Fluid leakage adaptive detection method for electrical control devices

Abstract

This invention provides an adaptive detection method for fluid leakage in electrical control equipment, comprising: reading the cabinet structure information, cooling pipe joint location, terminal proximity area location, internal air duct location, sensor installation location, and internal temperature, humidity, equipment load, and surface dielectric response data of the electrical control equipment; establishing a basic mapping model of the cabinet; generating a set of easily leakable monitoring areas and a monitoring parameter table for each area; applying low-power pulse heating to the easily leakable monitoring areas; acquiring temperature recovery curves, humidity recovery curves, and surface dielectric recovery curves; subtracting temperature rise changes caused by load fluctuations; generating condensation reference features; acquiring operating thermo-humidity-electrical data sequences during operation; calculating the humidity rise rate, temperature recovery time, surface dielectric offset, and load change; generating a fingerprint of impending thermo-humidity-electrical coupling leakage; and matching this fingerprint with the condensation reference features under the same area identifier to generate a condensation determination record.

Description

Technical Field

[0001] This invention relates to the field of electrical equipment safety monitoring technology, and in particular to an adaptive detection method for fluid leakage in electrical control equipment. Background Technology

[0002] Electrical control equipment is widely used in industrial production lines, liquid-cooled power cabinets, and frequency converter control cabinets. It typically houses components such as controllers, contactors, and cooling pipes. As equipment power density increases, the use of liquid-cooled pipes and cooling connectors inside the cabinets also increases. If coolant, liquefied water vapor, or external fluids enter the area near the terminals, it can easily cause a decrease in insulation performance, terminal creepage, localized corrosion, or even short circuits and shutdowns.

[0003] Existing leak detection methods mostly employ leak ropes, humidity sensors, bottom liquid accumulation detection pads, or single-point liquid level detectors, typically triggering alarms based on fixed thresholds. While these methods are simple in structure, they are easily affected by fan heat dissipation, load heating, air conditioning dehumidification, and ambient temperature differences within electrical control equipment, misinterpreting short-term condensation or localized humidity fluctuations as actual leaks. Furthermore, early micro-leaks in liquid-cooled pipe joints often do not yet form obvious liquid accumulation, only manifesting as localized temperature and humidity changes, surface dielectric response shifts, or abnormal recovery processes, making them difficult to identify in a timely manner using traditional detection methods.

[0004] Therefore, this invention proposes an adaptive detection method for fluid leakage in electrical control equipment. The information disclosed above in the background section is only for enhancing understanding of the background of this disclosure and may therefore contain prior art information that is not common knowledge to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing an adaptive detection method for fluid leakage in electrical control equipment, thereby solving the technical problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An adaptive detection method for fluid leakage in electrical control equipment includes the following steps:

[0008] S1. Read the cabinet structure information of the electrical control equipment, as well as the detection and response data inside the cabinet, establish the basic mapping model inside the cabinet, and generate a set of easily leaking monitoring areas and a table of monitoring parameters for the areas.

[0009] S2. According to the set of easily leakable monitoring areas and the monitoring parameter table, apply low-power pulse heating to the easily leakable monitoring areas, collect feedback data, and generate condensation reference characteristics;

[0010] S3. Collect the operating thermo-humidity and electrical data sequence, calculate the change in the corresponding data, and generate a fingerprint of the early signs of thermo-humidity and electrical coupling leakage by combining the positional relationship.

[0011] S4. Match the fingerprint of the early signs of thermal-wet-electric coupling leakage with the condensation reference features under the same area identification to obtain the condensation matching results and leakage deviation results, and generate condensation judgment records or fluid leakage judgment records.

[0012] S5. Update the adaptive alarm threshold in the area monitoring parameter table based on the condensation judgment record, fluid leakage judgment record and historical judgment results. Determine the handling level based on the leakage deviation result and electrical risk location. Output the detection result and feed the updated area monitoring parameter table back to the operation acquisition process.

[0013] S1 specifically includes: reading the cabinet structure information, cooling pipe joint location, terminal proximity area location, cabinet air duct location, and sensor installation location of the electrical control equipment; establishing a basic mapping model of the cabinet interior using a preset corner point on the cabinet base plate as the coordinate origin; based on the basic mapping model of the cabinet interior, collecting and binding cabinet interior temperature data, cabinet interior humidity data, equipment load data, and surface dielectric response data under normal operating conditions to form a normal reference response; associating the basic mapping model of the cabinet interior with the collected data areas, filtering easily leakable monitoring areas according to the proximity relationship of cooling pipe joints, proximity relationship of terminals, downstream relationship of cabinet air ducts, and liquid accumulation path relationship in low-lying areas of the cabinet, and generating a set of easily leakable monitoring areas and a monitoring parameter table for the areas.

[0014] S2 specifically includes: applying low-power pulse heating to the leak-prone monitoring areas according to the set of leak-prone monitoring areas and the area monitoring parameter table, recording the area identification, heating power, duration, start and end times, temperature before and after heating, and load data to form a pulse heating execution record; after heating, collecting temperature recovery curves, humidity recovery curves, and surface dielectric recovery curves, and combining the pulse heating execution record and equipment load data to deduct the temperature rise caused by load fluctuations to obtain a set of area recovery curves; and extracting the temperature recovery time range, humidity decrease slope range, surface dielectric response regression time range, and load correspondence from the set of area recovery curves to generate condensation reference features bound to the area identification.

[0015] S3 specifically includes: according to the set of easily leakable monitoring areas and the area monitoring parameter table, continuously collecting real-time temperature data, real-time humidity data, real-time equipment load data and real-time surface dielectric response data of each easily leakable monitoring area during operation, forming an operational thermo-humidity-electrical data sequence with area identifiers and timestamps; based on the operational thermo-humidity-electrical data sequence, calculating the humidity rise rate, temperature recovery time, surface dielectric offset and load change; combining the above calculation results and corresponding positional relationships to generate a thermo-humidity-electrical coupling leakage precursor fingerprint including area identifier, abnormal location, humidity change characteristics, temperature recovery characteristics, dielectric offset characteristics, load association characteristics, abnormal duration and position weight.

[0016] S4 specifically includes: matching the fingerprint of the precursor to thermal-humidity-electric coupling leakage with the condensation reference features under the same area identification, and normalizing and comparing the correspondence between temperature recovery, humidity recovery, surface dielectric recovery and load to obtain condensation matching results and leakage deviation results; when the condensation matching results show that the humidity rises and then falls back, the surface dielectric response does not continue to deviate and the temperature recovery time is within the reference range, the abnormal location is determined to be a condensation area and a condensation determination record is generated; when the leakage deviation results show that the humidity continues to rise, the surface dielectric response continues to deviate, the temperature recovery time exceeds the reference range, and the distance between the abnormal location and the adjacent area of ​​the cooling pipe joint or terminal is less than the preset association distance, a fluid leakage determination record is generated.

[0017] S5 specifically includes: updating the humidity alarm threshold, surface dielectric offset threshold, temperature recovery time threshold, and abnormal duration threshold in the area monitoring parameter table based on condensation judgment records, fluid leakage judgment records, and historical judgment results of easily leakable monitoring areas; determining the handling level based on condensation judgment records, fluid leakage judgment records, leakage deviation results, and electrical risk locations, with handling levels including recording observation, enhanced sampling, alarm prompt, alarm shutdown, or maintenance prompt; outputting detection results including area identification, abnormal location, abnormal type, handling level, condensation matching results, leakage deviation results, and adaptive alarm thresholds, and feeding back the updated area monitoring parameter table to the operation and acquisition process.

[0018] The beneficial effects of this invention are as follows:

[0019] This invention establishes a basic mapping model within the cabinet, unifying the association between cooling pipe joints, terminal adjacency areas, cabinet air ducts, and sensor installation locations. This transforms leak detection from single-point sensing to regional assessment, improving the accuracy of the correspondence between abnormal locations and potential leak sources. By acquiring temperature recovery curves, humidity recovery curves, and surface dielectric recovery curves through low-power pulse heating and generating condensation reference features, it can identify the formation and dissipation process of short-term condensation, reducing false alarms caused by environmental humidity fluctuations, load heating, or temperature differences within the cabinet.

[0020] This invention combines the rate of humidity rise, temperature recovery time, surface dielectric offset, load change, and location weights to create a pre-leakage fingerprint for thermo-humid-electric coupling (THFC). This fingerprint can identify micro-leakage at cooling pipe joints before significant condensation forms, improving the sensitivity of early leak detection. By matching the pre-leakage fingerprint with condensation reference features in the same area, condensation determination records and fluid leak determination records are generated separately. This clearly distinguishes between recoverable condensation and actual fluid leaks, improving the reliability of detection results.

[0021] This invention updates the adaptive alarm threshold in the area monitoring parameter table based on condensation determination records, fluid leakage determination records, and historical determination results. This reduces false alarms in condensation-prone areas and increases alarm sensitivity in areas showing early signs of leakage, enhancing the adaptability of the detection method to different operating environments. The invention determines the handling level based on the anomaly type, leakage deviation result, and electrical risk location, and outputs detection results including anomaly location, anomaly type, handling level, and adaptive alarm threshold. This provides a clear basis for alarm shutdown, enhanced sampling, and maintenance prompts, reducing the risk of insulation degradation, terminal creepage, and short circuits in electrical control equipment caused by fluid intrusion. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the adaptive fluid leakage detection method for electrical control equipment according to the present invention;

[0023] Figure 2 This is a schematic diagram of the detection system interface according to an embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of the cabinet layout and area division of the system according to an embodiment of the present invention;

[0025] Figure 4 This is a schematic diagram of the low-power pulse heating and reference acquisition interface according to an embodiment of the present invention;

[0026] Figure 5 This is a schematic diagram of pulse heating execution status monitoring according to an embodiment of the present invention;

[0027] Figure 6 This is a schematic diagram of the condensation reference feature comparison interface in an embodiment of the present invention;

[0028] Figure 7 This is a schematic diagram of the leakage impact path and risk analysis interface in an embodiment of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Example 1: As Figure 1 As shown, this embodiment provides an adaptive detection method for fluid leakage in electrical control equipment, including the following steps:

[0031] S1. Read the cabinet structure information of electrical control equipment, cooling pipe joints, terminal adjacent areas, air ducts and sensor locations inside the cabinet, cabinet temperature, humidity, equipment load and surface dielectric response data, establish a basic mapping model inside the cabinet, and generate a set of easily leaking monitoring areas and a table of monitoring parameters for each area.

[0032] S2. According to the set of easily leakable monitoring areas and the monitoring parameter table, apply low-power pulse heating to the easily leakable monitoring areas, collect temperature recovery curves, humidity recovery curves and surface dielectric recovery curves, deduct the temperature rise changes caused by load fluctuations, and generate condensation reference characteristics.

[0033] S3. Collect and run thermal and humidity electrical data sequences, calculate humidity rise rate, temperature recovery time, surface dielectric offset and load change, and generate a thermal and humidity electrical coupling leakage precursor fingerprint by combining the positional relationship.

[0034] S4. Match the fingerprint of the early signs of thermal-wet-electric coupling leakage with the condensation reference features under the same area identification to obtain the condensation matching results and leakage deviation results, and generate condensation judgment records or fluid leakage judgment records.

[0035] S5. Update the adaptive alarm threshold in the area monitoring parameter table based on the condensation judgment record, fluid leakage judgment record and historical judgment results. Determine the handling level based on the leakage deviation result and electrical risk location. Output the detection result and feed the updated area monitoring parameter table back to the operation acquisition process.

[0036] S1 specifically includes the following sub-steps:

[0037] S110. Read the cabinet structure information, cooling pipe joint position, terminal adjacent area position, cabinet air duct position and sensor installation position of the electrical control equipment to be tested, and establish a basic mapping model inside the cabinet.

[0038] The electrical control equipment to be tested refers to a control cabinet, control box, or control compartment that contains controllers, terminal blocks, power modules, relays, contactors, cooling pipes, or liquid cooling interfaces. The cabinet structure information comes from the equipment installation drawing, cabinet assembly drawing, cooling pipe layout drawing, or on-site calibration data. The location of the cooling pipe joint comes from the cooling pipe installation record or on-site coordinate measurement data. The location of the terminal adjacent area comes from the terminal block layout drawing or wiring terminal coordinates. The location of the air duct inside the cabinet comes from the fan installation location, air inlet location, air outlet location, and guide plate location. The sensor installation location comes from the installation coordinates of the temperature sensor, humidity sensor, surface dielectric detection plate, and miniature heating plate.

[0039] The cabinet's internal basic mapping model uses a preset corner point on the cabinet's base plate as the coordinate origin and establishes spatial mapping data according to the cabinet's width, depth, and height directions. This data is used to record the spatial correspondence between cooling pipe joints, terminal adjacent areas, internal air ducts, and sensors. This cabinet internal basic mapping model is output to S120 and S130 for subsequent data acquisition, location, and screening of areas prone to leakage monitoring.

[0040] For example, to avoid insufficient disclosure of "positional relationships", the left front corner of the cabinet bottom plate can be set as the origin of the coordinate system. One cooling pipe joint can be recorded as area A1, the area within 100mm below the terminal block can be recorded as terminal adjacent area B1, and the area within 300mm in the direction of the fan outlet can be recorded as downstream area C1 of the cabinet air duct. The humidity sensor H1, temperature sensor T1, surface dielectric detection sheet E1, and miniature heating sheet P1 can be bound to A1, B1, or C1 respectively. When the distance between A1 and B1 is 70mm and A1 is located upstream of C1, A1, B1, and C1 are recorded as the same leakage influence link in the basic mapping model inside the cabinet. This leakage influence link is used in S130 to determine whether a leakage-prone monitoring area is formed.

[0041] S120. Based on the cabinet basic mapping model, collect the cabinet temperature data, cabinet humidity data, equipment load data and surface dielectric response data of the electrical control equipment under test under normal operating conditions. Normal operating conditions refer to the state where the cabinet door is closed, the fan is within the set speed range, the equipment has not triggered a leakage alarm and the equipment load is within the preset stable range.

[0042] The cabinet internal temperature data is acquired by a surface-mount temperature sensor, an infrared temperature sensor, or a temperature acquisition module; the cabinet internal humidity data is acquired by a miniature humidity sensor or a humidity acquisition module; the equipment load data is read by a programmable logic controller, frequency converter, servo driver, bus current sensor, or power module controller, including operating current, bus power, duty cycle, or output power; the surface dielectric response data is acquired by capacitive detection pads, impedance detection electrodes, or dielectric detection circuits arranged in the terminal proximity area or on the insulating support surface, to reflect the presence of water film, coolant film, or condensation on the monitored surface. This cabinet internal temperature data, cabinet internal humidity data, equipment load data, and surface dielectric response data are bound to area identifiers in the cabinet's basic mapping model and output to S130 and S220.

[0043] For example, to avoid ambiguity regarding the "normal reference response," temperature, humidity, load, and surface dielectric response data can be continuously collected for 10 minutes under conditions where the equipment load is between 40% and 70% of the rated load, the cabinet door is closed, and the fan speed is stable. The average capacitance response of the surface dielectric detection sheet E1 within region A1 is used as the normal reference response for region A1, the average temperature over the same period is used as the normal temperature reference for region A1, and the average humidity is used as the normal humidity reference for region A1. When S320 subsequently calculates the surface dielectric offset, only the real-time surface dielectric response data of region A1 is compared with the normal reference response of region A1; it is not used in conjunction with the reference responses of other regions.

[0044] S130. Associate the cabinet basic mapping model, cabinet temperature data, cabinet humidity data, equipment load data and surface dielectric response data with regions. According to the proximity relationship of cooling pipe joints, terminal proximity relationship, downstream relationship of cabinet air duct and liquid accumulation path relationship in low-lying areas of the cabinet, select the easy-to-leak monitoring areas from the candidate areas and generate a set of easy-to-leak monitoring areas and a regional monitoring parameter table.

[0045] The leak-prone monitoring area refers to a local area inside the cabinet that meets the criteria of a high probability of leakage or a high impact on electrical safety after leakage. The area monitoring parameter table is used to record the area identification, corresponding sensor number, normal reference response, sampling period, low-power pulse heating parameters, humidity alarm threshold, surface dielectric offset threshold, temperature recovery time threshold, abnormal duration threshold, enhanced sampling threshold, and leakage deviation alarm threshold.

[0046] Candidate areas can be screened by the area correlation distance. The smaller the area correlation distance, the closer the candidate area is to the cooling pipe joint, terminal adjacent area, downstream path of the air duct in the cabinet, or low-lying liquid accumulation path, and the more suitable it is as a leak-prone monitoring area.

[0047]

[0048] in, This represents the regional association distance of the i-th candidate region; This represents the distance from the i-th candidate region to the cooling pipe joint; This represents the distance from the i-th candidate region to the terminal's adjacent region; This represents the distance from the i-th candidate region to the downstream path of the air duct inside the cabinet; This represents the height difference between the i-th candidate region and the path of the low-lying liquid accumulation in the cabinet. , , , These represent the weights corresponding to the distance to the cooling pipe joint, the distance to the adjacent area of ​​the terminal, the downstream path distance of the air duct inside the cabinet, and the height difference, respectively.

[0049] In this embodiment, the setting of each weight coefficient satisfies ,and Preferably, =0.4、 =0.3、 =0.2、 =0.1. This weight allocation highlights the sensitivity of the cooling pipe joint as the primary physical leakage source to the local area, while also taking into account the insulation safety risks of the terminal block and the physical interference of the air duct in the cabinet on moisture diffusion, so that the constructed regional correlation distance has a clear physical diagnostic logic.

[0050] For example, to avoid using relative terms like "close to" or "low-lying," candidate areas that are less than 100mm from the cooling pipe joint, less than 120mm from the terminal adjacent area, within 300mm downstream of the air duct, or more than 5mm below the adjacent support surface can be included in the leak-prone monitoring area set. If area A1 simultaneously meets both conditions—a distance of 60mm from the cooling pipe joint and a distance of 90mm from the terminal adjacent area—then area A1 will be marked as a key monitoring area in the area monitoring parameter table, and its sampling period will be set shorter than that of ordinary areas. The leak-prone monitoring area set and the area monitoring parameter table are output to S210 and S310.

[0051] S2 specifically includes the following sub-steps:

[0052] S210. According to the set of easily leakable monitoring areas and the monitoring parameter table, apply low-power pulse heating to each easily leakable monitoring area and form a pulse heating execution record; low-power pulse heating refers to the process of applying short-term thermal excitation to the monitoring surface through a micro heating element, flexible heating film or resistance heating element, so that the monitoring surface generates a detectable temperature rise but does not change the normal operating state of the electrical control equipment under test.

[0053] The pulse heating execution record includes the area identifier, heating element number, heating power, heating duration, heating start and end times, pre-heating temperature, post-heating temperature, and equipment load data. The safety boundaries for low-power pulse heating are determined based on the area monitoring parameter table; the heating power, heating duration, and allowable temperature rise must not exceed the upper limits specified in the table. If the equipment load experiences a sudden change, the low-power pulse heating operation is paused to prevent the load temperature rise from affecting the recovery curve. The pulse heating execution record is output to S220.

[0054] For example, to avoid insufficient disclosure of "low power", a miniature heating element P1 can be installed on the outer insulating support surface of the terminal adjacent area. The single heating power is set to 0.5W-2W, and the heating duration is set to 5s-20s, so that the temperature rise of the monitoring surface is controlled at 0.5K-2K. If the equipment load current of area A1 is 20A before heating, and the equipment load current suddenly changes to 35A during heating, then this heating is recorded as a pulse heating execution record affected by load interference, and is not used as a valid sample for condensation reference feature extraction in S220.

[0055] S220. After the low-power pulse heating ends, collect the temperature recovery curve, humidity recovery curve, and surface dielectric recovery curve of the corresponding easily leaking monitoring area according to the area monitoring parameter table. Combined with the pulse heating execution record and equipment load data, deduct the temperature rise caused by equipment load fluctuations to obtain the set of area recovery curves. The temperature recovery curve refers to the sequence of the monitored surface temperature falling back over time after the pulse heating ends. The humidity recovery curve refers to the sequence of the humidity change over time in the same area. The surface dielectric recovery curve refers to the sequence of the surface dielectric response data regressing to the normal reference response.

[0056] Load deduction refers to eliminating temperature changes caused by equipment load fluctuations based on the relationship between equipment load data and historical temperature rise, ensuring that the regional recovery curve set primarily reflects changes in condensation adhesion, condensation dissipation, or surface wetness. The regional recovery curve set is output to S230.

[0057] Specifically, the relationship between equipment load data and historical temperature rise is obtained through the following pre-calibration method: During the equipment's factory calibration phase or under normal dry operation with no fluid leakage and no condensation, the electrical control equipment under test is controlled to operate under different load current (or power) steps, and the steady-state temperature rise of each easily leakable monitoring area when reaching thermal equilibrium is continuously recorded, thereby constructing a linear fitting formula for 'load current - steady-state temperature rise':

[0058]

[0059] In the formula, The temperature rise component caused by the load. Let be the real-time equipment load current, k be the temperature rise characteristic coefficient obtained by least squares fitting, and b be a constant correction term. In the calculation of step S220, the system calculates based on the real-time collected equipment load data. The current value is calculated in real time using the above fitting formula. The load temperature rise component is subtracted from the measured temperature recovery curve to eliminate interference.

[0060] For example, to avoid the inability to implement "subtracting load fluctuations," data can be collected for 60 seconds at 1-second sampling intervals after the pulse heating ends. If the equipment load current in region A1 increases from 20A to 30A during the collection period, historical records show that this load change corresponds to a temperature rise of approximately 0.4K on the monitored surface. Therefore, the 0.4K load temperature rise effect is subtracted from the temperature recovery curve, retaining only the recovery curve resulting from the change in the moisture content of the monitored surface. The resulting set of regional recovery curves can be used in S230 to extract the true condensation reference characteristics, rather than mistaking a load temperature rise for undissipated condensation.

[0061] S230. Based on the set of regional recovery curves, extract the temperature recovery time range, humidity decrease slope range, surface dielectric response regression time range, and load correspondence for each easily leakable monitoring area to generate condensation reference characteristics. Condensation reference characteristics refer to the regionalized reference data exhibited when the temperature, humidity, and surface dielectric response recover to the normal baseline state after condensation is formed due to temperature difference or short-term moisture adhesion under conditions without actual fluid leakage.

[0062] Condensation reference characteristics are bound to area identifiers in the set of easily leakable monitoring areas and written into the area monitoring parameter table for S410 to match the precursor fingerprints of thermo-humid-electric coupling leaks. Temperature recovery time refers to the time required for the temperature to drop back to the normal temperature reference allowable range after the pulse heating ends; humidity decrease slope refers to the decrease in humidity per unit time when it drops from the peak to the normal humidity reference; surface dielectric response regression time refers to the time required for the surface dielectric response data to return to the normal reference response allowable range.

[0063] For example, to avoid the "condensation reference characteristics" being merely a result-oriented description, the effective samples formed by multiple pulse heatings of region A1 under leak-free conditions can be statistically analyzed. This yields a temperature recovery time range of 25-40 seconds, a humidity decrease slope range of 0.2%RH-0.5%RH per second, and a surface dielectric response that returns to ±5% of the normal baseline response within 45 seconds. Furthermore, the above recovery process is not continuously synchronized with changes in equipment load. These condensation reference characteristics of region A1 are written into the region monitoring parameter table. Subsequently, S410 only matches the thermal-humidity-electric coupling leakage precursor fingerprint of region A1 with the condensation reference characteristics of region A1.

[0064] S3 specifically includes the following sub-steps:

[0065] S310. During the operation of the electrical control equipment under test, according to the set of easily leakable monitoring areas and the monitoring parameter table of areas, real-time temperature data, real-time humidity data, real-time equipment load data and real-time surface dielectric response data of each easily leakable monitoring area are continuously collected to form an operating thermal and humidity data sequence; the operating thermal and humidity data sequence refers to the synchronous data set of temperature, humidity, equipment load and surface dielectric response with area identification and timestamp.

[0066] Real-time temperature data is acquired by temperature sensors within the area, real-time humidity data by humidity sensors within the area, real-time equipment load data by the controller or load acquisition module of the electrical control equipment under test, and real-time surface dielectric response data by surface dielectric detection plates or impedance detection electrodes. All real-time data are aligned according to the same area identifier and the same time base to ensure that subsequent calculations do not reference data across areas or time periods. The thermo-humidity data sequence is output to the S320.

[0067] For example, to avoid the lack of implementation methods for "continuous acquisition," the sampling period for region A1 can be set to 2 seconds, and the sampling period for region B1 to 5 seconds. During each sampling, the region identifier, sampling time, temperature value, humidity value, real-time surface dielectric response data, and device load current are recorded simultaneously. If temperature, humidity, and real-time surface dielectric response data are acquired in region A1 at 10:00:00, only the device load data from the same time or after timestamp alignment is read, avoiding the use of load data from 10:00:30 to interpret the humidity anomaly at 10:00:00.

[0068] S320. Based on the operating thermo-humidity and electrical data sequence, calculate the humidity rise rate, temperature recovery time, surface dielectric offset, and load change for each leak-prone monitoring area. The humidity rise rate is used to characterize the speed at which the humidity of the area increases per unit time. The temperature recovery time is used to characterize the time required for the temperature of the area to return from an abnormal temperature state to the normal reference range. The surface dielectric offset is used to characterize the degree of deviation of the real-time surface dielectric response data from the normal reference response. The load change is used to distinguish between temperature and humidity changes caused by equipment load increases or decreases and those caused by fluid adhesion.

[0069] When the rate of increase in humidity exceeds the preset start value or the surface dielectric offset exceeds the preset start value, the abnormal start time is used as the starting time for calculating the temperature recovery time.

[0070]

[0071] in, Indicates the rate of increase in humidity; This represents the real-time humidity data at the current moment. Indicates the time before the current moment Real-time humidity data over time; This indicates the preset anomaly detection time window or sampling time interval.

[0072]

[0073] in, Indicates the surface dielectric offset; This represents the real-time surface dielectric response data at the current moment; This indicates the normal baseline response for the corresponding easily leak-prone monitoring area.

[0074]

[0075] in, Indicates the amount of load change; This represents the device load data at the current moment; Indicates the time before the current moment Device load data over time; This indicates the rated load data of the electrical control equipment to be tested.

[0076] For example, to avoid uncertainty about "continuous offset", we can assume that the humidity in region A1 increases from 55%RH to 68%RH within 30 seconds. Then, we can calculate the rate of increase in humidity based on the humidity sampling time interval. At the same time, if the real-time surface dielectric response data of region A1 deviates from the normal reference response by more than 8%, and this deviation lasts for more than 60 seconds, it is recorded as a continuous offset of the surface dielectric response. If the load change is less than 5% at the same time, it means that the temperature and humidity change cannot be fully explained by the increase in load, and we can proceed to S330 to generate a precursor fingerprint of thermal-humidity-electrical coupling leakage.

[0077] S330. Combine the humidity rise rate, temperature recovery time, surface dielectric offset, load change, and the positional relationship of the corresponding leak-prone monitoring area in the cabinet's basic mapping model to generate a thermal-humidity-electric coupling leakage precursor fingerprint. The thermal-humidity-electric coupling leakage precursor fingerprint refers to a set of features used to distinguish between condensation and actual fluid leakage, including area identification, abnormal location, humidity change characteristics, temperature recovery characteristics, dielectric offset characteristics, load association characteristics, abnormal duration, and location weight.

[0078] Location weights are determined based on the spatial distance between the abnormal location and the location of the cooling pipe joint, the terminal adjacent area, and the downstream path of the air duct within the cabinet. This weight is used to characterize the correlation strength between the abnormal location and potential leakage sources or electrical risk structures. This thermal-wet-electric coupling leakage precursor fingerprint is output to S410.

[0079] For example, to avoid insufficient disclosure of "early warning fingerprints," the early warning fingerprint of thermal-humidity-electric coupling leakage generated in region A1 can be represented as follows: Region identifier A1, abnormal location 80mm below the cooling pipe connector, humidity rise rate is continuously increasing, temperature recovery time exceeds the upper limit of the condensation reference characteristic, surface dielectric offset is greater than 8% and lasts for more than 60s, load change is less than 5%, abnormal duration is 120s, and the location weight is set high due to proximity to the cooling pipe connector and terminal adjacency area. Each field in this fingerprint participates in condensation matching or leakage deviation judgment in S410, S420, or S430, without generating useless output.

[0080] S4 specifically includes the following sub-steps:

[0081] S410. Match the fingerprint of the precursor to the thermal-wet-electric coupling leak with the condensation reference features under the same area identification to obtain the condensation matching result and the leakage deviation result. The condensation matching result is used to determine whether the current anomaly conforms to the condensation formation and dissipation law, and the leakage deviation result is used to determine whether the current anomaly deviates from the condensation reference features and has real fluid adhesion characteristics.

[0082] The condensation matching result can be represented by the condensation matching score, and the leakage deviation result can be represented by the leakage deviation score. Before calculation, the temperature recovery similarity, humidity recovery similarity, surface dielectric recovery similarity, load correspondence similarity, humidity rise rate, surface dielectric offset, temperature recovery time and location weight are normalized so that data of different dimensions can participate in the judgment together.

[0083]

[0084] in, Indicates the condensation matching score; Indicates the similarity of temperature recovery curves; Indicates the similarity of humidity recovery curves; Indicates the similarity of surface dielectric recovery curves; Indicates the similarity of load correspondence; , , , These represent the weights corresponding to the similarity of temperature recovery curves, humidity recovery curves, surface dielectric recovery curves, and load correspondence, respectively.

[0085]

[0086] in, This indicates the deviation score from the leakage score; This represents the rate of increase in humidity after normalization. This represents the normalized surface dielectric offset. This indicates the temperature recovery time after normalization. This represents the weight of outlier locations after normalization. , , , These represent the weights corresponding to the humidity rise rate, surface dielectric offset, temperature recovery time, and abnormal location weight, respectively.

[0087] Specifically, the similarity scores of the temperature recovery curve (ST), humidity recovery curve (SH), and surface dielectric recovery curve (SE) are all calculated using the inverse Euclidean distance algorithm. Taking the temperature recovery curve similarity score (ST) as an example, its specific calculation formula is as follows:

[0088]

[0089] In the formula, This represents the value of the temperature recovery curve at the m-th time sampling point after real-time acquisition and load subtraction in the current round. This represents the value of the standard temperature recovery curve corresponding to the condensation reference feature under the same area identification at the m-th time sampling point, where M is the total number of sampling points in the current recovery process; SH and SE are calculated similarly.

[0090] The load correspondence similarity SL is determined by calculating the Pearson correlation coefficient r between the device load change sequence and the real-time temperature change sequence within the current acquisition period, and then taking... The correlation coefficient is used to characterize the degree of decoupling between the temperature recovery process and the internal load fluctuation of the equipment (the closer the correlation coefficient is to 0, the less the temperature rise is caused by the load, and the closer the similarity score is to 1).

[0091] Furthermore, the weighting coefficients are preferably fixed as follows: =0.3, =0.3, =0.3, =0.1; the weighting coefficient is preferably fixed as follows: =0.3, =0.4, =0.2, =0.1. This weighting configuration specifically increases the weighting of the surface dielectric offset (OE) determination. =0.4), to significantly improve the system's sensitivity to capturing microleakage before fluid accumulation in the early stages.

[0092] For example, to avoid the term "matching" being a result-oriented description, the precursor fingerprint of thermal-humidity-electric coupling leakage in region A1 can be compared item by item with the condensation reference characteristics of region A1. If the temperature recovery curve, humidity recovery curve, and surface dielectric recovery curve all fall within the range of the condensation reference characteristics of region A1, the condensation matching score increases. If the humidity continues to rise, the surface dielectric offset continues to exceed the surface dielectric offset threshold, the temperature recovery time exceeds the upper limit of the condensation reference characteristics, and the abnormal location is close to the cooling pipe joint, the leakage deviation score increases. The condensation matching result is output to S420, and the leakage deviation result is output to S430.

[0093] S420. When the condensation matching result shows that the humidity can fall back according to the condensation reference characteristics after the humidity rises, the surface dielectric response does not shift continuously, and the temperature recovery time is within the range of the condensation reference characteristics, the corresponding abnormal location is determined as a condensation area, and a condensation determination record is generated. A condensation area refers to an area where the humidity rises due to temperature difference, short-term moisture adhesion, or changes in the cabinet environment, but can return to the normal baseline state within a preset recovery time.

[0094] The condensation determination record includes the area identifier, condensation area, anomaly type, condensation occurrence time, condensation recovery status, condensation matching score, corresponding condensation reference feature number, and the reason for not triggering fluid leakage; among which, the anomaly type is condensation. This condensation determination record is output to S510 and S520 for subsequent updates to the adaptive alarm threshold, determination of the handling level, and suppression of false alarms caused by recoverable condensation.

[0095] For example, to avoid unclear boundaries between "condensation" and "leakage," it can be stipulated that if the humidity in area B1 rises briefly after the equipment is started, but begins to fall back within 45 seconds, the surface dielectric response returns to ±5% of the normal reference response within 60 seconds, the temperature recovery time remains within the condensation reference characteristic limit range of area B1, and the surface dielectric offset does not exceed the surface dielectric offset threshold within a preset duration, then area B1 is determined to be a condensation area. The condensation determination record states "humidity has fallen back, dielectric response has returned, and temperature recovery time has not exceeded the limit." This record is used in S510 to improve the tolerance of area B1 to recoverable condensation and in S520 to determine the treatment level of the recorded observation.

[0096] S430. When the leakage deviation results indicate that the humidity continues to rise, the surface dielectric response continues to deviate, the temperature recovery time exceeds the limit range of the condensation reference feature, and the regional correlation distance between the abnormal location and the corresponding easy-to-leak monitoring area of ​​the cooling pipe joint location or terminal adjacent area in the cabinet basic mapping model is less than the preset judgment threshold, the corresponding abnormal location will be judged as a fluid leakage area, and a fluid leakage judgment record will be generated. A fluid leakage area refers to an area where there is a risk of continuous adhesion of coolant, water vapor liquefaction or other fluids, and this risk cannot be explained by the condensation reference feature or equipment load changes.

[0097] The fluid leak determination record includes area identification, fluid leak area, leak-related structure, anomaly type, leak duration, leak deviation score, anomaly duration, and electrical risk location; the anomaly type is fluid leak. This fluid leak determination record is output to S510 and S520 to update adaptive alarm thresholds and determine the handling level.

[0098] For example, to avoid the lack of time boundaries for "continuous increase" and "continuous shift," it can be defined that if the humidity in area A1 is located 80mm below the cooling pipe joint, the humidity continuously increases for 120 seconds, the surface dielectric shift continuously exceeds 8%, the temperature recovery time exceeds the upper limit of the condensation reference characteristic of area A1 by more than 30%, and the equipment load change is less than 5%, the increase in humidity and dielectric shift cannot be explained. In this case, area A1 is identified as a fluid leakage area, the leakage-related structure is recorded as the cooling pipe joint, and the electrical risk location is recorded as the terminal adjacent area. This fluid leakage determination record is used in S520 to determine whether to alarm shutdown or provide maintenance prompts.

[0099] S5 specifically includes the following sub-steps:

[0100] S510. Based on the condensation judgment record, fluid leakage judgment record, and the historical judgment results of the corresponding easily leakable monitoring area, update the adaptive alarm threshold in the area monitoring parameter table. The adaptive alarm threshold refers to the alarm judgment parameter that is dynamically adjusted according to the historical status of the area, the frequency of condensation, and the fluid leakage judgment result, including the humidity alarm threshold, surface dielectric offset threshold, temperature recovery time threshold, and abnormal duration threshold.

[0101] If multiple condensation detection records are generated in the same area and all can be recovered, the humidity alarm threshold for that area is increased or the humidity weight is decreased to suppress false condensation alarms. If a fluid leak detection record is generated in the same area, the surface dielectric offset threshold, the abnormal duration threshold, or the location weight is increased to improve the alarm sensitivity for similar leak anomalies. The updated area monitoring parameter table is output to S530 and fed back to the subsequent round S310.

[0102]

[0103] in, This indicates the updated adaptive alarm threshold; This indicates the adaptive alarm threshold before the update; This indicates the observation threshold obtained in this round of detection. This observation threshold is determined based on the trigger value of the corresponding feature in the condensation judgment record or fluid leakage judgment record of this round. This represents the threshold update coefficient.

[0104] Threshold update coefficient The value range is limited to 0.05~0.20 to prevent drastic threshold fluctuations caused by a single disturbance. In this embodiment, its adaptive adjustment logic is as follows: when a condensation determination record is generated in the same area and it is confirmed as recoverable condensation, the value is... Setting the sensitivity threshold to a low empirical value of 0.05 allows for a smooth and gentle increase in the humidity alarm threshold, thereby improving tolerance to environmental fluctuations; when fluid leak detection records are generated in the same area, [the threshold will be adjusted accordingly]. The switch is set to a high-sensitivity empirical value of 0.15 to accelerate convergence and significantly reduce the surface dielectric offset threshold, thereby enabling highly sensitive tracking and early warning of subsequent abnormal states in the region in space.

[0105] For example, to avoid a lack of direction in "adaptive updates," it can be stipulated that if area B1 is determined to have recoverable condensation three times consecutively without any sustained deviation in surface dielectric response, then the humidity alarm threshold for area B1 will be increased by 2%RH-5%RH, or the weight of the humidity rise rate in the leakage deviation score will be reduced. If area A1 has one fluid leakage determination record, then the surface dielectric deviation threshold for area A1 will be reduced from 10% to 8%, and the abnormal duration threshold will be shortened from 120s to 90s. In this way, condensation determination records and fluid leakage determination records will generate threshold update results with different directions, avoiding the treatment of all abnormalities according to the same rules.

[0106] S520. Based on the condensation judgment record, fluid leakage judgment record, leakage deviation results and corresponding electrical risk locations, determine the handling level; the handling level refers to the response level taken by the electrical control equipment under test in response to the current abnormality, including recording and observation, enhanced sampling, alarm prompts, alarm shutdown or maintenance prompts.

[0107] If the anomaly type is condensation and the condensation recovery status is "recovered," the handling level is "record and observe." If the leakage deviation score is higher than the enhanced sampling threshold but lower than the alarm threshold, the anomaly type is determined to be a suspected fluid leak. If the anomaly type is a suspected fluid leak, the anomaly location is not close to the terminal adjacent area, and the leakage deviation score is in the medium range, the handling level is "enhanced sampling." If the anomaly type is a fluid leak, and the anomaly location corresponds to the terminal adjacent area or the cooling pipe joint location, the handling level is "alarm shutdown or maintenance prompt." Both the enhanced sampling threshold and the alarm threshold are provided by the area monitoring parameter table. The handling level is output to the S530 to generate the detection results.

[0108] For example, to avoid unclear boundaries of "handling levels," the handling level can be defined as "record and observe" if the anomaly type in area B1 is condensation that recovers within 60 seconds; if the humidity in area C1 increases but the surface dielectric offset does not exceed the threshold, and the anomaly location is far from the terminal proximity area, the handling level is "enhanced sampling"; if area A1 is determined to be a fluid leakage area and is located within 100mm of the terminal proximity area or below the cooling pipe joint, the handling level is "alarm shutdown or maintenance prompt." This handling level, along with the updated adaptive alarm threshold in S510, is written into the final detection result.

[0109] S530 outputs the detection results and feeds back the updated area monitoring parameter table to the S310 operation and acquisition process; the detection results include area identification, abnormal location, abnormal type, handling level, condensation matching result, leakage deviation result, adaptive alarm threshold and corresponding easy-to-leak monitoring area identification.

[0110] The test results can be output to the local controller, host computer, alarm module or maintenance platform of the electrical control equipment under test; the updated area monitoring parameter table is written into the running acquisition task, so that subsequent rounds of testing are executed according to the updated sampling period, humidity alarm threshold, surface dielectric offset threshold, temperature recovery time threshold and abnormal duration threshold.

[0111] Therefore, the cabinet basic mapping model established in S110, the set of easily leakable monitoring areas generated in S130, the condensation reference features generated in S230, the thermal-humid-electric coupling leakage precursor fingerprint generated in S330, and the judgment records generated in S420 and S430 all form result output and closed-loop feedback in S530.

[0112] For example, to avoid insufficient disclosure of the "feedback loop", the final detection result can be output as follows: Area identifier A1, abnormal location is 80mm below the cooling pipe joint, abnormality type is fluid leakage, handling level is alarm shutdown, condensation matching score is lower than the condensation matching threshold, leakage deviation score is higher than the leakage deviation threshold, and the updated humidity alarm threshold, surface dielectric offset threshold, and temperature recovery time threshold have been written into the area monitoring parameter table; subsequently, in subsequent rounds S310, the system performs shorter sampling cycles and higher sensitivity detection on area A1 according to the updated area monitoring parameter table, thereby completing the adaptive detection closed loop for fluid leakage of electrical control equipment.

[0113] Example 2: Figures 2 to 7 As shown, this embodiment provides an adaptive fluid leakage detection system for electrical control equipment, including:

[0114] The area modeling module is used to read the cabinet structure information, cooling pipe joint location, terminal adjacent area location, cabinet air duct location and sensor installation location of electrical control equipment, and to collect cabinet temperature data, cabinet humidity data, equipment load data and surface dielectric response data, establish a basic mapping model inside the cabinet, and generate a set of easily leaking monitoring areas and a table of area monitoring parameters.

[0115] The condensation reference generation module is used to apply low-power pulse heating to the leak-prone monitoring areas according to the set of leak-prone monitoring areas and the area monitoring parameter table, collect temperature recovery curves, humidity recovery curves and surface dielectric recovery curves, and deduct the temperature rise changes caused by equipment load fluctuations to generate condensation reference characteristics bound to the area identifier.

[0116] The precursor fingerprint generation module is used to collect the operating thermal and humidity data sequence during the operation of electrical control equipment. Based on the operating thermal and humidity data sequence, it calculates the humidity rise rate, temperature recovery time, surface dielectric offset and load change, and combines the positional relationship of the easy-to-leak monitoring area in the cabinet basic mapping model to generate the precursor fingerprint of thermal and humidity coupling leakage.

[0117] The anomaly determination module is used to match the early fingerprint of thermal-wet-electric coupling leakage with the condensation reference features under the same area identification to obtain condensation matching results and leakage deviation results, and generate condensation determination records based on condensation matching results, or generate fluid leakage determination records based on leakage deviation results.

[0118] The adaptive output module is used to update the adaptive alarm threshold in the area monitoring parameter table based on the condensation judgment record, fluid leakage judgment record and historical judgment results of the easily leakable monitoring area, and determine the handling level based on the leakage deviation result and electrical risk location. It outputs the detection results including area identification, abnormal location, abnormality type, handling level, condensation matching result, leakage deviation result and adaptive alarm threshold, and feeds back the updated area monitoring parameter table to the precursor fingerprint generation module.

[0119] All the above formulas are performed using dimensionless numerical calculations; the relevant formulas are based on empirical models that approximate the real situation, obtained through extensive data collection and software simulation fitting. The preset parameters and thresholds involved in the formulas can be conventionally set and adjusted by those skilled in the art according to the physical constraints of the actual application scenario.

[0120] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0121] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0122] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An adaptive detection method for fluid leakage in electrical control equipment, characterized in that, Includes the following steps: S1. Read the cabinet structure information of the electrical control equipment, as well as the detection and response data inside the cabinet, establish the basic mapping model inside the cabinet, and generate a set of easily leaking monitoring areas and a table of monitoring parameters for the areas. S2. According to the set of easily leakable monitoring areas and the monitoring parameter table, apply low-power pulse heating to the easily leakable monitoring areas, collect feedback data, and generate condensation reference characteristics; S3. Collect the operating thermo-humidity and electrical data sequence, calculate the change in the corresponding data, and generate a fingerprint of the early signs of thermo-humidity and electrical coupling leakage by combining the positional relationship; S4. Match the fingerprint of the early signs of thermal-wet-electric coupling leakage with the condensation reference features under the same area identification to obtain the condensation matching result and the leakage deviation result, and generate a condensation judgment record or a fluid leakage judgment record.

2. The adaptive fluid leakage detection method for electrical control equipment according to claim 1, characterized in that, Also includes: S5. Update the adaptive alarm threshold in the area monitoring parameter table based on the condensation judgment record, fluid leakage judgment record and historical judgment results, determine the handling level based on the leakage deviation result and electrical risk location, and output the detection result.

3. The adaptive fluid leakage detection method for electrical control equipment according to claim 1, characterized in that, S1 specifically includes: Read the cabinet structure information, cooling pipe joint position, terminal adjacent area position, cabinet air duct position and sensor installation position of electrical control equipment, and establish the cabinet basic mapping model with the preset corner point of the cabinet bottom plate as the coordinate origin; Based on the cabinet's basic mapping model, under normal operating conditions, the cabinet's temperature data, humidity data, equipment load data, and surface dielectric response data are collected and bound to form a normal reference response.

4. The adaptive fluid leakage detection method for electrical control equipment according to claim 3, characterized in that, Also includes: The basic mapping model inside the cabinet and the data collection area are associated. The leakage-prone monitoring areas are screened according to the proximity of cooling pipe joints, the proximity of terminals, the downstream relationship of air ducts inside the cabinet, and the path relationship of liquid accumulation in low-lying areas of the cabinet. A set of leakage-prone monitoring areas and a table of regional monitoring parameters are generated.

5. The adaptive fluid leakage detection method for electrical control equipment according to claim 1, characterized in that, S2 specifically includes: According to the set of easily leakable monitoring areas and the monitoring parameter table, low-power pulse heating is applied to the easily leakable monitoring areas, and the area identification, heating power, duration, start and end time, temperature before and after heating and load data are recorded to form a pulse heating execution record; After heating is completed, temperature recovery curve, humidity recovery curve and surface dielectric recovery curve are collected. Combined with pulse heating execution record and equipment load data, the temperature rise caused by load fluctuation is deducted to obtain a set of regional recovery curves. Based on the set of regional recovery curves, the range of temperature recovery time, the range of humidity decrease slope, the range of surface dielectric response regression time, and the load correspondence are extracted to generate condensation reference features bound to the regional identifier.

6. The adaptive fluid leakage detection method for electrical control equipment according to claim 1, characterized in that, S3 specifically includes: According to the set of easily leakable monitoring areas and the monitoring parameter table for each area, real-time temperature data, real-time humidity data, real-time equipment load data and real-time surface dielectric response data of each easily leakable monitoring area are continuously collected during operation to form a sequence of operating thermo-humidity and electrical data with area identifiers and timestamps. Based on the running thermo-humidity and electrical data sequence, the humidity rise rate, temperature recovery time, surface dielectric offset, and load change are calculated.

7. The adaptive fluid leakage detection method for electrical control equipment according to claim 6, characterized in that, Also includes: The above calculation results and corresponding positional relationships are combined to generate a thermal-humid-electric coupling leakage precursor fingerprint that includes regional identification, abnormal location, humidity change characteristics, temperature recovery characteristics, dielectric offset characteristics, load association characteristics, abnormal duration, and positional weight.

8. The adaptive fluid leakage detection method for electrical control equipment according to claim 1, characterized in that, S4 specifically includes: The fingerprint of the early leakage of thermal-humid-electric coupling is matched with the condensation reference features under the same area identification. The normalized comparison of the correspondence between temperature recovery, humidity recovery, surface dielectric recovery and load is obtained to obtain the condensation matching result and the leakage deviation result. When the condensation matching results show that the humidity rises and then falls back, the surface dielectric response does not continue to shift, and the temperature recovery time is within the reference range, the abnormal location is determined to be a condensation area, and a condensation determination record is generated. When the leakage deviation results indicate that the humidity continues to rise, the surface dielectric response continues to deviate, the temperature recovery time exceeds the reference range, and the distance between the abnormal location and the adjacent area of ​​the cooling pipe joint or terminal is less than the preset associated distance, a fluid leakage judgment record is generated.

9. The adaptive fluid leakage detection method for electrical control equipment according to claim 2, characterized in that, S5 specifically includes: Based on the condensation judgment records, fluid leakage judgment records, and historical judgment results of easily leaking monitoring areas, update the humidity alarm threshold, surface dielectric offset threshold, temperature recovery time threshold, and abnormal duration threshold in the area monitoring parameter table. The handling level is determined based on the condensation judgment record, fluid leakage judgment record, leakage deviation results and electrical risk location. The handling level includes recording and observation, enhanced sampling, alarm prompt, alarm shutdown or maintenance prompt.

10. The adaptive fluid leakage detection method for electrical control equipment according to claim 9, characterized in that, Also includes: The output includes the detection results such as area identifier, abnormal location, abnormal type, handling level, condensation matching result, leakage deviation result, and adaptive alarm threshold, and feeds back the updated area monitoring parameter table to the operation and acquisition process.

Images