification. Then, the temperature is adjusted by the temperature compensation module to match the oil temperature inside the oil-filling equipment. Finally, it is returned to the equipment body through the upper sampling valve to form a closed-loop oil circuit.

[0019] Furthermore,

[0020] In step S2:

[0021] The ultrasonic measurement module uses 1MHz frequency ultrasound to detect changes in sound velocity in response to dissolved gases and microbubbles in the oil.

[0022] The laser spectroscopy measurement module uses a tunable diode laser with a wavelength of 1653nm to detect the concentration of methane gas based on infrared absorption spectroscopy.

[0023] The wideband dielectric measurement module scans within a frequency range of 10Hz-1kHz, analyzing the overall state of the oil and gas dissolution by measuring the dielectric spectrum response.

[0024] Furthermore,

[0025] In step S3, the dynamic weighting coefficients α, β, and γ are adaptively optimized and adjusted based on historical equipment operating data and comparison with offline gas chromatography analysis results.

[0026] The adaptive optimization adjustment process of the dynamic weight coefficients α, β, and γ includes two stages:

[0027] Initialization and calibration phase: In the initial stage of device operation, offline gas chromatography analysis results are collected as the true value of gas content. The calculated value of the three-dimensional feature matrix G is compared with the result. Based on the principle of minimizing deviation, the initial values ​​of α, β, and γ are corrected through iterative algorithm.

[0028] Operational optimization phase: During continuous operation of the device, historical operating data of the equipment is extracted periodically, and combined with gas content monitoring data under operating conditions of 20-80℃ temperature and 0.1-0.5MPa pressure, the weight ratio of α, β, and γ is dynamically adjusted; for gas content sensitive operating conditions, the weight coefficient of the corresponding detection dimension is increased to ensure the matching degree between the G value calculation result and the actual gas content of the insulating oil.

[0029] Furthermore,

[0030] In step S3, the diagnostic and hierarchical early warning rules based on the value of the three-dimensional dynamic feature matrix G are as follows:

[0031] When G < 1%, the gas content of the insulating oil is considered normal;

[0032] When 1%≤G≤3%, the gas content is determined to be too high, an early warning is issued, the detection cycle is shortened, and trend tracking is strengthened.

[0033] When G>3%, the gas content is determined to be seriously excessive, indicating a risk of partial discharge or overheating faults, and the warning information is pushed to the substation intelligent monitoring platform at the same time.

[0034] A rapid diagnostic device for the gas content of insulating oil in substation oil-filled equipment.

[0035] The device constitutes a closed-loop online monitoring system connected to the main body of the oil-filling equipment. It eliminates the need for manual sampling and enables real-time continuous monitoring of the gas content in the insulating oil. It comprises the following components connected sequentially along the oil path:

[0036] The online sampling and pretreatment unit includes a lower sampling valve, a first oil pump, and a first filter module and a second filter module connected in series.

[0037] The multimodal synchronous sensing unit includes a first four-way valve, and an ultrasonic measurement module, a laser spectroscopy measurement module, and a broadband dielectric measurement module connected in parallel downstream thereto, for the purpose of achieving synchronous measurement;

[0038] The data diagnostic unit communicates with each measurement module and is used to execute the three-dimensional dynamic feature matrix fusion diagnostic algorithm.

[0039] The oil sample purification and closed-loop return unit includes a second four-way valve, a vacuum oil filtration module, a temperature compensation module, a second oil pump, and the upper sampling valve, which are used to purify and temperature-adjust the oil sample before returning it to the main body of the equipment.

[0040] Furthermore,

[0041] The ultrasonic measurement module, laser spectroscopy measurement module, and broadband dielectric measurement module in the multimodal synchronous sensing unit are independent of each other, and all have a detection resolution of 0.01% gas content, and support independent calibration and maintenance.

[0042] Furthermore,

[0043] The first oil pump is a variable frequency oil-resistant gear pump with a maximum flow rate of 3L / min, used to precisely control the oil sample flow rate;

[0044] The temperature compensation module uses electric heating and has a temperature control accuracy of ±0.1℃ to ensure that the return oil temperature is consistent with the oil temperature of the oil filling equipment body.

[0045] Furthermore,

[0046] The first filter module has a filtration accuracy of 5 μm, the second filter module has a filtration accuracy of 0.1 μm, and the vacuum oil filtration module has a vacuum degree of 100 Pa.

[0047] Furthermore,

[0048] The device is made entirely of stainless steel and is equipped with a fault self-protection unit that automatically cuts off the oil circuit between the oil filling equipment body and the device in case of an abnormality.

[0049] Furthermore,

[0050] The data diagnostic unit also integrates a communication module for uploading diagnostic results to the substation intelligent monitoring platform.

[0051] The beneficial effects of this invention are as follows:

[0052] The rapid diagnostic method and device for gas content in insulating oil of substation oil-filled equipment described in this invention effectively solves the technical pain points of traditional detection methods, such as long cycle, large human interference, single diagnostic methods, and insufficient early fault identification capability, and has significant practical and economic value.

[0053] Adopting a closed-loop oil circuit design, no manual sampling is required throughout the process, enabling real-time continuous monitoring of the gas content in insulating oil. Compared to the 4-hour diagnostic time of traditional offline chromatographic analysis, this invention can complete the entire diagnostic process within 30 minutes, reducing the time by more than 95%. Furthermore, by continuously monitoring and eliminating abnormal data, the diagnostic accuracy is improved by more than 10%, completely avoiding false alarms and missed alarms.

[0054] Meanwhile, this invention innovatively employs three methods—ultrasound, laser spectroscopy, and broadband dielectric analysis—for simultaneous detection, each achieving a resolution of 0.01% gas content. Combined with a three-dimensional feature matrix and dynamic weight adjustment algorithm, it eliminates interference from temperature fluctuations of 20-80℃ and pressure changes of 0.1-0.5MPa, accurately identifying gas content variation characteristics of faults such as low-temperature overheating, partial discharge, and poor sealing, thus enabling early warning of latent faults.

[0055] Furthermore, the entire device is made of stainless steel, with a pressure rise rate ≤1Pa / min. Its miniaturized and modular design requires no modification to existing equipment. Sampling and return utilize a fully sealed integrated pipeline, with a maximum oil consumption of only 3L / min, having no impact on the equipment's oil level and pressure. The filter module has a clearly defined filter element replacement cycle and controllable costs. Each sensor supports independent maintenance and calibration, significantly reducing maintenance complexity. The device complies with substation explosion-proof and electromagnetic interference prevention standards and is equipped with a fault self-protection unit that automatically cuts off the oil circuit in case of power failure or malfunction, without affecting normal equipment operation. The temperature compensation module has a temperature control accuracy of ±0.1℃, accurately matching the body oil temperature to avoid disrupting the oil dissolution balance. The built-in communication module can interface with an intelligent monitoring platform, employing a "local cache + background archiving" storage architecture to support fault tracing and trend analysis. In addition, the two-stage filtration module removes impurities from the oil step by step. The vacuum oil filtration module achieves a vacuum degree of 100Pa, with a particulate matter removal efficiency ≥99% and moisture reduced to 10ppm, effectively removing bubbles and impurities generated during the detection process, ensuring detection accuracy, extending the device's service life, and providing comprehensive and reliable technical support for the safe and stable operation of substation oil-filling equipment. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a flowchart of the rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment according to the present invention;

[0058] Figure 2 This is a schematic diagram of the structure of the rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to the present invention.

[0059] Figure 3 This is a flowchart of the three-dimensional feature matrix diagnostic logic in the rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment of the present invention. Attached Figure Description

[0060] 1-Oil filling equipment body; 2-Lower sampling valve; 3-Upper sampling valve; 4-First valve; 5-First pipeline; 6-First oil pump; 7-Second valve; 8-Second pipeline; 9-First filter module; 10-Third valve; 11-Third pipeline; 12-Second filter module; 13-Fourth valve; 14-Fourth pipeline; 15-First four-way valve; 16-Ultrasonic measurement module; 17-Fifth valve; 18-Sixth valve; 19-Fifth pipeline; 20-Laser spectroscopy measurement. Modules: 21-Seventh Valve, 22-Eighth Valve, 23-Sixth Pipeline, 24-Wideband Dielectric Measurement Module, 25-Ninth Valve, 26-Tenth Valve, 27-Seventh Pipeline, 28-Second Four-Way Valve, 29-Tenth Valve, 30-Eighth Pipeline, 31-Vacuum Oil Filter Module, 32-Ninth Pipeline, 33-Second Oil Pump, 34-Eleventh Valve, 35-Tenth Pipeline, 36-Temperature Compensation Module, 37-Data Diagnostic Module, 38-Insulating Oil Gas Content Monitoring and Diagnostic Device. Detailed Implementation

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0064] Example 1

[0065] This embodiment combines Figure 1 and Figure 2 The overall implementation scheme of the method and apparatus described in this invention will be explained in detail.

[0066] like Figure 2As shown, the present invention provides a rapid diagnostic device for the gas content of insulating oil in substation oil-filled equipment. This device constitutes a closed-loop online monitoring system connected to the main body 1 of the oil-filled equipment.

[0067] The system is encapsulated in a stainless steel housing for monitoring and diagnosing the gas content of insulating oil. It eliminates the need for manual sampling and enables real-time, continuous, and in-situ monitoring of the gas content of insulating oil.

[0068] Please combine Figure 1 The method flowchart shown illustrates that the device operates according to the following steps:

[0069] The entire system is encapsulated within a stainless steel housing for monitoring and diagnosing the gas content of insulating oil. It operates as follows:

[0070] S1, Online Sampling and Pre-processing: Upon monitoring startup, the first valve 4 is opened. Insulating oil within the oil-filled equipment body 1, under internal pressure and the extraction action of the first oil pump 6, is drawn out from the lower sampling valve 2 and enters the monitoring circuit via the first pipeline 5. The first oil pump 6 is a variable frequency oil-resistant gear pump with a maximum flow rate of 3 L / min. Its speed is steplessly adjusted by a frequency converter, precisely and stably controlling the oil sample flow rate within a preset range of 0.5-2 L / min. This optimized flow rate aims to suppress turbulence, prevent air bubbles from forming in the detection section, and ensure sufficient response time for subsequent sensors. Subsequently, the oil sample passes sequentially through the first filter module 9 (5 μm filtration accuracy) and the second filter module 12 (0.1 μm filtration accuracy) connected in series to remove impurities such as metal particles and cellulose of different sizes from the oil, providing a clean oil sample for precision measurement. The second valve 7, third valve 10, and fourth valve 13 remain open during normal monitoring.

[0071] S2, Multimodal Synchronous Sensing: The pretreated clean oil sample reaches the first four-way valve 15 through the fourth pipeline 14. This valve synchronously and proportionally divides the single oil flow into three streams, which are then delivered to:

[0072] Ultrasonic measurement module 16 (via fifth pipe 19, on which fifth valve 17 and sixth valve 18 are provided).

[0073] Laser spectroscopy measurement module 20 (via the sixth pipe 23, which is equipped with the seventh valve 21 and the eighth valve 22).

[0074] Wideband dielectric measurement module 24 (via the seventh conduit 27, which is equipped with the ninth valve 25 and the tenth valve 26).

[0075] The three sensing modules measure the insulating oil in parallel, simultaneously, and from the same source, and each transmits the real-time signal to the data diagnostic unit 37.

[0076] S3, 3D Dynamic Feature Matrix Fusion Diagnosis: Data Diagnosis Unit 37 receives and processes synchronization signals from the three modules, and executes the core diagnostic algorithm (its internal logic flow is described in [link]). Figure 3 The specific process will be detailed in Example 3.

[0077] S4, Oil Sample Purification and Closed-Loop Return: After measurement, the three oil samples re-converge at the second four-way valve 28 and enter the vacuum oil filtration module 31 via the eighth pipeline 30 (equipped with the tenth valve 29) for degassing and deep purification. This module maintains a vacuum of approximately 100 Pa, effectively removing trace bubbles and impurities generated during the testing process, with a particulate matter removal efficiency ≥99%. The purified oil sample flows into the temperature compensation module 36 via the ninth pipeline 32. This module uses an electric heating tape wrapped around the outer wall of the return oil pipeline for PID closed-loop temperature control, achieving a temperature control accuracy of ±0.1℃, precisely adjusting the oil sample temperature to match the real-time oil temperature of the oil filling equipment body 1 (temperature difference controlled within ±1℃). Finally, driven by the second oil pump 33, the temperature-matched clean insulating oil returns in full to the oil filling equipment body 1 via the tenth pipeline 35 (equipped with the eleventh valve 34) and the upper sampling valve 3, forming a complete, safe, and non-destructive closed-loop circulation of insulating oil. The entire monitoring process can be completed within 30 minutes, enabling rapid diagnosis.

[0078] Example 2

[0079] This embodiment is a further refinement of the multimodal synchronous sensing unit described in Embodiment 1, such as... Figure 2 As shown.

[0080] The core of the ultrasonic measurement module 16 is an ultrasonic transducer with a transmission frequency of 1 MHz and a high-precision timing circuit. Its working principle is based on the fact that dissolved gas in insulating oil significantly reduces the ultrasonic propagation speed, while suspended microbubbles cause sound wave scattering and attenuation; by accurately measuring the change in sound velocity Δv, a high-sensitivity response to gas content can be achieved, with a detection resolution of 0.01%.

[0081] The laser spectroscopy measurement module 20 employs tunable diode laser absorption spectroscopy technology. Its laser emission wavelength is rapidly scanned near 1653 nm (the characteristic absorption line of methane). By detecting the absorption intensity after the laser passes through the oil sample, the concentration of dissolved methane in the oil is directly calculated according to the Beer-Lambert law, exhibiting excellent selectivity and anti-interference capabilities. The broadband dielectric measurement module 24 performs alternating electric field scanning on the oil sample within a frequency range of 10 Hz to 1 kHz, analyzing the dielectric constant ε and dielectric loss factor ε... 0The frequency-varying spectrum comprehensively assesses the degradation, moisture absorption, and gas content of the insulating oil; the low-frequency range is sensitive to conductive impurities and moisture, while the mid-to-high-frequency range is more responsive to oil molecule polarization and dissolved gases. The three sensing modules are independent of each other and all support periodic calibration and independent maintenance based on the external standard method. The system integrates power-on self-test and periodic self-diagnosis functions to ensure long-term operational stability.

[0082] Example 3

[0083] This embodiment combines Figure 3 The diagnostic algorithm and early warning logic executed by the data diagnostic unit 37 are explained in detail.

[0084] The software algorithm execution of data diagnostic unit 37 is as follows: Figure 3 The diagnostic logic flow shown is based on the calculation and application of the three-dimensional dynamic feature matrix G value:

[0085] Data fusion and G-value calculation: The unit synchronously receives the sound velocity change Δv, methane concentration A, and dielectric loss tanδ. After temperature and pressure compensation, the normalized eigenvalue is calculated. ), ( ), ( Subsequently, the adaptively optimized weight coefficients α, β, γ are called, according to the formula... Calculate the comprehensive state index G value;

[0086] in, The normalized rate of change of sound speed is Δv, which is the difference between the current measured sound speed v and the reference sound speed v0.

[0087] The normalized characteristic gas concentration change rate is A, which is the characteristic gas concentration value measured by the laser spectroscopy measurement module (20), and A0 is the reference concentration threshold of the characteristic gas.

[0088] The normalized rate of change of dielectric properties is given by ε, where ε is the dielectric constant. The dielectric loss factor of insulating oil;

[0089] α is ( The weighting coefficients and β are () The weighting coefficients and γ are () The dynamic weighting coefficients of )

[0090] Dynamic weight optimization: The weight coefficients α, β, and γ are not fixed values, and the optimization process consists of two stages:

[0091] Initial calibration: In the initial stage of device operation, the results of offline gas chromatography analysis are used as a benchmark, and the initial values ​​of α, β, and γ are corrected through iterative algorithms to minimize the deviation between the calculated G value and the actual gas content.

[0092] Adaptive operation: During continuous operation, the weights are dynamically adjusted by combining historical equipment data with real-time operating conditions (temperature 20-80℃, pressure 0.1-0.5MPa). For example, when the equipment operating temperature is high and the dielectric signal is easily interfered with, γ is automatically reduced; when the quality of the laser spectral signal deteriorates, β is reduced, relying more on data from other dimensions.

[0093] Grading diagnosis and early warning: Rapid grading judgment based on G value, the process is as follows: Figure 3 The decision node is shown below:

[0094] Judgment Node 1 (Normal): If G<1%, the gas content of the insulating oil is considered normal and the equipment is in good operating condition.

[0095] Judgment Node 2 (Warning): If 1%≤G≤3%, the gas content is considered too high. The system issues a "Caution" warning, automatically shortens the monitoring cycle (e.g., adjusts to intensive sampling every 10 minutes), strengthens trend tracking, and uploads the information to the monitoring platform.

[0096] Judgment Node 3 (Fault Alarm): If G > 3%, the gas content is deemed severely excessive, the insulating oil performance is significantly degraded, and the risk of latent faults such as partial discharge or overheating inside the equipment is extremely high. The system immediately generates the highest level alarm and pushes alarm information containing detailed data to the substation intelligent monitoring platform in real time, prompting an emergency inspection. The entire diagnostic process can be completed within 3 minutes.

[0097] To illustrate the diagnostic process more specifically, a typical application scenario will be used as an example.

[0098] Suppose that during continuous monitoring of the main transformer in a substation, the system acquires the following synchronous data and operating conditions at a specific moment: the equipment oil temperature is 62℃, and the internal pressure is 0.28MPa. The real-time sound velocity v measured by the ultrasonic measurement module is 1472m / s. Based on the reference sound velocity v0 under this operating condition, set to 1488 m / s, calculate the normalized rate of change of sound velocity (…). The value is -0.0108;

[0099] The laser spectroscopy measurement module 20 detected a methane gas concentration of 115 μL / L. Compared to the set attention threshold A0 = 100 μL / L, its normalized concentration change rate (…) The value is 1.15;

[0100] The wideband dielectric measurement module 24 measures a dielectric loss tangent (tanδ) of 0.016 at power frequency (50Hz), relative to the reference standard under the current operating conditions. (0.011), calculated to be ( The value is 1.45.

[0101] Based on the current operating condition of "temperature above 60℃", the data diagnostic unit 37 automatically executes a weight adaptive strategy, appropriately reducing the weight of the temperature-sensitive dielectric characteristics and increasing the weight of the sound velocity characteristics, dynamically generating the weight coefficients for this round of diagnosis as follows: α=0.45, β=0.35, γ=0.20.

[0102] Substitute the above data into the formula for calculating the three-dimensional dynamic feature matrix:

[0103] G = 0.45 × (-0.0108) + 0.35 × 1.15 + 0.20 × 1.45 ≈ 0.821, meaning the value of G is 82.1%. This value far exceeds the 3% fault alarm threshold.

[0104] Accordingly, the system immediately triggers the highest-level fault alarm, determining that the gas content in the transformer's insulating oil is severely abnormal, strongly suggesting the possible presence of partial discharge or overheating latent faults within the equipment. The data diagnostic unit simultaneously pushes complete alarm information, including the G value, component data, dynamic weights, and diagnostic conclusions, to the substation's intelligent monitoring platform via the communication module. This example fully demonstrates the closed-loop process from multimodal data synchronous acquisition, dynamic weight adaptive adjustment, three-dimensional feature matrix fusion calculation to intelligent hierarchical early warning, showcasing the effectiveness and superiority of the method of this invention in achieving rapid, accurate, and early fault diagnosis.

[0105] Example 4

[0106] This embodiment describes the engineering design and safety characteristics of the device. The reliability of the device is firstly guaranteed by the fault self-protection mechanism. The built-in protection unit monitors the system power supply, oil pressure (safe range 0.1-0.5 MPa) and key sensor status in real time. Once a power failure, pressure over-limit or serious fault is detected, the main oil valve can be automatically cut off within milliseconds to ensure the absolute safety of the oil filling equipment body (1). Secondly, a modular design is adopted to facilitate maintenance. For example, the filter elements of the two-stage filtration module have a clear replacement cycle (approximately 12 months for the first stage and approximately 36 months for the second stage). Each sensor module supports online isolation and rapid replacement, which significantly reduces the complexity and cost of operation and maintenance. In terms of data management, the device integrates an industrial Ethernet communication module, supports standard protocols such as IEC 61850, and can realize the real-time uploading of diagnostic results. It adopts a hierarchical storage architecture of "local cache (capacity not less than 30 days of historical data) + centralized archiving in the background", which effectively supports fault tracing and long-term trend analysis. Finally, the device adopts a stainless steel structure, which meets the strict explosion-proof, electromagnetic interference-proof and protection level requirements of substations and has good environmental adaptability.

[0107] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0108] Furthermore, although the operations of the method of the present invention are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

Claims

1. A rapid diagnostic method for the gas content of insulating oil in substation oil-filled equipment. Its features are, include The following steps: S1, Online sampling and pretreatment: Insulating oil is extracted online from the sampling valve (2) at the bottom of the oil filling equipment body (1), and its flow rate is controlled within the range of 0.5-2L / min by the first oil pump (6), and the oil is then passed through the first filter module (9) and the second filter module (12) connected in series to remove particulate impurities in the oil. S2, Multimodal synchronous sensing: The pre-treated insulating oil is diverted through the first four-way valve (15) and flows synchronously through the parallel ultrasonic measurement module (16), laser spectral measurement module (20) and broadband dielectric measurement module (24) to simultaneously and from the same source acquire the sound velocity change signal, characteristic gas absorption spectrum signal and broadband dielectric spectrum signal characterizing the gas state in the oil; S3, Three-dimensional dynamic feature matrix fusion diagnosis: The sound speed change signal, the characteristic gas absorption spectrum signal, and the broadband dielectric spectrum signal are fused to construct and calculate the three-dimensional dynamic feature matrix G value. ; in, The normalized rate of change of sound speed is Δv, which is the difference between the current measured sound speed v and the reference sound speed v0. The normalized characteristic gas concentration change rate is A, which is the characteristic gas concentration value measured by the laser spectroscopy measurement module (20), and A0 is the reference concentration threshold of the characteristic gas. The normalized rate of change of dielectric properties is given by ε, where ε is the dielectric constant. The dielectric loss factor of insulating oil; α is ( The weighting coefficients and β are () The weighting coefficients and γ are () The dynamic weighting coefficients of ) Based on the G value, the gas content of insulating oil is used for condition diagnosis and graded early warning. S4, Oil sample purification and closed-loop return: The insulating oil that has completed the test is collected through the second four-way valve (28) and sent to the vacuum oil filter module (31) for degassing and purification. Then, the temperature is adjusted by the temperature compensation module (36) to be consistent with the oil temperature in the main body (1) of the oil filling equipment. Finally, it is returned to the main body of the equipment through the upper sampling valve (3) to form a closed-loop oil circuit.

2. The rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment according to claim 1, Its features are, In step S2: The ultrasonic measurement module (16) uses ultrasonic waves at a frequency of 1MHz to detect changes in sound velocity in response to dissolved gases and microbubbles in the oil. The laser spectroscopy measurement module (20) uses a tunable diode laser with a wavelength of 1653nm to detect the concentration of methane gas based on infrared absorption spectroscopy. The wideband dielectric measurement module (24) scans within a frequency range of 10Hz-1kHz, and analyzes the overall state of the oil and the gas dissolution by dielectric spectrum response.

3. The rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment according to claim 1. Its features are, In step S3, the dynamic weighting coefficients α, β, and γ are adaptively optimized and adjusted based on historical equipment operating data and comparison with offline gas chromatography analysis results. The adaptive optimization adjustment process of the dynamic weight coefficients α, β, and γ includes two stages: Initialization and calibration phase: In the initial stage of device operation, offline gas chromatography analysis results are collected as the true value of gas content. The calculated value of the three-dimensional feature matrix G is compared with the result. Based on the principle of minimizing deviation, the initial values ​​of α, β, and γ are corrected through iterative algorithm. Operational optimization phase: During continuous operation of the device, historical operating data of the equipment is extracted periodically, and combined with gas content monitoring data under operating conditions of 20-80℃ temperature and 0.1-0.5MPa pressure, the weight ratio of α, β, and γ is dynamically adjusted; for gas content sensitive operating conditions, the weight coefficient of the corresponding detection dimension is increased to ensure the matching degree between the G value calculation result and the actual gas content of the insulating oil.

4. The rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment according to claim 1. Its features are, In step S3, the diagnostic and hierarchical early warning rules based on the value of the three-dimensional dynamic feature matrix G are as follows: When G < 1%, the gas content of the insulating oil is considered normal; When 1%≤G≤3%, the gas content is determined to be too high, an early warning is issued, the detection cycle is shortened, and trend tracking is strengthened. When G>3%, the gas content is determined to be seriously excessive, indicating a risk of partial discharge or overheating faults, and the warning information is pushed to the substation intelligent monitoring platform at the same time.

5. A rapid diagnostic device for the gas content of insulating oil in substation oil-filled equipment. The rapid diagnostic method for gas content in insulating oil of substation oil-filled equipment as described in any one of claims 1 to 4 is adopted. Its features are, The device constitutes a closed-loop online monitoring system connected to the oil-filled equipment body (1), eliminating the need for manual sampling and enabling real-time continuous monitoring of the gas content in the insulating oil. It includes components connected sequentially along the oil path: The online sampling and pretreatment unit includes a lower sampling valve (2), a first oil pump (6), and a first filter module (9) and a second filter module (12) connected in series. The multimodal synchronous sensing unit includes a first four-way valve (15), and an ultrasonic measurement module (16), a laser spectroscopy measurement module (20), and a broadband dielectric measurement module (24) connected in parallel downstream thereto, for the purpose of achieving synchronous measurement; The data diagnostic unit (37) is connected in communication with each measurement module and is used to execute the three-dimensional dynamic feature matrix fusion diagnostic algorithm. The oil sample purification and closed-loop return unit includes a second four-way valve (28), a vacuum oil filtration module (31), a temperature compensation module (36), a second oil pump (33), and the upper sampling valve (3), which is used to purify and adjust the temperature of the oil sample before returning it to the main body of the equipment.

6. The rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to claim 5, Its features are, The ultrasonic measurement module (16), laser spectroscopy measurement module (20), and broadband dielectric measurement module (24) in the multimodal synchronous sensing unit are independent of each other, and all have a detection resolution of 0.01% gas content, and support independent calibration and maintenance.

7. The rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to claim 5. Its features are, The first oil pump (6) is a variable frequency oil-resistant gear pump with a maximum flow rate of 3L / min, used to precisely control the flow rate of the oil sample; The temperature compensation module (36) uses electric heating and has a temperature control accuracy of ±0.1℃ to ensure that the return oil temperature is consistent with the oil temperature of the oil filling equipment body (1).

8. The rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to claim 5. Its features are, The first filter module (9) has a filtration accuracy of 5 μm, the second filter module (12) has a filtration accuracy of 0.1 μm, and the vacuum oil filter module (31) has a vacuum degree of 100 Pa.

9. The rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to claim 5. Its features are, The device is made of stainless steel and is equipped with a fault self-protection unit that automatically cuts off the oil circuit between the oil filling equipment body (1) and the device in case of an abnormality.

10. The rapid diagnostic device for gas content in insulating oil of substation oil-filled equipment according to claim 5. Its features are, The data diagnostic unit (37) also integrates a communication module for uploading diagnostic results to the substation intelligent monitoring platform.

Images