Branch box maintenance cycle prediction method based on historical data

Abstract

The present application relates to power distribution equipment operation and maintenance technical field, specifically to a branch box maintenance cycle prediction method based on historical data, comprising: obtaining the internal cavity temperature and humidity time series data and external environment temperature sudden drop gradient data of the target closed power distribution branch box; inputting the preset condensation along the surface creepage prediction model, calculating the water droplet condensation rate and water film thickness at the cable inlet room wall bushing based on the closed space thermodynamics phase change parameters; mapping it to the busbar interphase insulation clearance parameter of the target closed power distribution branch box, calculating the dynamic creepage distance degradation of the interphase insulation baffle surface; comparing the degradation with the preset interphase flashover breakdown critical value; in response to reaching the critical value, generating a power distribution switchgear dehumidification maintenance work order and sending it to the power distribution asset management master station, and establishing a preventive power outage maintenance schedule. The present application quantifies the influence of microscopic condensation phase change on electrical insulation clearance, and realizes accurate scheduling of equipment maintenance.

Description

Technical Field

[0001] This invention relates to the field of power distribution equipment operation and maintenance technology, specifically a method for predicting the maintenance cycle of branch boxes based on historical data. Background Technology

[0002] In modern power distribution systems, metal-enclosed distribution boxes serve as critical switching and distribution devices, widely deployed at outdoor cable junctions. Because these devices are exposed to complex and changing natural environments for extended periods and possess highly sealed enclosures, the stability of their internal microclimate directly determines the electrical operational safety of the internal busbars, wall bushings, and insulating partitions.

[0003] Existing anti-condensation maintenance of enclosed power distribution equipment mainly relies on fixed-cycle manual inspections or basic temperature and humidity controllers for threshold management (i.e., directly activating heaters to remove moisture when the internal humidity reaches a certain static threshold). However, this conventional technology has significant drawbacks: First, a single and fixed humidity threshold cannot reflect the rapid thermodynamic phase change process caused by a sudden drop in external ambient temperature, making it prone to forming hidden condensation at physical interfaces with intense heat exchange, such as cable entry chamber wall bushings. Second, existing technologies view temperature and humidity data in isolation, failing to map them to core physical dimensions of the power distribution equipment, such as the inter-phase insulation clearance of busbars and the surface contamination and salt density. Therefore, it is impossible to accurately quantify the dynamic creepage distance attenuation caused by the water film bridging effect on the surface of the insulation baffle. This disconnect leads to a lack of scientific prediction of the depth of insulation degradation in the actual operation and maintenance system, and preventive power outage scheduling often relies on subjective experience, making it difficult to balance power supply reliability and maintenance economy.

[0004] Existing maintenance methods for enclosed power distribution equipment cannot dynamically and accurately quantify the deteriorating impact of closed microclimate phase changes on the actual electrical insulation clearance of the equipment. This leads to serious delays in the scheduling of anti-condensation maintenance or blind dispatching, which can easily cause serious power outage accidents such as phase-to-phase flashover breakdown of branch boxes.

[0005] To address this, a branch box maintenance cycle prediction method based on historical data is proposed. Summary of the Invention

[0006] The purpose of this invention is to provide a method for predicting the maintenance cycle of branch boxes based on historical data. By quantifying the dynamic impact of condensation phase change on electrical insulation clearance, the method can achieve accurate scheduling of equipment and avoid phase-to-phase flashover breakdown accidents.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for predicting branch box maintenance cycles based on historical data includes: Acquire time-series data of internal temperature and humidity of the target enclosed power distribution branch box and data of the gradient of sudden temperature drop in the external environment; The internal temperature and humidity time series data and the external ambient temperature drop gradient data are input into the preset condensation creepage prediction model. Based on the thermodynamic phase change parameters of the enclosed space, the water droplet condensation rate and water film thickness at the cable inlet bushing of the target enclosed distribution branch box are calculated. The water droplet condensation rate and the water film thickness are mapped to the bus phase-to-phase insulation clearance parameters of the target enclosed distribution branch box, and the dynamic creepage distance degradation degree of the phase-to-phase insulation baffle surface is calculated. Compare the dynamic creepage distance degradation with the preset phase-to-phase flashover breakdown threshold; When the dynamic creepage distance degradation reaches the phase-to-phase flashover breakdown critical value, a dehumidification maintenance work order for the power distribution switchgear is generated. Send the dehumidification maintenance work order of the power distribution switchgear to the power distribution asset management master station to establish a preventive power outage maintenance schedule for the target enclosed power distribution branch box.

[0008] Preferably, the step of acquiring the time series data of internal cavity temperature and humidity and the gradient data of sudden temperature drop in the external environment includes: A temperature and humidity sensor unit installed inside the target enclosed power distribution branch box collects the original internal temperature and humidity sequence at a first preset sampling frequency; a preset outlier removal algorithm is used to perform data cleaning and timestamp alignment on the original internal temperature and humidity sequence to generate the internal temperature and humidity time series data; a micro-meteorological sensor terminal deployed outside the target enclosed power distribution branch box collects the external ambient temperature sequence at a second preset sampling frequency; the time derivative of the external ambient temperature sequence within adjacent sampling periods is calculated to obtain the temperature change rate; it is determined whether the temperature change rate reaches a preset temperature drop threshold; in response to the temperature change rate reaching the temperature drop threshold, the temperature change rate reaching the preset temperature drop threshold is extracted as the external ambient temperature drop gradient data.

[0009] Preferably, the condensation creepage prediction model includes a cavity thermodynamic evaluation unit, an interface phase change calculation unit, and a surface water film integration unit. The cavity thermodynamic evaluation unit receives the internal cavity temperature and humidity time series data and the external ambient temperature drop gradient data, extracts the saturated vapor pressure parameter and the current absolute vapor pressure parameter of the condensation surface of the target enclosed power distribution branch box based on the thermodynamic phase change parameters of the sealed space, and calculates the pressure difference between the saturated vapor pressure parameter and the current absolute vapor pressure parameter. The interface phase change calculation unit extracts the preset surface mass transfer coefficient of the cable inlet chamber wall bushing, multiplies the surface mass transfer coefficient by the pressure difference value output by the cavity thermodynamic evaluation unit, and calculates the water droplet condensation rate at the cable inlet chamber wall bushing. The surface water film integration unit performs time integration on the water droplet condensation rate output by the interface phase change calculation unit within a preset continuous monitoring time window to generate the water film thickness.

[0010] Preferably, the busbar phase-to-phase insulation clearance parameter is obtained by: analyzing the structural dimension database of the target enclosed distribution branch box, extracting the initial physical creepage distance value of the phase-to-phase insulation baffle and the minimum air gap value between two adjacent phase busbars; and combining the initial physical creepage distance value and the minimum air gap value to generate the busbar phase-to-phase insulation clearance parameter.

[0011] Preferably, the dynamic creepage distance degradation degree of the phase-to-phase insulation baffle surface is calculated as follows: The two-dimensional spatial insulation feature vector corresponding to the bus phase-to-phase insulation clearance parameter is analyzed, and the initial physical creepage distance value and the minimum air gap value are extracted; the feature ratio of the initial physical creepage distance value and the minimum air gap value is calculated and defined as the spatial electric field coupling enhancement factor; the water film thickness and the spatial electric field coupling enhancement factor are input into a preset water film conductivity effect evaluation function, and the pre-recorded surface contamination salt density parameter is retrieved to calculate the water film conductivity penalty coefficient on the phase-to-phase insulation baffle surface; the initial physical creepage distance value is multiplied by the water film conductivity penalty coefficient to calculate the water film surface bridging distance value; the initial physical creepage distance value is subtracted from the water film surface bridging distance value to calculate the current effective creepage distance value; the difference between the initial physical creepage distance value and the current effective creepage distance value is calculated, and the difference is divided by the initial physical creepage distance value to obtain the dynamic creepage distance degradation degree.

[0012] Preferably, the phase-to-phase flashover breakdown critical value is obtained through the following steps: extracting historical phase-to-phase flashover fault records of the same type of enclosed distribution branch box from the preset distribution asset management master station database; parsing the historical phase-to-phase flashover fault records, retrieving the time series data of temperature and humidity inside the fault cavity and the gradient data of sudden temperature drop in the external environment within a preset tracing time window before the fault occurred; inputting the time series data of temperature and humidity inside the fault cavity and the gradient data of sudden temperature drop in the external environment into the condensation surface creepage prediction model to calculate the maximum cumulative water film thickness before the fault; mapping the maximum cumulative water film thickness before the fault to the bus phase-to-phase insulation clearance parameter to calculate the historical limit creepage distance degradation degree; extracting the pre-entered insulation safety margin coefficient; multiplying the historical limit creepage distance degradation degree by the insulation safety margin coefficient to obtain the phase-to-phase flashover breakdown critical value.

[0013] Preferably, the dehumidification maintenance work order for the power distribution switchgear includes the following structured data fields: a unique asset identification code and spatial location coordinates pointing to the target enclosed power distribution branch box; a record of the dynamic creepage distance degradation value on the surface of the phase-to-phase insulation partition that triggers the alarm; a set preventive power outage maintenance mandatory execution timestamp based on the remaining time for the dynamic creepage distance degradation value to reach the phase-to-phase flashover breakdown critical value; a remote mandatory activation command pointing to the anti-condensation heating device inside the target enclosed power distribution branch box; and a list of cleaning actions for the insulation surface of the through-wall bushing in the cable entry compartment and a baseline for the quantity of silica gel desiccant to be replaced.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention introduces a surface creepage prediction model for condensation. By extracting the temperature and humidity gradients of the internal cavity and the sudden drop in external temperature, and combining them with the thermodynamic phase change parameters of the enclosed space, it accurately calculates the condensation rate of microscopic water droplets and the dynamic water film thickness at the wall bushing of the incoming line chamber. This mechanism upgrades lagging environmental monitoring to forward-looking microclimate phase change prediction, locking in the evolution trajectory of condensation hazards from a physical perspective.

[0015] 2. This invention provides an in-depth analysis of the structural dimensions of the enclosed branch box, deeply integrating core physical quantities such as water film thickness, inter-phase insulation clearance, and surface contamination density, to derive the dynamic creepage distance degradation degree on the surface of the inter-phase insulation baffle. This approach transforms the invisible phenomenon of water vapor condensation in the sealed cavity into an intuitive and quantifiable electrical surface bridging indicator, thereby improving the confidence level of equipment health status assessment.

[0016] 3. This invention changes the subjective power outage scheduling model that relies on manual experience. By reverse-analyzing historical phase-to-phase flashover fault records of the same model of equipment, an absolutely objective phase-to-phase flashover breakdown threshold is deduced. When the real-time degradation level reaches this threshold, the system automatically issues a structured dehumidification maintenance work order containing spatial coordinates, a forced execution timestamp, remote heating activation instructions, and cleaning and replacement criteria. This technology not only avoids both "under-maintenance" and "over-maintenance," but also establishes a lean operation and maintenance closed loop from bottom-level status perception to top-level management. Attached Figure Description

[0017] Figure 1 This is a flowchart of a branch box maintenance cycle prediction method based on historical data proposed in this invention; Figure 2 This is a flowchart of a branch box maintenance cycle prediction method based on historical data proposed in this invention. Figure 3 This is a model structure diagram of the surface creepage prediction model for condensation proposed in this invention. Detailed Implementation

[0018] 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.

[0019] Please see Figures 1 to 3 This invention provides a method for predicting the maintenance cycle of branch boxes based on historical data, the technical solution of which is as follows: A method for predicting branch box maintenance cycles based on historical data, such as Figures 1-2 As shown, it includes: Acquire time-series data of internal temperature and humidity of the target enclosed power distribution branch box and data of the gradient of sudden temperature drop in the external environment; The internal temperature and humidity time series data and the external ambient temperature drop gradient data are input into the preset condensation creepage prediction model. Based on the thermodynamic phase change parameters of the enclosed space, the water droplet condensation rate and water film thickness at the cable inlet bushing of the target enclosed distribution branch box are calculated. The water droplet condensation rate and the water film thickness are mapped to the bus phase-to-phase insulation clearance parameters of the target enclosed distribution branch box, and the dynamic creepage distance degradation degree of the phase-to-phase insulation baffle surface is calculated. Compare the dynamic creepage distance degradation with the preset phase-to-phase flashover breakdown threshold; When the dynamic creepage distance degradation reaches the phase-to-phase flashover breakdown critical value, a dehumidification maintenance work order for the power distribution switchgear is generated. Send the dehumidification maintenance work order of the power distribution switchgear to the power distribution asset management master station to establish a preventive power outage maintenance schedule for the target enclosed power distribution branch box.

[0020] Further, the steps for acquiring time-series data of internal cavity temperature and humidity and gradient data of sudden temperature drop in the external environment include: A temperature and humidity sensor unit installed inside the target enclosed power distribution branch box collects the original internal temperature and humidity sequence at a first preset sampling frequency; a preset outlier removal algorithm is used to perform data cleaning and timestamp alignment on the original internal temperature and humidity sequence to generate the internal temperature and humidity time series data; a micro-meteorological sensor terminal deployed outside the target enclosed power distribution branch box collects the external ambient temperature sequence at a second preset sampling frequency; the time derivative of the external ambient temperature sequence within adjacent sampling periods is calculated to obtain the temperature change rate; it is determined whether the temperature change rate reaches a preset temperature drop threshold; in response to the temperature change rate reaching the temperature drop threshold, the temperature change rate reaching the preset temperature drop threshold is extracted as the external ambient temperature drop gradient data.

[0021] The first preset sampling frequency is set to 1Hz to 0.1Hz, and the second preset sampling frequency is set to 0.1Hz to 0.016Hz. The preset outlier removal algorithm specifically employs an outlier detection method based on the 3σ criterion (Laida criterion). The process involves calculating the mean and standard deviation of the original internal cavity temperature and humidity sequence within a set sliding time window. Data points deviating from the mean by more than three times the standard deviation are identified as abnormal electromagnetic interference data and removed. Subsequently, Lagrange interpolation is used to re-interpolate the removed gaps. Numerical completion is performed to complete the timestamp alignment process; the preset temperature drop threshold is specifically set to a range of -0.2℃ / min to -0.5℃ / min. The preset temperature drop threshold is obtained by analyzing the historical data of the same type of enclosed distribution branch box in the artificial climate chamber for high and low temperature alternating condensation test in the distribution asset management master station database, differentiating the external environmental temperature time series curve within 15 minutes before the initial condensation on the inner wall, and setting the threshold to the critical minimum value of the derivative of the time series curve.

[0022] This embodiment employs differentiated sampling frequencies combined with the 3σ criterion and Lagrange interpolation, effectively filtering out complex electromagnetic interference from the power distribution network and ensuring the fidelity of multidimensional time-series data and absolute alignment of timestamps. Furthermore, its temperature drop threshold is scientifically tuned based on the derivative of real historical experimental curves in an artificial climate chamber, achieving precise early detection 15 minutes before initial condensation on the equipment, thus enhancing the sensitivity of microclimate change perception and the reliability of underlying data.

[0023] Furthermore, such as Figure 3 As shown, the condensation creepage prediction model includes a cavity thermodynamic evaluation unit, an interface phase change calculation unit, and a surface water film integration unit. The cavity thermodynamic evaluation unit receives the internal cavity temperature and humidity time series data and the external ambient temperature drop gradient data. Based on the thermodynamic phase change parameters of the sealed space, it extracts the saturated vapor pressure parameter and the current absolute vapor pressure parameter of the condensation surface of the target closed distribution branch box, and calculates the pressure difference between the saturated vapor pressure parameter and the current absolute vapor pressure parameter. The interface phase change calculation unit extracts the preset surface mass transfer coefficient of the cable inlet chamber wall bushing, multiplies the surface mass transfer coefficient by the pressure difference value output by the cavity thermodynamic evaluation unit, and calculates the water droplet condensation rate at the cable inlet chamber wall bushing. The surface water film integration unit performs time integration on the water droplet condensation rate output by the interface phase change calculation unit within a preset continuous monitoring time window to generate the water film thickness.

[0024] The thermodynamic phase transition parameters of the enclosed space specifically include the molar mass of water vapor. (Take 0.018 kg / mol), the universal gas constant R (take 8.314 J / (mol·K)), and the latent heat of vaporization of water at a specific temperature. The cavity thermodynamic evaluation unit extracts the saturated water vapor pressure parameters. Compared with the current absolute water vapor pressure parameters Specifically, the Magnus-Tertens empirical formula is used. The real-time temperature from the time series data of the internal cavity temperature and humidity. Substituting the values, we obtain the saturated water vapor pressure parameters. And combined with the real-time relative humidity in the time series data of the internal cavity temperature and humidity ,according to The current absolute water vapor pressure parameters are obtained. The pressure difference value .

[0025] The preset surface mass transfer coefficient of the cable inlet chamber wall bushing The specific value range is set to 0.005 m / s to 0.025 m / s, and this surface mass transfer coefficient... The value of can be determined by referring to the Chilton-Colburn analogy correlation for convective mass transfer; the surface mass transfer coefficient is an inherent property parameter of the insulation component obtained by placing the cable inlet bushing of the same material and process in an artificial climate test chamber for temperature and humidity alternating boundary layer wind tunnel test, and calibrating and fitting based on the empirical correlation of convective mass transfer.

[0026] The preset continuous monitoring time window is specifically set to 15 to 60 minutes. The specific length of this time window is obtained by calculating the average physical hysteresis period from the time the external ambient temperature drops sharply to the time when the condensation film thickness on the inner wall of the target enclosed distribution branch box reaches a dynamic equilibrium state, based on the historical database of the distribution asset management master station, and adding a redundancy constant.

[0027] This embodiment introduces the Magnus-Tertens empirical formula and a proprietary surface mass transfer coefficient calibrated by wind tunnel tests to achieve precise physical quantification of the condensation rate of microscopic water droplets at the wall-penetrating sleeve. Simultaneously, a continuous monitoring window of 15 to 60 minutes is scientifically tuned based on historical physical hysteresis periods, perfectly aligning with the unique thermodynamic phase transition laws of the sealed power distribution cavity. This mechanism overcomes the lag and blindness of traditional empirical threshold estimations, improving the fidelity of dynamic water film thickness integral calculations and the reliability of flashover prevention early warning.

[0028] Furthermore, the busbar phase-to-phase insulation clearance parameter is obtained by: analyzing the structural dimension database of the target enclosed distribution branch box, extracting the initial physical creepage distance value of the phase-to-phase insulation baffle and the minimum air gap value between two adjacent phase busbars; combining the initial physical creepage distance value and the minimum air gap value to generate the busbar phase-to-phase insulation clearance parameter.

[0029] The structural dimension database of the target enclosed distribution branch box is specifically a digital twin space ledger of the equipment built based on the three-dimensional CAD digital design drawings and insulation type test factory reports of the power distribution equipment at the time of its manufacture. The ledger data is pre-stored in the server of the power distribution asset management master station for remote system access. The specific process of generating the busbar phase-to-phase insulation clearance parameter by combining the initial physical creepage distance value and the minimum air gap value is as follows: The initial physical creepage distance value is used as the first feature dimension, and the minimum air gap value is used as the second feature dimension. A matrix concatenation operation is used to construct a two-dimensional spatial insulation feature vector from these two dimensions. This two-dimensional spatial insulation feature vector is then defined as the busbar phase-to-phase insulation clearance parameter. Here, the initial physical creepage distance value is denoted as... (Unit: mm), the minimum air gap value is denoted as (Unit: mm) and These correspond to the creepage distance and clearance specified in GB / T 16935.1 (IEC 60664-1), respectively. They are treated as two independent scalars for subsequent joint mapping calculation of dynamic creepage distance degradation.

[0030] This embodiment relies on 3D CAD digital design drawings and factory reports to construct a digital twin spatial ledger for the equipment, achieving accurate extraction of busbar phase-to-phase insulation baseline data and avoiding the risk of deviations from manual measurements. Simultaneously, matrix splicing operations are used to construct a two-dimensional spatial insulation feature vector from the initial creepage distance and minimum air gap. This mechanism abandons traditional one-dimensional scalar mathematical operations, preserving the multi-dimensional spatial physical properties of the branch box phase-to-phase insulation system. This provides a high-fidelity structured input for subsequent joint mapping calculations of degradation degree, ensuring the reliability of phase-to-phase flashover early warning from the underlying data dimension.

[0031] Further, the dynamic creepage distance degradation degree on the surface of the phase-to-phase insulation baffle is calculated: the two-dimensional spatial insulation feature vector corresponding to the bus phase-to-phase insulation clearance parameter is analyzed, and the initial physical creepage distance value and the minimum air gap value are extracted; the feature ratio of the initial physical creepage distance value and the minimum air gap value is calculated and defined as the spatial electric field coupling enhancement factor; it should be noted that this feature ratio... This reflects the degree of geometric asymmetry between the surface creepage path and the air breakdown path in the phase-to-phase insulation structure, according to GB / T 16935.1 (IEC 16935.1). (60664-1) Regarding the provisions on insulation coordination, the ratio of creepage distance to clearance is a core geometric parameter characterizing the anti-flashover capability of the insulation structure. Under the condition of coexistence of water film and pollution, this ratio amplifies the surface conductive bridging effect through nonlinear exponential mapping. The water film thickness and the spatial electric field coupling enhancement factor are input into a preset water film conductivity effect evaluation function, and the pre-recorded surface pollution salt density parameter is retrieved to calculate the water film conductivity penalty coefficient on the surface of the interphase insulation baffle. The initial physical creepage distance value is multiplied by the water film conductivity penalty coefficient to calculate the water film surface bridging distance value. The initial physical creepage distance value is subtracted from the water film surface bridging distance value to calculate the current effective creepage distance value. The difference between the initial physical creepage distance value and the current effective creepage distance value is calculated, and the difference is divided by the initial physical creepage distance value to obtain the dynamic creepage distance degradation degree.

[0032] The specific value range of the pre-entered surface contamination salt density parameter is set as follows: the equivalent salt density (ESDD) ranges from 0.03 mg / cm² to 0.35 mg / cm², and the non-soluble deposit density (NSDD) ranges from 0.1 mg / cm² to 5.0 mg / cm². The above value ranges are determined with reference to the pollution level classification standard of IEC TS 60815-1. This surface contamination salt density parameter is obtained by maintenance personnel periodically sampling and wiping the surface of the phase-to-phase insulation baffles of decommissioned enclosed distribution branch boxes deployed in the same operating environment according to the anti-flashover standard level of the target area, and the average salt density benchmark value obtained from the test is pre-entered into the system database through the distribution asset management master station.

[0033] The preset water film conductivity effect evaluation function is a nonlinear penalty exponential function that incorporates spatial electric field distortion effect, based on the flashover test of the insulating component. Its specific mathematical model is as follows: ; Where k is the calculated conductivity penalty coefficient of the water film, δ is the thickness of the water film (unit: μm), and S is the surface fouling salt density parameter (unit: mg / cm²). The hydrophilicity attenuation constant of the insulating material is given in μm⁻¹·(mg / cm²)⁻¹. The initial physical creepage distance is the value (unit: mm). The minimum air gap value (unit: mm). That is, the spatial electric field coupling enhancement factor derived analytically from the two-dimensional spatial insulation eigenvector. The electric field distortion amplification index (usually ranging from 1.2 to 1.5).

[0034] The hydrophilic attenuation constant of the insulating material The inherent constants of the equipment are obtained by calibration through the following steps: (1) Referring to the solid layer method specified in GB / T 4585 (IEC 60507) "Artificial Pollution Test for High Voltage Insulators for AC Systems", a pollution layer of different salt density levels (salt density S is set to 0.03, 0.06, 0.10, 0.25, and 0.35 mg / cm²) is uniformly coated on the interphase insulation baffle test specimens of the same material in an artificial climate chamber; (2) In a constant environment with a temperature of 25±2℃ and a relative humidity of 95%±3%, a water film of different thicknesses is formed on the surface of the test specimens by a controllable spray device, and an AC voltage of power frequency is applied at a voltage increase rate of 1 kV / s until surface flashover occurs, and the inflection point of the surface leakage current surge under each combination of conditions is recorded; (3) The test data are substituted into the mathematical model of the water film conductivity effect evaluation function, and the results are obtained by calibration through calibration. For the parameters to be fitted, the Levenberg-Marquardt least squares algorithm is used to perform multivariate nonlinear surface fitting, and the value corresponding to the minimum fitting residual is selected. The value is used as the hydrophilic attenuation constant of the insulating material, and the goodness of fit R² should not be less than 0.95; (4) the constant The typical reference range is usually between 0.01 and 0.5 μm⁻¹·(mg / cm²)⁻¹. The specific value varies depending on the insulation material formulation (such as epoxy resin, SMC or DMC). It should be recorded in the equipment factory type test report and entered into the power distribution asset management master station.

[0035] This invention couples water film thickness with the pollution and salt density parameters of the real environment to construct a nonlinear penalty exponential function based on high-voltage flashover test fitting. This mechanism overcomes the limitations of traditional linear estimation, accurately calibrating the insulation hydrophilicity attenuation constant using the least squares method, and reconstructing the real physical process of the surge in insulation leakage current under the combined effects of pollution and wetness. This improves the fidelity of dynamic creepage distance degradation calculations and provides data support for the scientific assessment of branch box insulation status and accurate early warning of flashover in complex and harsh environments.

[0036] Further, the phase-to-phase flashover breakdown critical value is obtained through the following steps: extracting historical phase-to-phase flashover fault records of the same type of enclosed distribution branch box from the preset distribution asset management master station database; parsing the historical phase-to-phase flashover fault records, retrieving the time series data of temperature and humidity inside the fault cavity and the temperature drop gradient data of the external environment within the preset tracing time window before the fault occurred; inputting the time series data of temperature and humidity inside the fault cavity and the temperature drop gradient data of the external environment into the condensation surface creepage prediction model to calculate the maximum cumulative water film thickness before the fault; mapping the maximum cumulative water film thickness before the fault to the bus phase-to-phase insulation clearance parameter to calculate the historical limit creepage distance degradation degree; extracting the pre-entered insulation safety margin coefficient; multiplying the historical limit creepage distance degradation degree by the insulation safety margin coefficient to obtain the phase-to-phase flashover breakdown critical value.

[0037] The preset database of the power distribution asset management master station contains historical phase-to-phase flashover fault records of the same model of enclosed distribution branch boxes. Specifically, the system automatically filters out actual phase-to-phase flashover breakdown accident logs and corresponding micro-meteorological logs from batches of the same type of enclosed distribution branch box with the same internal insulation structure and deployed in similar climate zones (such as coastal high-salt-fog areas) within a set period (e.g., 3 to 5 years) that occurred during the past. The preset traceability time window is specifically set to 1 to 2 hours before the historical phase-to-phase flashover fault occurred. By extracting data within this time window for steady-state and transient calculations, the physical vaporization damage of the water film caused by the high heat of the electric arc at the moment of breakdown can be completely avoided. This accurately restores the maximum amount of water vapor that can adhere to the condensed surface just before the insulation failure, i.e., the maximum cumulative water film thickness in the transient state before the fault, ensuring that the acquisition of the underlying data strictly follows the laws of thermodynamics and the physical evolution of the electric arc.

[0038] The specific value range of the pre-entered insulation safety margin coefficient is set to 0.75 to 0.85. This insulation safety margin coefficient is obtained by statistically analyzing the breakdown probability distribution curve obtained from the ultimate withstand voltage test of the same type of insulation structure in an artificial climate chamber, extracting the lower confidence limit quantile corresponding to the extremely low breakdown probability (such as less than 1% or 5%), and comprehensively calibrating it in conjunction with the insulation reliability engineering margin required by the power distribution equipment operation regulations. Finally, it is pre-entered and distributed to the prediction model by the power distribution network operation and maintenance management personnel through the human-computer interaction interface of the power distribution asset management master station.

[0039] This invention provides in-depth analysis of historical flashover fault records for the same type of equipment under real-world weather conditions, deducing the absolute and objective limit of degradation. Simultaneously, it introduces an insulation safety margin coefficient based on a breakdown probability distribution curve and national regulations. This mechanism abandons the subjective experience-based reliance on traditional threshold settings, deeply integrating theoretical degradation calculations with real-world engineering tolerance margins. This enables the scientific setting of phase-to-phase flashover breakdown critical values, improving the accuracy of flashover prevention and early warning, and enhancing the safety of distribution network operation.

[0040] Furthermore, the dehumidification maintenance work order for the power distribution switchgear includes the following structured data fields: a unique asset identification code and spatial location coordinates pointing to the target enclosed power distribution branch box; a record of the dynamic creepage distance degradation value on the surface of the phase-to-phase insulation partition that triggers the alarm; a set preventive power outage maintenance mandatory execution timestamp based on the remaining time for the dynamic creepage distance degradation value to reach the phase-to-phase flashover breakdown critical value; a remote mandatory activation command pointing to the anti-condensation heating device inside the target enclosed power distribution branch box; and a list of cleaning actions for the insulation surface of the wall bushing in the cable entry compartment and a baseline for the quantity of silica gel desiccant to be replaced.

[0041] The process for obtaining the replacement quantity benchmark for silica gel desiccant is as follows: the effective volume parameter of the inner cavity of the target enclosed distribution branch box is extracted from the distribution asset management master station, and multiplied by the water film thickness calculated in the previous step to obtain the estimated total amount of liquid water condensation in the inner cavity. Then, the estimated total amount of liquid water condensation is divided by the pre-entered limiting hygroscopic saturation mass constant of a single pack of silica gel desiccant, and rounded up to obtain the replacement quantity benchmark for silica gel desiccant. The process of obtaining the list of cleaning actions for the insulating surface is as follows: extract the dynamic creepage distance degradation value of the phase-to-phase insulating partition surface when the alarm is triggered, compare it with the preset insulation degradation grading strategy table in the power distribution asset management master station database, and match and retrieve the standardized operation and maintenance operation code corresponding to the current degradation value range. The standard execution code specifically points to specific operation steps such as live cleaning process, power outage insulation wiping or anti-flashover coating respray.

[0042] This invention transforms traditional, extensive dispatching into structured, precise execution instructions. By scientifically quantifying condensate volume by combining internal cavity volume and water film thickness, it accurately outputs desiccant replacement benchmarks, eliminating waste or insufficient dehumidification caused by blindly replacing desiccant based on experience. Simultaneously, it automatically matches standardized insulation cleaning and coating processes based on real-time degradation levels, eliminating subjective differences in manual grading. This mechanism achieves quantitative closed-loop management from status warnings to material allocation and on-site maintenance actions, improving equipment operation and maintenance efficiency and economy.

[0043] This invention overcomes the blindness of traditional fixed-period inspections by accurately quantifying the dynamic water film thickness at wall bushings using a thermodynamic phase change model and mapping it to the inter-phase insulation clearance of busbars, scientifically deriving the dynamic creepage distance degradation degree. This method transforms the hidden microclimate condensation phenomenon into a quantified electrical flashover prevention early warning indicator, achieving precise closed-loop scheduling of dehumidification maintenance and preventative power outages. This effectively eliminates the operation of faulty equipment and the waste of maintenance resources, significantly improving the safety and economy of power distribution network operation and maintenance.

[0044] Example 2 Taking the operation and maintenance scenario of a 10kV target enclosed distribution branch box deployed in a coastal high salt fog industrial park when it encounters a sudden attack of strong cold air (such as an autumn cold wave) as an example, the workflow of the present invention will be specifically explained.

[0045] A temperature and humidity sensor unit installed inside the branch box continuously collects the original internal temperature and humidity sequence at a first preset sampling frequency of 1Hz. An outlier detection method based on the 3σ criterion is used to remove electromagnetic interference data caused by surrounding industrial equipment. Then, Lagrange interpolation is used to fill in gaps and align the timestamps. Simultaneously, an external micro-meteorological sensor terminal collects the external ambient temperature at a frequency of 0.1Hz. Calculations show that the temperature change rate within the current adjacent sampling period reaches -0.4℃ / min, successfully triggering a preset temperature drop threshold of -0.2℃ / min to -0.5℃ / min. The system automatically extracts this value as the external ambient temperature drop gradient data.

[0046] After receiving the aforementioned time-series data, the cavity thermodynamic evaluation unit, combining the molar mass of water vapor and the latent heat of vaporization of water at a specific temperature, accurately calculates the pressure difference between the current saturated vapor pressure and the absolute vapor pressure using the Magnus-Tertens empirical formula. Next, the interface phase change calculation unit extracts the preset 0.015 m / s surface mass transfer coefficient of the through-wall sleeve and multiplies it by the pressure difference to obtain the water droplet condensation rate. Within a set 30-minute continuous monitoring time window, the surface water film integration unit performs time integration on this rate, generating the current dynamic water film thickness value at the condensation interface.

[0047] The system remotely analyzes the 3D CAD digital twin ledger of this type of equipment in the power distribution asset management master station, extracts the initial physical creepage distance value of the phase-to-phase insulation baffle and the minimum air gap value between two adjacent phase busbars, and uses matrix splicing operations to construct a two-dimensional spatial insulation feature vector, namely the busbar phase-to-phase insulation clearance parameter. Subsequently, the system extracts the feature ratio of the initial creepage distance to the minimum air gap in this feature vector and defines it as the spatial electric field coupling enhancement factor. The system retrieves the pre-recorded surface salt density parameter of the high pollution level in the coastal area (e.g., 0.25 mg / cm²), inputs the current water film thickness and the spatial electric field coupling enhancement factor into the preset model, and calculates the water film conductivity penalty coefficient using a nonlinear penalty exponential function that incorporates the spatial electric field distortion effect. Multiplying the initial creepage distance by the penalty coefficient yields the water film surface bridging distance value, and subtracting the latter to calculate the current effective creepage distance. Finally, the dynamic creepage distance degradation of the current phase-to-phase insulation baffle surface is calculated to have climbed to 78%.

[0048] The system automatically filters historical phase-to-phase flashover fault records of the same model equipment in the main station database over the past three years under similar coastal high-salt-fog climates. It extracts micro-meteorological time-series data from a preset tracing time window (1-2 hours before the fault) for extrapolation, determining the maximum transient cumulative water film thickness and historical extreme degradation before the fault. Combined with an insulation safety margin coefficient of 0.8, it scientifically extrapolates the current phase-to-phase flashover breakdown critical value to be 75%. Since the current degradation (78%) has exceeded this critical value, the system immediately responds, generating a structured dehumidification maintenance work order containing the branch box's unique asset identification code and spatial location coordinates. This work order not only issues a remote forced activation command for the anti-condensation heating device but also accurately generates a baseline for replacing three packs of silica gel desiccant based on the effective internal volume and the calculated estimated total liquid water condensation. Simultaneously, based on the degradation value and matching the insulation degradation grading strategy table, the work order includes a detailed operation list for power outage insulation wiping and anti-flashover coating re-spraying. Ultimately, the work order was automatically sent to the power distribution asset management master station, successfully establishing a precise preventative power outage maintenance schedule before an actual phase-to-phase flashover failure occurred.

[0049] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the maintenance cycle of branch boxes based on historical data, characterized in that, include: Acquire time-series data of internal temperature and humidity of the target enclosed power distribution branch box and data of the gradient of sudden temperature drop in the external environment; The internal temperature and humidity time series data and the external ambient temperature drop gradient data are input into the preset condensation creepage prediction model. Based on the thermodynamic phase change parameters of the enclosed space, the water droplet condensation rate and water film thickness at the cable inlet bushing of the target enclosed distribution branch box are calculated. The water droplet condensation rate and the water film thickness are mapped to the bus phase-to-phase insulation clearance parameters of the target enclosed distribution branch box, and the dynamic creepage distance degradation degree of the phase-to-phase insulation baffle surface is calculated. Compare the dynamic creepage distance degradation with the preset phase-to-phase flashover breakdown threshold; When the dynamic creepage distance degradation reaches the phase-to-phase flashover breakdown critical value, a dehumidification maintenance work order for the power distribution switchgear is generated. Send the dehumidification maintenance work order of the power distribution switchgear to the power distribution asset management master station to establish a preventive power outage maintenance schedule for the target enclosed power distribution branch box.

2. The branch box maintenance cycle prediction method based on historical data according to claim 1, characterized in that, The steps for obtaining time-series data of internal temperature and humidity and gradient data of sudden temperature drop in the external environment include: A temperature and humidity sensor unit installed inside the target enclosed power distribution branch box collects the original internal temperature and humidity sequence at a first preset sampling frequency; a preset outlier removal algorithm is used to perform data cleaning and timestamp alignment on the original internal temperature and humidity sequence to generate the internal temperature and humidity time series data; a micro-meteorological sensor terminal deployed outside the target enclosed power distribution branch box collects the external ambient temperature sequence at a second preset sampling frequency; the time derivative of the external ambient temperature sequence within adjacent sampling periods is calculated to obtain the temperature change rate; it is determined whether the temperature change rate reaches a preset temperature drop threshold; in response to the temperature change rate reaching the temperature drop threshold, the temperature change rate reaching the preset temperature drop threshold is extracted as the external ambient temperature drop gradient data.

3. The branch box maintenance cycle prediction method based on historical data according to claim 1, characterized in that, The condensation creepage prediction model includes a cavity thermodynamic evaluation unit, an interface phase change calculation unit, and a surface water film integration unit. The cavity thermodynamic evaluation unit receives the internal cavity temperature and humidity time series data and the external ambient temperature drop gradient data. Based on the thermodynamic phase change parameters of the sealed space, it extracts the saturated vapor pressure parameter and the current absolute vapor pressure parameter of the condensation surface of the target closed distribution branch box cavity, and calculates the pressure difference between the saturated vapor pressure parameter and the current absolute vapor pressure parameter. The interface phase change calculation unit extracts the preset surface mass transfer coefficient of the cable inlet chamber wall bushing, multiplies the surface mass transfer coefficient by the pressure difference value output by the cavity thermodynamic evaluation unit, and calculates the water droplet condensation rate at the cable inlet chamber wall bushing. The surface water film integration unit is used to perform time integration calculation on the water droplet condensation rate output by the interface phase change calculation unit within a preset continuous monitoring time window to generate the water film thickness.

4. The branch box maintenance cycle prediction method based on historical data according to claim 1, characterized in that, The busbar phase-to-phase insulation clearance parameter is obtained by analyzing the structural dimension database of the target enclosed distribution branch box, extracting the initial physical creepage distance value of the phase-to-phase insulation baffle and the minimum air gap value between two adjacent phase busbars; and combining the initial physical creepage distance value and the minimum air gap value to generate the busbar phase-to-phase insulation clearance parameter.

5. The branch box maintenance cycle prediction method based on historical data according to claim 4, characterized in that, Calculate the dynamic creepage distance degradation degree on the surface of the phase-to-phase insulation baffle: Analyze the two-dimensional spatial insulation feature vector corresponding to the bus phase-to-phase insulation clearance parameter, and extract the physical initial creepage distance value and the minimum air gap value; The characteristic ratio of the initial physical creepage distance to the minimum air gap is calculated and defined as the spatial electric field coupling enhancement factor. The water film thickness and the spatial electric field coupling enhancement factor are input into a preset water film conductivity effect evaluation function. The pre-recorded surface contamination salt density parameter is retrieved, and the water film conductivity penalty coefficient on the surface of the interphase insulating baffle is calculated. The initial physical creepage distance value is multiplied by the water film conductivity penalty coefficient to calculate the water film surface bridging distance value. The initial physical creepage distance value is subtracted from the water film surface bridging distance value to calculate the current effective creepage distance value. Calculate the difference between the initial physical creepage distance value and the current effective creepage distance value, and divide the difference by the initial physical creepage distance value to obtain the dynamic creepage distance degradation degree.

6. The method for predicting the maintenance cycle of a branch box based on historical data according to claim 1, characterized in that, The phase-to-phase flashover breakdown threshold is obtained through the following steps: extracting historical phase-to-phase flashover fault records of the same type of enclosed distribution branch box from the preset distribution asset management master station database; parsing the historical phase-to-phase flashover fault records, and retrieving the time series data of temperature and humidity inside the fault cavity and the gradient data of sudden temperature drop in the external environment of the fault within the preset tracing time window before the fault occurred. The time series data of temperature and humidity inside the fault cavity and the gradient data of sudden temperature drop in the external environment are input into the condensation creepage prediction model to calculate the maximum cumulative water film thickness before the fault. The maximum cumulative water film thickness before the fault is mapped to the interphase insulation clearance parameter of the busbar to calculate the historical limit creepage distance degradation. The pre-entered insulation safety margin coefficient is extracted. The historical limit creepage distance degradation is multiplied by the insulation safety margin coefficient to obtain the interphase flashover breakdown critical value.

7. The method for predicting the maintenance cycle of a branch box based on historical data according to claim 1, characterized in that, The dehumidification maintenance work order for the power distribution switchgear includes the following structured data fields: a unique asset identification code and spatial positioning coordinates pointing to the target enclosed power distribution branch box; a record of the dynamic creepage distance degradation value on the surface of the phase-to-phase insulation partition that triggers the alarm; and a set preventive power outage maintenance mandatory execution timestamp based on the remaining time for the dynamic creepage distance degradation value to reach the phase-to-phase flashover breakdown critical value. The remote forced activation command is directed to the anti-condensation heating device inside the target enclosed power distribution branch box; the cleaning action list and silica gel desiccant replacement quantity reference are directed to the insulation surface of the through-wall bushing in the cable entry chamber.

Images