A power distribution equipment defect diagnosis method and system based on multi-physical parameter analysis

Abstract

The application provides a power distribution equipment defect diagnosis method and system based on multi-physical parameter analysis, and relates to the technical field of power distribution equipment management. The method comprises the following steps: deploying a multi-parameter synchronous sensor group, acquiring original time sequence data of the power distribution equipment, and constructing an internal and external multi-dimensional original time sequence database; performing data cleaning, key derived feature calculation, heat balance correction and dry force correction on the internal and external multi-dimensional original time sequence database, and acquiring a standardized feature database; constructing an equipment dynamic response fingerprint feature vector; constructing a wall surface phase change and fluid feature vector; establishing a dynamic health baseline library, combining the equipment dynamic response fingerprint feature vector and the wall surface phase change and fluid feature vector, and performing three-level fusion diagnosis. The technical problems of lacking early quantitative diagnosis means, serious environmental factor interference, high false alarm rate, single diagnosis dimension and lacking multi-physical quantity collaborative fusion joint diagnosis mechanism in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of power distribution equipment management technology, and in particular to a method and system for diagnosing power distribution equipment defects based on multi-physical parameter analysis. Background Technology

[0002] Power distribution equipment is the most numerous and widely distributed basic equipment in a power system, and its safe and stable operation is directly related to the reliability of power supply and the quality of electricity consumption on the user side. With the continuous advancement of my country's new power system construction, the number of power distribution equipment is growing rapidly, the operating environment is becoming increasingly complex, and the management of equipment throughout its entire life cycle faces severe challenges. Defects such as internal insulation moisture absorption, condensation, and localized abnormal heating due to the degradation of the sealing performance of power distribution equipment are one of the main causes of insulation breakdown, flashover accidents, and even power outages.

[0003] Currently, the detection and diagnosis of insulation and sealing conditions of power distribution equipment mainly rely on insulation resistance testing, dielectric loss factor measurement, partial discharge detection, and infrared thermal imaging. However, these methods all have significant limitations: changes in insulation resistance and dielectric loss factor have a significant lag; electrical parameters only show measurable shifts when the moisture content inside the insulation material accumulates to a considerable level, making it impossible to provide quantitative early warning of early sealing performance degradation. Partial discharge detection can only capture discharge defects and is insensitive to non-discharge-related slow moisture absorption and sealing aging. Infrared thermal imaging is an offline sampling method and is difficult to achieve online continuous monitoring. More importantly, existing online monitoring technologies generally use a fixed threshold method to judge whether the monitored parameters exceed the standard, treating the equipment as an independent system isolated from the external environment. This completely ignores the continuous impact of external environmental factors (temperature, humidity, wind speed, solar radiation, etc.) on the internal condition parameters of the equipment, leading to frequent false alarms during the rainy season, drastic diurnal temperature variations, or windy weather. Furthermore, existing technologies only focus on static snapshot information such as whether condensation occurs, without analyzing the dynamic evaporation process after condensation forms, and cannot distinguish between environment-driven condensation and defect-driven condensation.

[0004] In summary, existing defect diagnosis technologies for power distribution equipment mainly suffer from the following technical problems: First, there is a lack of early quantitative diagnostic methods for sealing performance degradation and non-discharge-related moisture absorption, making it impossible to provide effective early warnings during the sub-healthy stage of declining sealing performance; second, environmental factors cause significant interference, and the fixed threshold method struggles to distinguish between normal changes caused by environmental fluctuations and defects in the equipment itself, resulting in a high false alarm rate; third, the diagnostic dimensions are limited, lacking a joint diagnostic mechanism that integrates multiple physical quantities. Therefore, a method is urgently needed to address these issues. Summary of the Invention

[0005] This disclosure provides a method and system for diagnosing defects in power distribution equipment based on multi-physical parameter analysis. It addresses the technical problems in the prior art, such as the lack of quantitative diagnostic methods for early stages of sealing performance degradation and non-discharge moisture absorption, the inability to provide effective early warnings in the sub-healthy stage of declining sealing performance, severe interference from environmental factors, difficulty in distinguishing between normal changes caused by environmental fluctuations and defects in the equipment itself using fixed threshold methods, high false alarm rate, single diagnostic dimensions, and lack of a collaborative diagnostic mechanism integrating multiple physical quantities.

[0006] According to a first aspect of this disclosure, a method for diagnosing defects in power distribution equipment based on multi-physical parameter analysis is provided, comprising the following steps:

[0007] By deploying multi-parameter synchronous sensor groups using an internal-external differential distribution point method, the original time series data of the power distribution equipment is obtained, and an internal and external multi-dimensional original time series database is constructed. The multi-parameter synchronous sensor groups include an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group.

[0008] Data cleaning, key derived feature calculation, thermal balance correction and drying force correction are performed on the internal and external multidimensional original time series databases to obtain a standardized feature database;

[0009] Based on a standardized feature database, dynamic response feature parameters of the power distribution equipment casing are extracted, and a dynamic response fingerprint feature vector of the equipment is constructed. The dynamic response feature parameters of the casing include thermal response time constant, wet response time constant, and damping ratio.

[0010] Based on a standardized feature database, condensation and evaporation relaxation analysis and thermal plume extraction are performed to construct wall phase change and fluid feature vectors.

[0011] A dynamic health baseline database is established. By combining the device's dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, a three-level fusion diagnosis is performed, and the diagnosis results are output. The diagnosis results include false alarm suppression, sealing warning, moisture warning, and overheating location.

[0012] According to a second aspect of this disclosure, a power distribution equipment defect diagnosis system based on multi-physical parameter analysis is provided, comprising:

[0013] The multi-parameter synchronous acquisition module is used to deploy a multi-parameter synchronous sensor group through an internal and external differential distribution point method to acquire the original time series data of the power distribution equipment and construct an internal and external multi-dimensional original time series database. The multi-parameter synchronous sensor group includes an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group.

[0014] The thermal and humidity environment correction module is used to perform data cleaning, key derived feature calculation, heat balance correction and drying force correction on the internal and external multidimensional original time series databases to obtain a standardized feature database.

[0015] The device dynamic response fingerprint feature construction module is used to extract dynamic response feature parameters of the power distribution equipment shell based on a standardized feature database, and construct a device dynamic response fingerprint feature vector. The shell dynamic response feature parameters include thermal response time constant, wet response time constant and damping ratio.

[0016] A wall phase change and fluid feature construction module is used to perform condensation and evaporation relaxation analysis and thermal plume extraction based on a standardized feature database, and to construct a wall phase change and fluid feature vector.

[0017] The fusion diagnostic module is used to establish a dynamic health baseline library, combine the device's dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, perform three-level fusion diagnostics, and output diagnostic results, including false alarm suppression, sealing warning, moisture warning, and overheating location.

[0018] One or more technical solutions provided in this disclosure have at least the following technical effects or advantages: A multi-parameter synchronous sensor group is deployed using an internal / external differential distribution point method to acquire the original time-series data of the power distribution equipment, constructing an internal / external multi-dimensional original time-series database. The multi-parameter synchronous sensor group includes an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group. Data cleaning, key derived feature calculation, thermal balance correction, and drying force correction are performed on the internal / external multi-dimensional original time-series database to obtain a standardized feature database. Based on the standardized feature database, dynamic response feature parameters of the power distribution equipment casing are extracted to construct a dynamic response fingerprint feature vector. The casing dynamic response feature parameters include thermal response time constant, wet response time constant, and damping ratio. Based on the standardized feature database, condensation and evaporation relaxation analysis and thermal plume extraction are performed to construct a wall phase change and fluid feature vector. A dynamic health baseline library is established, and a three-level fusion diagnosis is performed by combining the equipment dynamic response fingerprint feature vector and the wall phase change and fluid feature vector, outputting diagnostic results. The diagnostic results include false alarm suppression, sealing warning, moisture warning, and overheating location. This invention addresses the technical problems in existing technologies, including the lack of quantitative diagnostic methods for early stages of sealing performance degradation and non-discharge-induced moisture absorption, the inability to provide effective early warnings during the sub-healthy stage of declining sealing performance, severe interference from environmental factors, the difficulty of distinguishing normal changes caused by environmental fluctuations from equipment defects using fixed threshold methods, high false alarm rates, limited diagnostic dimensions, and the lack of a multi-physical quantity collaborative diagnostic mechanism. It achieves a significant improvement in the reliability and operational efficiency of power distribution equipment condition monitoring.

[0019] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0021] Figure 1 A flowchart illustrating a method for diagnosing defects in power distribution equipment based on multi-physical parameter analysis, provided in an embodiment of this application;

[0022] Figure 2 This is a schematic diagram of the structure of a power distribution equipment defect diagnosis system based on multi-physical parameter analysis, provided in an embodiment of this application.

[0023] Figure labeling: 11 Multi-parameter synchronous acquisition module, 12 Thermal and humid environment correction module, 13 Equipment dynamic response fingerprint feature construction module, 14 Wall phase change and fluid feature construction module, 15 Fusion diagnosis module. Detailed Implementation

[0024] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0025] Example 1: This disclosure provides a method for diagnosing defects in power distribution equipment based on multi-physical parameter analysis, which is referred to below. Figure 1 The methods include:

[0026] S1: By deploying multi-parameter synchronous sensor groups through the internal and external differential distribution point method, the original time series data of the power distribution equipment is obtained, and an internal and external multi-dimensional original time series database is constructed. The multi-parameter synchronous sensor groups include an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group.

[0027] Furthermore, step S1 also includes:

[0028] Basic meteorological parameters are collected by an external environment sensor group, which includes an ambient temperature sensor and an ambient relative humidity sensor. The basic meteorological parameters include the ambient temperature and the ambient relative humidity.

[0029] Environmental correction parameters are obtained by an external correction sensor group, which includes a wind speed sensor and a solar irradiance sensor. The environmental correction parameters include wind speed and solar irradiance.

[0030] Internal air condition information and shell temperature response information are collected by an internal response sensor group. The internal response sensor group includes an internal air temperature sensor, an internal relative humidity sensor, and three wall temperature sensors. The internal air condition information includes internal air temperature and internal relative humidity, and the shell temperature response information is the wall temperature.

[0031] Information on local air disturbances within the interior is collected by an internal disturbance sensor group, which includes a polymer thin-film capacitive condensation sensor and a near-wall micro-airflow monitoring unit. The information on local air disturbances within the interior is a near-wall micro-airflow velocity sequence.

[0032] By integrating external ambient temperature, external relative humidity, wind speed, solar irradiance, internal air temperature, internal relative humidity, wall temperature, and near-wall micro-airflow velocity sequences, a multi-dimensional original time-series database is constructed.

[0033] Specifically, this step aims to deploy a multi-parameter synchronous sensor group using an internal-external differential distribution point method to simultaneously acquire internal and external information of the power distribution equipment, providing a unified data source for subsequent extraction of shell dynamic response characteristic parameters, condensation and evaporation relaxation analysis, and thermal plume extraction.

[0034] The multi-parameter synchronous sensor group is divided into four categories according to the monitored object: external environment sensor group, external correction sensor group, internal response sensor group, and internal disturbance sensor group. Each type of sensor group undertakes the task of collecting different physical quantities and together constitutes the data acquisition system.

[0035] The external environment sensor group is used to collect basic meteorological parameters of the external environment where the power distribution equipment is located. The group includes an ambient temperature sensor and an ambient relative humidity sensor. The ambient temperature sensor is preferably a Pt100 industrial-grade platinum resistance temperature sensor, installed in an unobstructed location near the power distribution equipment casing and fixed within a protective louvered box approximately 0.5 to 1.0 m outside the equipment using a mounting bracket to avoid the influence of direct sunlight and equipment heat dissipation on the measurement results. The ambient relative humidity sensor is preferably a polymer thin-film capacitive humidity sensor. This type of sensor has advantages such as fast response speed, good long-term stability, and strong anti-condensation ability, making it suitable for long-term outdoor monitoring of power distribution equipment. The ambient relative humidity sensor and the ambient temperature sensor are integrated and installed in the same protective louvered box to ensure that their measurement environments are consistent. This sensor group is used to continuously acquire the external ambient temperature T. out and external relative humidity (RH) out It characterizes the external environmental excitation conditions that power distribution equipment experiences during operation.

[0036] An external correction sensor group is used to acquire environmental correction parameters affecting the heat exchange process of the power distribution equipment. The group includes a wind speed sensor and a solar irradiance sensor. The wind speed sensor is preferably an ultrasonic wind speed sensor or a hot-film wind speed sensor, installed on top of or around the power distribution equipment in an unobstructed area. Its measurement direction is consistent with the prevailing wind direction to collect real-time data on the airflow around the equipment and obtain the wind speed V. wind The solar irradiance sensor is preferably a silicon photovoltaic total irradiance sensor, installed horizontally on the top of the equipment in an unobstructed position. It is used to continuously measure the solar irradiance intensity received by the equipment, obtaining the solar irradiance (Solar). The data acquired by this sensor group is used for thermal balance correction and drying force correction in subsequent steps.

[0037] The internal response sensor group is used to collect information on the internal air condition and enclosure temperature response of the power distribution equipment. The group includes an internal air temperature sensor, an internal relative humidity sensor, and three wall temperature sensors. The internal air temperature sensor and the internal relative humidity sensor are installed in the air circulation area inside the distribution cabinet, avoiding proximity to heat-generating components or vents, to ensure that the measured data reflects the overall air condition inside the cabinet. The internal air temperature T is continuously acquired. in and internal relative humidity RH in The wall temperature sensors are preferably thin-film platinum resistance temperature sensors or thermocouples, and are installed in three key typical areas of the power distribution equipment housing that are prone to condensation or large temperature changes, including the top, side walls, and bottom, to obtain three corresponding wall temperature T values. S1 T S2 T S3 Multiple wall temperature measurement points can reflect the differences in thermal response at different locations of the shell, providing basic data for subsequent calculations of wall undercooling.

[0038] The internal disturbance sensor group is used to collect information on local air disturbances inside the power distribution equipment. The group includes a polymer film capacitive condensation sensor and a near-wall micro-airflow monitoring unit. The polymer film capacitive condensation sensor is installed in the area of ​​the bottom or lower side wall where the temperature is lowest and condensation is most likely to occur. Its sensitive layer is installed close to the inner wall of the casing to detect changes in the humidity near the wall in real time. The near-wall micro-airflow monitoring unit preferably uses a miniature hot-wire anemometer, distributed along different locations on the inner wall of the casing. Typically, 1 to 2 measuring points are arranged along each of the four walls (front, back, left, and right) inside the equipment, for a total of 4 to 8 measuring points. Each measuring point is maintained at a distance of approximately 5 to 10 mm from the wall to avoid the wall boundary layer effect affecting measurement accuracy, while simultaneously acquiring the airflow status near the wall. Multiple measuring points continuously collect the near-wall micro-airflow velocity V at their respective locations. flow1 V flow2 ... V flown This generates a near-wall micro-airflow velocity sequence, which reflects the characteristics of air disturbance and thermal plume distribution inside the equipment, providing basic data for subsequent condensation and evaporation relaxation analysis and thermal plume location.

[0039] After installation, each sensor group is connected to a unified data acquisition system terminal. The data acquisition system terminal synchronously collects data according to a preset sampling period. In this embodiment, the sampling period is preferably set to 5 seconds, which balances the accuracy of dynamic response capture with the economy of data storage and transmission. The IEEE 1588 precision time protocol is used to achieve unified time synchronization between each sampling node, so that all collected data have a unified timestamp. The data acquisition system terminal synchronously records the external ambient temperature, external relative humidity, wind speed, solar irradiance, internal air temperature, internal relative humidity, wall temperature, and near-wall micro-airflow velocity sequences based on the unified time reference, and stores them continuously according to the sampling period. Finally, a multi-dimensional raw time series database containing all original time series data with strictly aligned timestamps is constructed.

[0040] S2: Perform data cleaning, key derived feature calculation, thermal balance correction and drying force correction on the internal and external multidimensional original time series databases to obtain a standardized feature database;

[0041] Furthermore, step S2 also includes:

[0042] Obtain internal and external multidimensional raw time series databases, and perform data cleaning on all raw time series data in the database. The data cleaning includes integrity checks and smoothing processes.

[0043] The internal air dew point temperature is calculated based on the internal air temperature and relative humidity after data cleaning. The specific calculation formula is as follows:

[0044] ;

[0045] Among them, T d T is the internal air dew point temperature. in RH (Relative Tolerance) is the internal air temperature. in Let be the internal relative humidity, and a and b be empirical coefficients of the Magnus formula, where a is a dimensionless coefficient and b is a temperature-dimensional coefficient, a = 17.62, b = 243.12℃. As an intermediate variable;

[0046] The wall subcooling at the corresponding location is calculated based on the wall temperature and the internal air dew point temperature. The specific calculation formula is as follows:

[0047] ;

[0048] in, T represents the wall subcooling. s T represents the corresponding wall temperature. d This refers to the dew point temperature of the internal air.

[0049] A convective heat transfer coefficient is set, and a heat balance correction is performed based on the convective heat transfer coefficient to obtain the equivalent external temperature. The specific calculation formula is as follows:

[0050] ;

[0051] Among them, T eq For equivalent external temperature, T out The external ambient temperature is represented by Solar, and the solar irradiance is represented by h. conv The convective heat transfer coefficient is... The heat absorption coefficient of the shell surface;

[0052] A dryness index is constructed, and dryness is corrected by comprehensively considering wind speed, external ambient temperature, and external relative humidity. The specific calculation formula is as follows:

[0053] ;

[0054] Where DI is the drying power index, and V wind For wind speed, RH out For external relative humidity, T out External ambient temperature;

[0055] The raw time-series data after data cleaning, along with the calculated internal air dew point temperature, wall subcooling, equivalent external temperature, and drying index, constitute a standardized feature database.

[0056] Furthermore, step S2 also includes:

[0057] The average wall temperature is calculated based on the wall surface temperature. The specific calculation formula is as follows:

[0058] ;

[0059] in, The average wall temperature, The wall temperatures at three measuring points on the wall are shown respectively.

[0060] The wall temperature difference is calculated based on the average wall temperature and the ambient temperature. The specific calculation formula is as follows:

[0061] ;

[0062] in, Due to the temperature difference in the wall environment, The average wall temperature, External ambient temperature;

[0063] The characteristic length L is determined based on the geometric parameters of the power distribution equipment housing;

[0064] The qualitative temperature is calculated based on the average wall temperature and the external ambient temperature. The specific calculation formula is as follows:

[0065] ;

[0066] in, For qualitative temperature, The average wall temperature, The external ambient temperature is represented by the constant 273.15, which is the conversion constant between the Celsius scale and the thermodynamic temperature scale.

[0067] The thermal conductivity and kinematic viscosity of air at the qualitative temperature are calculated based on the air's thermophysical properties. The specific calculation formulas are as follows:

[0068] ;

[0069] in, The thermal conductivity of air, The constant 293.15 is the Kelvin temperature corresponding to 20℃, the constant 0.0242 is the fitting intercept of the thermal conductivity of air, and the constant 0.000072 is the temperature sensitivity coefficient of the thermal conductivity of air.

[0070] ;

[0071] in, For air kinematic viscosity, This is the qualitative temperature, with a constant of 1.48 × 10⁻⁶. -5 The constant 293.15 is the reference value for the kinematic viscosity of air at 20℃, the constant 110.4 is the Kelvin temperature corresponding to 20℃, and the constant 293.15 is the Sutherland constant for air.

[0072] The Reynolds number is calculated based on wind speed and characteristic length, using the following formula:

[0073] ;

[0074] in, Let V be the Reynolds number. wind For wind speed, Where L is the viscosity of air motion, and L is the characteristic length of the power distribution equipment;

[0075] Determine the airflow state on the shell surface based on the Reynolds number. When Re < 5 × 10⁻⁶ 5 The flow is determined to be laminar. The proportional gain of the Nusselt number correlation is set to 0.664, and the exponential term of the Reynolds number is set to 0.5. When Re ≥ 5 × 10⁻⁶, the flow is considered laminar. 5 The flow is determined to be in a turbulent state. The proportionality coefficient of the Nusselt number correlation is set to 0.037, and the exponential term of the Reynolds number is set to 0.8.

[0076] The formula for calculating the forced convection heat transfer coefficient component is as follows:

[0077] ;

[0078] in, For the forced convection heat transfer coefficient component, C is the proportionality coefficient of the Nusselt number correlation, Re is the Reynolds number, m is the exponential term of the Reynolds number, Pr is the Prandtl number, λ is the thermal conductivity of air, and L is the characteristic length;

[0079] Calculate the natural convection heat transfer coefficient components:

[0080] ;

[0081] in, This represents the natural convection heat transfer coefficient component. The temperature difference in the wall environment is denoted by L, the characteristic length is L, and the constant 1.42 is an empirical coefficient obtained by fitting experimental data of natural convection on a vertical wall.

[0082] The convective heat transfer coefficient is obtained by setting the forced convection heat transfer coefficient component and the natural convection heat transfer coefficient component. The specific calculation formula is as follows:

[0083] ;

[0084] in, The convective heat transfer coefficient is... For the forced convection heat transfer coefficient component, This represents the natural convection heat transfer coefficient component.

[0085] Specifically, this step aims to perform data cleaning, key derived feature calculation, thermal balance correction, and drying force correction on the internal and external multidimensional original time-series database obtained in step S1. This step does not directly perform defect diagnosis, but rather constructs standardized input data with clear physical meaning while ensuring the authenticity of the original data.

[0086] First, the original multi-dimensional time-series database output from step S1 is cleaned. Due to the potential for abnormal sampling points caused by factors such as communication interruptions, short-term sensor failures, or electromagnetic interference during long-term online monitoring, all original time-series data are first checked for integrity based on a unified timestamp. When a sampling point exhibits a significant abrupt change compared to its preceding and following consecutive sampling values, and does not conform to the thermal inertia variation pattern of the equipment, this sampling point is identified as an outlier. Linear interpolation of adjacent valid sampling values ​​is then used for correction to eliminate the impact of outliers on subsequent calculations. After outlier processing, each original time-series data is further smoothed using a median filter with a window length of 5 to suppress high-frequency random noise generated during on-site monitoring, while preserving the edge characteristics of changes in data such as temperature, humidity, and micro-airflow, thus retaining the true response process for subsequent dynamic parameter identification.

[0087] After data cleaning, key derived characteristics of the internal air conditions are calculated. The internal air dew point temperature is calculated based on the cleaned internal air temperature and relative humidity. The internal air dew point temperature represents the critical temperature at which water vapor begins to condense under the current air conditions, and is an important physical quantity for assessing the risk of condensation inside the equipment. Specifically, the Magnus formula is used for calculation, and the formula is as follows:

[0088] ;

[0089] Among them, T d T is the internal air dew point temperature. in RH (Relative Tolerance) is the internal air temperature. in Let represent the internal relative humidity. 'a' and 'b' are empirical coefficients of the Magnus formula, where 'a' is a dimensionless coefficient and 'b' is a temperature-dimensional coefficient. a = 17.62, b = 243.12℃. These empirical coefficients are the standard empirical coefficients of the Magnus formula recommended by the World Meteorological Organization, applicable to the calculation of saturated vapor pressure and dew point temperature at the water surface within a temperature range of -45℃ to +60℃. Within this temperature range, the calculation uncertainty of this formula is better than ±0.6%. As an intermediate variable, its physical meaning is the logarithmic scaling relationship between the actual water vapor pressure and the saturated water vapor pressure corresponding to the dew point temperature under the current temperature conditions. Its calculation expression is:

[0090] ;

[0091] In this formula, ln is the natural logarithm operator with the natural constant as the base. Its purpose is to convert the water vapor pressure ratio under the logarithmic scale into a linearly additive term and calculate it in a unified manner with the temperature term. The denominator 100 converts the internal relative humidity input in percentage form into a small value between 0 and 1, so that the logarithmic operation is physically meaningful.

[0092] Furthermore, the wall subcooling at the corresponding location is calculated based on the three wall temperature sensors and the internal air dew point temperature. The specific calculation formula is as follows:

[0093] ;

[0094] in, T represents the wall subcooling. s The corresponding wall temperatures are T. S1 T S2 T S3 T d This refers to the internal air dew point temperature. Wall subcooling reflects the temperature difference between the shell wall and the internal air dew point temperature. When... When the temperature is less than 0, it indicates that the wall temperature is lower than the internal air dew point temperature, and the wall surface meets the thermodynamic conditions for condensation to occur; when If the value is greater than 0, it indicates that the wall surface has not yet reached the condensation condition.

[0095] To eliminate the impact of external environmental changes on the thermal response analysis of the equipment, this step further introduces a thermal balance correction, performing an equivalent correction calculation for the external ambient temperature. Since the power distribution equipment is exposed to the outdoors for extended periods, its actual thermal state is affected not only by the external air temperature but also by the combined effects of solar radiation heating and enhanced convection cooling from wind speed. Therefore, directly using the ambient temperature cannot accurately reflect the true external thermal boundary conditions of the equipment; it is necessary to calculate the equivalent external temperature based on the thermal balance relationship of the casing. The specific calculation formula is as follows:

[0096] ;

[0097] Among them, T eq For equivalent external temperature, T out The external ambient temperature is represented by Solar, and the solar irradiance is represented by h. conv The convective heat transfer coefficient is... The heat absorption coefficient of the casing surface is related to the color and material of the casing surface. According to the "Estimation Method of Thermal Characteristics of Housing" and the provisions of IEC60890 standard on the solar radiation absorption coefficient of the casing surface, the solar radiation absorption coefficient of the casing surface with light color coating is about 0.4 to 0.6, and that of the casing surface with dark color coating is about 0.7 to 0.9. The casing of power distribution equipment is usually painted with light gray or light green. Therefore, in this embodiment, α is preferably taken as 0.60. For dark-colored equipment, it can be appropriately adjusted to 0.80.

[0098] It should be noted that the convective heat transfer coefficient in this invention is not a simple exponential function based on a single wind speed, but a comprehensive model that takes into account multiple factors such as forced convection and natural convection heat transfer mechanisms, equipment geometric parameters, and wall temperature. The specific method for obtaining the coefficient is as follows:

[0099] First, the average wall temperature is calculated based on the three wall temperatures collected in step S1. The specific calculation formula is as follows:

[0100] ;

[0101] in, The average wall temperature, These represent the wall temperatures at three measurement points on the wall. The purpose of using the average wall temperature from these three measurement points is to eliminate temperature distribution differences in different parts of the shell caused by uneven sunlight or local heat sources, making subsequent calculations more representative.

[0102] The wall temperature difference is calculated based on the average wall temperature and the ambient temperature. The specific calculation formula is as follows:

[0103] ;

[0104] in, Due to the temperature difference in the wall environment, The average wall temperature, The external ambient temperature. The temperature difference between the wall and the surrounding environment is the fundamental physical factor driving natural convection. When the wall temperature is higher than the surrounding air temperature, the air near the wall is heated, expands, and rises, forming natural convection driven by buoyancy.

[0105] The characteristic length L is determined based on the geometric parameters of the power distribution equipment casing. In this invention, the characteristic length is the vertical height of the power distribution equipment casing, expressed in meters (m). The specific reasons are as follows: The power distribution equipment casing mainly consists of four vertical walls and a horizontal top surface. The total area of ​​the vertical walls typically accounts for more than 70% of the total heat dissipation area of ​​the casing, making them the dominant walls for convective heat dissipation. The top surface area is relatively small, and the bottom surface is often close to the ground or installed in an elevated layer, where ventilation is limited. Therefore, its contribution to overall convective heat dissipation is far less than that of the vertical walls. Thus, this invention prioritizes the vertical walls and directly uses the vertical height of the power distribution equipment casing as the characteristic length. This parameter can be directly measured using a measuring tape or laser rangefinder during equipment commissioning. For standard models, it can also be obtained directly from the equipment nameplate or factory technical parameter sheet. Once entered, it serves as a fixed parameter in the full lifecycle calculation.

[0106] Next, the qualitative temperature is calculated based on the average wall temperature and the external ambient temperature. The specific calculation formula is as follows:

[0107] ;

[0108] in, For qualitative temperature, The average wall temperature, The external ambient temperature is represented by the constant 273.15, which is the conversion constant between the Celsius scale and the thermodynamic temperature scale. Its function is to convert the average Celsius temperature into the thermodynamic temperature.

[0109] The thermal conductivity and kinematic viscosity of air at the qualitative temperature are calculated based on the air's thermophysical properties. The specific calculation formulas are as follows:

[0110] ;

[0111] in, The thermal conductivity of air, For qualitative temperature, constant 293.15 is the Kelvin temperature corresponding to 20℃, constant 0.0242 is the fitting intercept of air thermal conductivity, and constant 0.000072 is the temperature sensitivity coefficient of air thermal conductivity, which represents the average increment of air thermal conductivity for every 1℃ increase in temperature. The values ​​of the above two constants are derived from the linear regression results of air thermal conductivity experimental data in the NIST REFPROP database. The deviation of linear fitting in the temperature range of -20℃ to +60℃ does not exceed ±1.5%, which fully meets the engineering calculation accuracy requirements for the outdoor operating environment temperature range of power distribution equipment.

[0112] ;

[0113] in, For air kinematic viscosity, This is the qualitative temperature, with a constant of 1.48 × 10⁻⁶. -5 This is the reference value for the kinematic viscosity of air at 20°C. This value comes from the standard data of the kinematic viscosity of air at 20°C under standard atmospheric pressure in the NIST REFPROP database. The constant 293.15 is the Kelvin temperature corresponding to 20°C, the exponent 1.5 is the dimensionless temperature exponent, derived from the theoretical derivation in the kinetic theory of gases that viscosity is proportional to the 3 / 2 power of absolute temperature, and the constant 110.4 is the Sutherland constant for air.

[0114] Subsequently, the Reynolds number is calculated based on wind speed and characteristic length. The specific calculation formula is as follows:

[0115] ;

[0116] in, V is the Reynolds number, representing the ratio of inertial forces to viscous forces. It is a core dimensionless number for determining the state of fluid flow. wind For wind speed, Let L be the viscosity of air motion, and L be the characteristic length of the power distribution equipment.

[0117] The airflow state on the shell surface is determined based on the Reynolds number. In this invention, when Re < 5 × 10⁻⁶... 5 When the flow is classified as laminar, C = 0.664 and m = 0.5 are used; when Re ≥ 5 × 10⁻⁶, the flow is classified as laminar. 5 The flow was determined to be in a turbulent state, with C = 0.037 and m = 0.8. Among these, 5 × 10⁻⁶... 5 is the critical Reynolds number at which air transitions from laminar to turbulent flow when it sweeps across a flat plate. C is the proportionality coefficient of the Nusselt number correlation, a dimensionless parameter reflecting the baseline intensity of convective heat transfer under the flow conditions; m is the exponential term of the Reynolds number, a dimensionless parameter reflecting the nonlinear influence of flow velocity changes on the intensity of convective heat transfer. Based on this, the forced convection heat transfer coefficient components are calculated, with the specific formula as follows:

[0118] ;

[0119] in, For forced convection heat transfer coefficient component, C is the proportionality coefficient of the Nusselt number correlation, Re is the Reynolds number (dimensionless), m is the exponential term of the Reynolds number, Pr is the Prandtl number (dimensionless parameter reflecting the ratio of fluid momentum diffusion capacity to heat diffusion capacity, taken as Pr≈0.71 under normal air conditions), λ is the thermal conductivity of air, and L is the characteristic length.

[0120] Simultaneously, the natural convection heat transfer coefficient component is calculated:

[0121] ;

[0122] in, This represents the natural convection heat transfer coefficient component. Let L be the temperature difference of the wall environment, L be the characteristic length, and 1.42 be an empirical coefficient obtained by fitting experimental data of natural convection on a vertical wall, which remains constant in an ambient temperature and pressure environment. The ratio of temperature difference to characteristic length in this formula... The 1 / 4 power originates from the natural convection boundary layer theory, where the boundary layer thickness is related to... It is directly proportional to, therefore the convective heat transfer coefficient is... Proportional.

[0123] Ultimately, the convective heat transfer coefficient is the sum of the forced convection heat transfer coefficient component and the natural convection heat transfer coefficient component:

[0124] ;

[0125] in, The convective heat transfer coefficient is... For the forced convection heat transfer coefficient component, Let be the component of the natural convection heat transfer coefficient. The physical basis for using linear superposition is that the contributions of forced convection and natural convection to wall heat transfer are independent of each other, and the total heat transfer can be regarded as the algebraic sum of the contributions of the two mechanisms.

[0126] As shown in the multi-factor integrated model above, the convective heat transfer coefficient is simultaneously affected by wind speed, the temperature difference between the wall surface and the environment, the equipment casing size, and the air temperature. All of these variables are collected in real time during step S1 or entered once when the equipment is put into operation. Therefore, this model is entirely based on existing data sources in the monitoring system and requires no additional sensors. Compared to the exponential function model that solely relies on wind speed, the integrated model of this invention can more accurately reflect the true convective heat transfer state of the power distribution equipment casing in the actual outdoor environment, thereby improving the calculation accuracy of the equivalent external temperature.

[0127] To further reflect the ability of the outside air to remove moisture, this step also involves a drying force correction to construct a drying force index. The drying force index comprehensively considers factors such as wind speed, ambient temperature, and relative humidity, and is used to describe the ability of the external environment to promote the evaporation process of the water film on the equipment wall. It provides a basis for subsequent standardization of evaporation relaxation time. The specific calculation formula is as follows:

[0128] ;

[0129] Where DI is the drying power index, and V wind For wind speed, RH out For external relative humidity, T out The external ambient temperature is used as the reference value. The constant 100 in the formula converts the relative humidity expressed as a percentage into a small value between 0 and 1 for calculation purposes. This drying driving term, which reflects the degree of air unsaturation, has a constant of 273.15. Its role is to convert Celsius temperature into thermodynamic temperature, because the ability of air to carry water vapor is proportional to absolute temperature. This conversion is a basic requirement of the classical drying kinetic equation in mass transfer.

[0130] As can be seen from the above relationship, the higher the wind speed and the lower the air humidity, the stronger the air's ability to carry water vapor, and the higher the dryness index. Conversely, when the air humidity is high and the wind speed is low, the dryness index decreases, indicating that the evaporation capacity after condensation forms is weak. The subsequent step S4 will use this dryness index to standardize and correct the evaporation process, thereby reducing the impact of different external environmental conditions on the evaluation of evaporation time.

[0131] After completing the above calculations, the cleaned raw time-series data, along with the calculated internal air dew point temperature, wall supercooling, equivalent external temperature, and drying index, constitute a standardized feature database. This standardized feature database serves as the output data for this step and is divided into two paths according to the data flow: the first path extracts the equivalent external temperature, internal air temperature, internal relative humidity, external relative humidity, and wind speed, serving as the input for shell response time constant identification in step S3; the second path extracts the wall supercooling, drying index, and the near-wall micro-airflow velocity sequence collected in step S1, serving as the input for condensation-evaporation relaxation and thermal plume extraction in step S4, thus completing the data splitting for the subsequent two analysis paths.

[0132] S3: Based on a standardized feature database, extract dynamic response feature parameters of the power distribution equipment housing and construct a dynamic response fingerprint feature vector of the equipment. The dynamic response feature parameters of the housing include thermal response time constant, wet response time constant and damping ratio.

[0133] Furthermore, step S3 also includes:

[0134] Obtain the equivalent external temperature, internal air temperature, internal relative humidity, external relative humidity, and wind speed from the standardized feature database;

[0135] The thermal response process of the shell is equivalent to a first-order RC network, and a continuous thermal response model is constructed, the specific expression of which is:

[0136] ;

[0137] in, For equivalent external temperature, The thermal response time constant, This refers to the internal air temperature.

[0138] The continuous thermal response model is discretized using the forward Euler difference method to obtain the discrete thermal response model, the specific expression of which is:

[0139] ;

[0140] in, Let K be the internal air temperature at the k-th sampling point. Let K be the internal air temperature at the (k-1)th sampling point. The equivalent external temperature of the (k-1)th sampling point For thermal response recursion coefficients, This is the thermal excitation gain coefficient;

[0141] The thermal response recursion coefficients are updated in real time using a recursive least squares algorithm. Then, based on the correspondence between the discrete model and the continuous model, the thermal response time constant under the current operating condition is calculated. The specific calculation formula is as follows:

[0142] ;

[0143] in, The thermal response time constant, The sampling period is The recursive coefficients of the thermal response obtained through recursive identification;

[0144] The wind speed correction factor is constructed based on the wind speed, and the specific formula is as follows:

[0145] ;

[0146] in, V is the wind speed correction factor. wind Where β is the wind speed and β is the wind speed correction factor;

[0147] A continuous wet response model is established, with the following formula:

[0148] ;

[0149] in, The wet response time constant is This refers to the internal relative humidity. External relative humidity, This is the wind speed correction factor;

[0150] The wet response continuous model is discretized using the same forward Euler difference method as the thermal response continuous model, and the parameters of the wet response discrete model are identified online using the recursive least squares algorithm to obtain the wet response time constant under the current operating conditions.

[0151] The specific formula for calculating the dynamic damping ratio of the equipment is as follows:

[0152] ;

[0153] in, For the damping ratio, A in A represents the amplitude of the internal response signal fluctuation. out This corresponds to the fluctuation amplitude of the external excitation signal;

[0154] By integrating thermal response time constant, wet response time constant, and damping ratio, a dynamic response fingerprint feature vector of the equipment is constructed.

[0155] Specifically, this step aims to extract dynamic response characteristic parameters of the casing from the equivalent external temperature and internal / external temperature and humidity data output in step S2 using system identification theory. When the casing of power distribution equipment is subjected to changes in the external environment, its internal temperature and humidity do not change immediately with the external environment. Instead, due to the combined effects of the casing's thermal resistance, thermal capacity, and sealing structure, it exhibits a certain degree of response hysteresis. Therefore, this invention equates the equipment casing to a first-order thermal resistance-thermal capacity dynamic network model. By analyzing the dynamic relationship between external environmental excitation and internal response, it identifies the thermal response time constant, wet response time constant, and damping ratio under the current operating state of the equipment, constructing a dynamic response fingerprint feature vector of the equipment. This provides key characteristics reflecting the equipment's sealing performance and thermal inertia for subsequent fusion diagnosis.

[0156] The data input in this step includes equivalent external temperature, internal air temperature, internal relative humidity, external relative humidity, and wind speed. Among them, the equivalent external temperature calculated in step S2 is used as the excitation input for system identification, replacing the traditional method of directly using the external ambient temperature.

[0157] Using the equivalent external temperature as the excitation signal has clear physical significance. Since power distribution equipment is exposed to the outdoors for extended periods, its actual thermal boundary conditions are affected not only by ambient air temperature but also by the combined effects of solar radiation heating and wind-induced heat dissipation. For example, under summer conditions of intense sunlight and low wind speed, although the external ambient temperature may only be 35°C, the equivalent external temperature calculated in step S2 may reach over 55°C due to continuous solar radiation heating of the casing surface. For equipment with normal sealing performance, the internal air temperature will gradually increase as the casing absorbs heat. If the external ambient temperature is still used as the system input, the input excitation will be underestimated, leading to the incorrect assumption that the internal response speed of the equipment is too slow, thus misjudging a decrease in sealing performance or abnormal thermal inertia. However, using the equivalent external temperature, which comprehensively considers the effects of solar radiation and wind speed, as the excitation signal can more accurately reflect the actual thermal excitation experienced by the equipment, allowing the identified response time constant to truly characterize the thermal inertia characteristics of the casing and improving the accuracy of sealing performance diagnosis.

[0158] Specifically, the thermal response process of the shell is equivalent to a first-order RC network, and a continuous thermal response model is constructed, the specific expression of which is:

[0159] ;

[0160] in, The equivalent external temperature calculated in step S2, The thermal response time constant, This refers to the internal air temperature.

[0161] To achieve online identification, the aforementioned continuous thermal response model is discretized. The continuous thermal response model is discretized using the forward Euler difference method, with a sampling period set to [value missing]. Let the discrete sampling time be k, then we can obtain:

[0162] ;

[0163] After simplification, the discretized recurrence form is obtained:

[0164] ;

[0165] make Then the formula can be expressed as:

[0166] ;

[0167] in, Let K be the internal air temperature at the k-th sampling point. Let K be the internal air temperature at the (k-1)th sampling point. The equivalent external temperature of the (k-1)th sampling point For thermal response recursion coefficients, The sum of the two is always equal to 1, which reflects the physical constraint of the first-order system: the output equals the input in steady state.

[0168] The thermal response recursion coefficients are updated in real time using a recursive least squares algorithm. Then, based on the correspondence between the discrete and continuous models of the system, the thermal response time constant under the current operating condition is calculated. The specific calculation formula is as follows:

[0169] ;

[0170] in, The thermal response time constant, The sampling period is the same as the data sampling period in step S1. These are the recursive coefficients of the thermal response obtained through recursive identification. The above derivation process shows that the thermal response time constant can be directly obtained after identifying the recursive coefficients of the thermal response using the recursive least squares algorithm, without the need to calculate the thermal excitation gain coefficient separately.

[0171] After completing the thermal response identification, the humidity response process of the equipment is identified. In addition to being affected by the sealing performance of the equipment, humidity changes are also affected by the external airflow velocity on the moisture exchange efficiency. Therefore, this invention introduces a wind speed correction factor in the humidity response identification process to weight and correct the moisture convection exchange process, thereby improving the physical accuracy of the humidity response time constant.

[0172] First, construct the wind speed correction factor based on the external wind speed output in step S2. The specific formula is as follows:

[0173] ;

[0174] in, V is the wind speed correction factor. wind Let β be the wind speed, and β be the wind speed correction coefficient, in s / m. The physical meaning of this coefficient is the relative rate of change in moisture exchange efficiency caused by a unit change in wind speed. According to the experimental research conclusions in the literature "The Influence of Temperature and Wind Speed ​​on the Dynamic Characteristics of Humidity Sensors," within the wind speed range of 0–5 m / s, the step response time constant of the humidity sensor decreases approximately linearly with increasing wind speed, and the wind speed influence coefficient obtained by regression is approximately 0.08–0.12 s / m. Considering both the airflow characteristics inside the power distribution equipment and the moisture permeation law of the casing gaps, this embodiment preferably selects β = 0.10 s / m.

[0175] Subsequently, a continuous wet response model was established, with the following formula:

[0176] ;

[0177] in, The wet response time constant is This refers to the internal relative humidity. External relative humidity, This is the wind speed correction factor.

[0178] The aforementioned continuous wet response model is discretized using the same forward Euler difference method as the continuous thermal response model, and the parameters of the discrete model are identified online using a recursive least squares algorithm to calculate the wet response time constant under the current operating condition. By introducing a wind speed correction factor, when the external wind speed increases, the system can reflect the accelerated response effect caused by the increased moisture exchange rate, thereby avoiding misinterpreting changes in wind speed as changes in equipment sealing performance.

[0179] To further evaluate the equipment housing's ability to isolate itself from external environmental disturbances, this step calculates the equipment's dynamic damping ratio. The damping ratio is defined as the ratio between the amplitude of the internal response fluctuation and the amplitude of the external excitation fluctuation; the specific calculation formula is as follows:

[0180] ;

[0181] in, For the damping ratio, A in A represents the amplitude of the internal response signal fluctuation. out This corresponds to the fluctuation amplitude of the external excitation signal.

[0182] It should be noted that the calculation method for the internal response signal fluctuation amplitude is as follows: the sliding analysis window is set to a length of 10 minutes, covering 120 sampling points. Then, the difference between the maximum and minimum internal air temperature within the current sliding analysis window is taken as the internal response signal fluctuation amplitude. The corresponding calculation method for the external excitation signal fluctuation amplitude is to take the difference between the maximum and minimum equivalent external temperature within the same analysis window.

[0183] The damping ratio reflects the equipment casing's ability to isolate high-frequency disturbances from the outside environment. A smaller damping ratio indicates a stronger attenuation effect of the casing on high-frequency fluctuations, a more stable internal environment, and better sealing and thermal insulation performance. Conversely, when the damping ratio increases and approaches 1, it indicates a decrease in the casing's isolation capability, making it easier for external disturbances to be transmitted into the equipment. Therefore, the damping ratio can be used as an important dynamic characteristic for evaluating the sealing condition of equipment.

[0184] After extracting the above-mentioned shell dynamic response characteristic parameters, the thermal response time constant, wet response time constant and damping ratio calculated under the current working condition are used to form the device dynamic response fingerprint feature vector, which is used as the only output data of this step and sent to step S5.

[0185] S4: Based on a standardized feature database, perform condensation and evaporation relaxation analysis and thermal plume extraction to construct wall phase change and fluid feature vectors;

[0186] Furthermore, step S4 also includes:

[0187] Obtain wall subcooling, drying force index, and near-wall micro-airflow velocity sequences from a standardized feature database;

[0188] Based on the wall supercooling and combined with the status of the polymer thin film capacitive condensation sensor, the original evaporation relaxation time corresponding to this condensation event is obtained.

[0189] A standardized correction model for evaporation time is established using the drying power index to calculate the standardized evaporation relaxation time. The specific calculation formula is as follows:

[0190] ;

[0191] in, To standardize the evaporation relaxation time, This is the original evaporation relaxation time. The drying index corresponds to the current operating environment. Standard drying power index;

[0192] Based on the near-wall micro-airflow velocity sequence, the average micro-airflow velocity at each measuring point is calculated using the following formula:

[0193] ;

[0194] in, Let M be the average micro-airflow velocity at the i-th measurement point within the current analysis window, M be the number of sampling points within the current analysis window, and k be the index number of the sampling point. Let be the near-wall micro-airflow velocity at the k-th sampling point.

[0195] Further calculation of the micro-airflow pulsation coefficient at the corresponding measuring point is performed using the following formula:

[0196] ;

[0197] in, Let be the micro-airflow pulsation coefficient at the i-th measuring point. The near-wall micro-airflow velocity at the k-th sampling point is the velocity of the airflow at the i-th measurement point. The average micro-airflow velocity corresponds to the measurement point, M is the number of sampling points in the current analysis window, and k is the index number of the sampling point;

[0198] By integrating standardized evaporation relaxation time and micro-airflow pulsation coefficients at each measuring point, a wall phase change and fluid characteristic vector are constructed.

[0199] Specifically, this step aims to jointly analyze the evaporation process of the liquid film on the internal wall of the power distribution equipment and the near-wall airflow characteristics based on the wall supercooling, drying index, and near-wall micro-airflow velocity sequence from the standardized feature database output in step S2. It extracts dynamic features that reflect the inherent moisture state and local overheating state of the equipment. This step consists of a condensation and evaporation relaxation analysis submodule and a thermal plume extraction submodule. The condensation and evaporation relaxation analysis uses the drying index calculated in step S2 to perform environmental normalization correction on the original evaporation relaxation time, eliminating the influence of different wind speeds and ambient humidity conditions on the evaporation rate. The thermal plume extraction uses the near-wall micro-airflow velocity sequence to spatially locate local thermal disturbances inside the equipment. Finally, the results of the two analyses are fused to form a wall phase change and fluid feature vector, which is then sent to step S5 for three-level fusion diagnosis.

[0200] First, based on the wall supercooling and the status of the polymer thin-film capacitive condensation sensor, the evaporation process of each condensation event is precisely timed.

[0201] Since the evaporation of the liquid film on the device wall is a dynamic process with a definite start and end state, this invention uses the moment when the wall supercooling crosses zero as the time reference and establishes the timing logic for the evaporation relaxation time in conjunction with the state of the condensation sensor. Specifically, when the wall supercooling rises continuously from a negative value and crosses zero for the first time, it indicates that the wall temperature has once again exceeded the internal air dew point temperature, and the wall has begun to meet the conditions for evaporation. If the polymer thin-film capacitive condensation sensor is still in a wet state at this time, i.e., it detects that a continuous liquid film still exists on the wall, then this moment is defined as the evaporation start time t. start Continue monitoring the output of the condensation sensor. When the condensation sensor returns from a wet state to a dry state, and the sensor impedance returns to within 95% of the corresponding dry baseline impedance value, it is determined that the liquid film on the wall has basically evaporated, and this moment is defined as the evaporation end time t. end The original evaporation relaxation time corresponding to this condensation event is calculated as follows:

[0202] ;

[0203] in, The original evaporation relaxation time is expressed in minutes (t). start t is the time when evaporation begins. end This is the time when evaporation ends.

[0204] The above-mentioned original evaporation relaxation time truly reflects the time it takes for the liquid film on the equipment wall to completely evaporate. However, since external wind speed and ambient humidity can significantly affect the evaporation rate, the original evaporation relaxation time obtained under different regions and weather conditions cannot be directly used to evaluate the degree of moisture absorption of the equipment.

[0205] If the wall undercooling always meets the requirements throughout the entire monitoring period If no condensation occurs inside the equipment, the system will not calculate the original evaporation relaxation time, and this submodule will not output the condensation and evaporation relaxation analysis results, but will only retain the thermal plume analysis results.

[0206] To eliminate the influence of environmental factors, a standardized correction model for evaporation time is established using the drying power index. This model uniformly converts evaporation times measured under actual environmental conditions to a standard reference environment. The formula for calculating the standardized evaporation relaxation time is as follows:

[0207] ;

[0208] in, To standardize the evaporation relaxation time, This is the original evaporation relaxation time. The drying index corresponds to the current operating environment. The standard drying power index.

[0209] It should be noted that the standard dryness index corresponds to a preset reference environment. In this invention, the standard damp heat test environment conditions specified in "Environmental Testing Part 2: Test Db: Alternating Damp Heat (12h+12h Cycle)" are used: temperature 25℃, relative humidity 95%. Taking into account the low wind speed (0.5m / s) under natural ventilation conditions, the standard reference environment is set as: wind speed 0.5m / s, ambient temperature 25℃, and relative humidity 60%. Substituting the above parameters into the dryness index calculation formula in step S2, we can obtain... ≈59.6 m·K / s, therefore, in this embodiment, we take... = 60m·K / s is used as the fixed reference value for the system.

[0210] The standardized calculations described above essentially convert evaporation times measured under different weather conditions to the same reference environment. For example, in windy and low-humidity environments, even if the actual evaporation rate of the liquid film is faster, the conversion can still restore the equipment's true dehumidification capacity, thus avoiding misjudgments caused by environmental differences.

[0211] To further enable individualized evaluation across different devices, a dynamic health baseline establishment mechanism is introduced. During the first three months of break-in and health maintenance after the equipment is put into operation, the system continuously records the standardized evaporation relaxation time corresponding to each condensation event. The data collected during this phase will be used as historical samples of the equipment's health status to establish a standardized dynamic health baseline for evaporation relaxation time specific to the equipment.

[0212] Suppose that N standardized evaporative relaxation times are collected during the healthy period, then first calculate their mean and standard deviation, using the following formulas:

[0213] ;

[0214] ;

[0215] in, The average evaporative relaxation time during the healthy period. The standard deviation of the standardized evaporative relaxation time during the healthy period. Let be the standardized evaporation relaxation time for the i-th time, where i is the index number and N is the total number of condensation events recorded during the healthy period.

[0216] Further establish a standardized dynamic health baseline for evaporation relaxation time specific to the equipment. The specific calculation formula is as follows:

[0217] ;

[0218] The baseline is the standardized evaporative relaxation time dynamic healthy baseline. According to the 68-95-99.7 rule of normal distribution in statistics, the μ±2σ interval covers approximately 95% of the normal fluctuation range. The average standardized evaporation time during the healthy period. Standard deviation of evaporation time during the healthy period;

[0219] During subsequent operation, the standardized evaporation relaxation time calculated each time was compared with the dynamic health baseline of the standardized evaporation relaxation time corresponding to this device. A two-level judgment mechanism was adopted to balance statistical stability and physical accuracy. The first level is the attention threshold, when... When this occurs, the event is marked as a concern event, and continuous tracking is initiated without immediate alarm. The 1.5 times threshold is set based on the warning limit setting principle commonly used in cumulative sum control charts (CUSUM)—early warning tracking is initiated when the process mean deviation reaches 1.5 times the standard deviation. This principle has been widely used in industrial process monitoring. The second level is the alarm threshold; if two consecutive independent condensation-evaporation events both meet the threshold... If the condensation is not detected, the equipment is deemed to have inherent dampness or a tendency for the anti-condensation coating to fail. The 1.8-fold threshold is established based on the following three criteria: First, according to the theory of droplet evaporation kinetics on material surfaces, when the anti-condensation coating on the equipment surface ages or accumulates contaminants, the wall surface gradually changes from a hydrophobic state to a hydrophilic state, significantly enhancing the adhesion of the liquid film to the wall surface and leading to an increase in evaporation activation energy. Numerous experimental studies have shown that when the surface contact angle decreases from approximately 95° in a healthy state to approximately 45° in a defective state, the complete evaporation time of the liquid film per unit area typically increases by approximately 1.7 to 1.9 times. Second, the "Guidelines for Measuring the Wettability of Insulator Surfaces" stipulates that the critical contact angle for the transition from hydrophobic to hydrophilic surface wettability is 50°, and there is a clear physical correspondence between the corresponding change in surface energy and the change in evaporation rate. Third, combined with the condensation-evaporation cycle evaluation method specified in the alternating damp heat test, if the evaporation time exceeds 1.8 times the benchmark value after the sample surface has undergone a specified number of temperature and humidity cycles, the surface anti-condensation performance is deemed unqualified. Based on the above three aspects, this invention selects 1.8 times as the engineering criterion for significant degradation of material surface properties. Furthermore, confirmation through two consecutive independent condensation events further avoids accidental false alarms caused by extreme weather, improving the reliability of moisture diagnosis.

[0220] After completing the condensation analysis, the near-wall micro-airflow velocity sequence collected in step S1 is further used to analyze the local thermal plume inside the equipment in order to identify local overheated areas.

[0221] Inside enclosed power distribution equipment, under normal operating conditions, air is driven by heat and buoyancy to form a stable laminar or weakly turbulent flow, with relatively gentle changes in the micro-airflow at each measuring point. When a component inside experiences localized overheating, the surrounding air rises rapidly, forming a localized thermal plume. This causes significant fluctuations in the near-wall micro-airflow velocity in the corresponding area, accompanied by turbulent flow direction. Therefore, by analyzing the dynamic fluctuations in the near-wall micro-airflow velocity sequence at different measuring points, the localized heat source can be spatially located.

[0222] Let the near-wall micro-airflow velocity corresponding to the i-th measuring point be... First, calculate its average speed within the current analysis window, which is typically set to 10 minutes. The specific calculation formula is as follows:

[0223] ;

[0224] in, Let M be the average micro-airflow velocity at the i-th measurement point within the current analysis window, M be the number of sampling points within the current analysis window, and k be the index number of the sampling point. Let be the near-wall micro-airflow velocity at the k-th sampling point.

[0225] Further calculation of the micro-airflow pulsation coefficient at the corresponding measuring point is performed using the following formula:

[0226] ;

[0227] in, Let be the micro-airflow pulsation coefficient at the i-th measuring point, whose physical meaning is the relative pulsation intensity of the flow velocity. The near-wall micro-airflow velocity at the k-th sampling point is the velocity of the airflow at the i-th measurement point. The average micro-airflow velocity corresponds to the measurement point, M is the number of sampling points in the current analysis window, and k is the index number of the sampling point.

[0228] This coefficient reflects the stability of airflow in the corresponding area. The larger the micro-airflow pulsation coefficient, the more intense the local thermal plume, and the higher the probability of local overheating in the corresponding area. Therefore, by spatially comparing the micro-airflow pulsation coefficients obtained from all measuring points, the distribution of thermal plumes inside the equipment can be obtained. The area with the largest pulsation coefficient corresponds to the local overheating hotspot mapping area, providing a basis for subsequent local overheating location.

[0229] After completing the above analysis, the standardized evaporation relaxation time and the micro-airflow pulsation coefficient at each measuring point are combined to form the wall phase change and fluid characteristic vector, which is used as the only output data of this step and sent to step S5.

[0230] S5: Establish a dynamic health baseline library, combine the device's dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, perform a three-level fusion diagnosis, and output the diagnosis results, which include false alarm suppression, sealing warning, moisture warning, and overheating location.

[0231] Specifically, this step aims to integrate the device dynamic response fingerprint feature vector output in step S3, the wall phase change and fluid feature vector output in step S4, and the dynamic health baseline library established during device operation. Through longitudinal historical comparison and multi-feature fusion analysis, a three-level fusion diagnosis of the operating status of power distribution equipment is performed to distinguish between normal responses caused by environmental changes and defects in the equipment itself. This enables accurate identification of different operating scenarios such as false alarm suppression, sealing defect warning, moisture warning, and local overheating location.

[0232] This step's input includes three parts of data: the device's dynamic response fingerprint feature vector output in step S3, the wall phase transition and fluid feature vector output in step S4, and the dynamic health baseline library. The dynamic health baseline library is not a fixed threshold, but rather a device-specific health model continuously updated during the device's long-term operation.

[0233] Specifically, the construction method of the dynamic health baseline library is as follows: During the first three months of break-in and health period after the equipment is put into operation, the system continuously records the thermal response time constant, wet response time constant, damping ratio obtained in step S3, and the standardized evaporation relaxation time and micro-airflow pulsation coefficient at each measuring point obtained in step S4. The mean and standard deviation of each characteristic parameter are calculated respectively to construct the dynamic health baseline for each characteristic. The specific construction process is similar to the construction process of the standardized evaporation relaxation time dynamic health baseline. For the thermal response time constant, the dynamic health baseline is μ. τT +2σ τT For the wet response time constant, the dynamic health baseline is μ. τRH +2σ τRH For the damping ratio, the dynamic health baseline is μ. ξ + 2σ ξ For the micro-airflow pulsation coefficient at each measuring point, the dynamic health baseline is μ. Ci +3σ Ci This value employs a more stringent 3σ criterion to reduce the false alarm rate of overheating. As equipment operating time increases, the dynamic health baseline library uses a sliding window update mechanism, with the window length preferably being the data from the most recent 6 months. This ensures that each device has its own exclusive dynamic health baseline that matches its operating environment and installation conditions, enabling longitudinal self-learning and individualized diagnosis.

[0234] After data reading is complete, the deviation of each feature from the dynamic health baseline under the current operating state is first calculated. For the thermal response time constant, wet response time constant, and damping ratio, the deviation rate is used as the evaluation index, and the specific calculation formula is as follows:

[0235] ;

[0236] in, Let X be the deviation rate of the i-th feature relative to the dynamic health baseline. i The feature value corresponding to the current time. The dynamic health baseline value is the corresponding feature in the dynamic health baseline library.

[0237] Based on the aforementioned deviation rates, each diagnostic dimension is assessed.

[0238] For the thermal response time constant, when it decreases by more than 15% relative to the dynamic health baseline, it indicates that the overall thermal inertia of the equipment casing has decreased significantly, the insulation and heat insulation capacity of the casing has weakened, and a clear hot and cold air exchange channel has been formed between the inside and outside air of the equipment. This usually corresponds to defects such as damaged cabinet door sealing strips, detached cable hole seals, or structural seal failure.

[0239] For the wet response time constant, a decrease of more than 15% relative to the dynamic health baseline indicates a decline in the equipment's buffering capacity against external water vapor intrusion. Since water vapor molecules are smaller than the main components of air, they more easily penetrate through microcracks and micropores in the sealing material. Therefore, when the equipment's sealing material undergoes early aging, the wet response time constant usually changes before the thermal response time constant, and can be used to identify micropore-penetrating seal failures.

[0240] For the damping ratio ξ, when it increases by more than 20% relative to the dynamic health baseline, it indicates that the equipment housing's ability to isolate high-frequency external disturbances such as diurnal temperature differences and sudden weather changes has significantly decreased, and the internal environment of the equipment is more sensitive to external changes. This usually reflects the elastic decay of the sealing strip, the formation of multiple small leaks, or a decline in overall isolation performance.

[0241] The dynamic health baseline for the normalized evaporation relaxation time output in step S4 has been explained in the above steps. When the normalized evaporation relaxation time meets... When this occurs, it indicates a significant decrease in the evaporation capacity of the liquid film on the wall and an increase in the hydrophilicity of the equipment surface. This usually corresponds to defects such as aging and peeling of the anti-condensation coating, deep moisture absorption of the insulating support components, or long-term dampness of the wall surface.

[0242] For the micro-airflow pulsation coefficients of each measuring point output in step S4, if the micro-airflow pulsation coefficient of a certain measuring point is greater than the dynamic health baseline, it is determined that there is a significant abnormal thermal plume in the area, and the area where the measuring point with the largest pulsation coefficient is located is determined as the location of the local hot spot.

[0243] After completing the deviation analysis of each feature, a three-level fusion diagnostic matrix is ​​constructed based on each dimension to comprehensively determine the equipment status.

[0244] First, when the thermal response time constant, wet response time constant, and damping ratio are all within ±5% of the healthy baseline, and the standardized evaporation relaxation time does not exceed 1.2 times the healthy baseline, and the micro-airflow pulsation at each measuring point remains stable, it is determined that the current equipment is only affected by changes in the external environment, and there is no abnormality in the equipment itself. At this time, the diagnostic result "false alarm suppression" is output. The system automatically filters alarm information caused by environmental factors such as diurnal temperature differences and sudden weather changes to avoid false alarms.

[0245] Secondly, when the thermal response time constant, wet response time constant, and damping ratio all deteriorate, while the standardized evaporation relaxation time and micro-airflow pulsation remain normal, it indicates a decrease in the overall sealing capability of the equipment. In this case, further subdivision is made based on the degree of decoupling between the thermal and wet response time constants: if the decrease in the wet response time constant is more than 5 percentage points greater than that of the thermal response time constant, it is determined that micro-cracks have occurred in the sealing strip or that material aging has led to micropore penetration; if the decreases in both are roughly synchronized, it is determined that the equipment has structural sealing leaks such as detached cabinet door sealing strips or failed cable hole sealing. Both situations will uniformly output a diagnostic result of "sealing warning," and planned maintenance will be arranged according to the degree of sealing performance degradation.

[0246] Furthermore, when the thermal response time constant, wet response time constant, and damping ratio remain normal, but the standardized evaporation relaxation time exceeds 1.8 times the dynamic health baseline, and the micro-airflow pulsation remains stable, it indicates that the overall sealing performance of the equipment is normal, but the wall evaporation capacity has significantly decreased, which suggests that the anti-condensation coating of the equipment has failed or the insulation material has become deeply damp. At this time, the diagnostic result "Moisture Warning" is output, prompting timely anti-condensation treatment or insulation condition inspection.

[0247] If the thermal response time constant, wet response time constant, and damping ratio all deteriorate, and the standardized evaporation relaxation time exceeds 1.8 times the dynamic health baseline, it indicates that the equipment has simultaneously experienced sealing failure and wall moisture absorption. These two defects working together represent the most severe operating condition. In this case, the output diagnostic result will still be "moisture warning," but the system will simultaneously raise the alarm level and recommend immediate power outage maintenance to prevent further deterioration of the equipment's insulation performance.

[0248] Finally, if no condensation event is detected in step S4, but the micro-airflow pulsation coefficient at a certain measuring point is significantly higher than the dynamic health baseline, and the thermal response time constant, wet response time constant, and damping ratio remain normal, it indicates that the overall sealing condition of the equipment is good, but there is an abnormal heat source in a localized area. The system determines the location of the overheating hotspot based on the measuring point with the largest micro-airflow pulsation coefficient, outputs the diagnostic result "overheating location," and suggests infrared temperature measurement and further inspection of the contacts, busbar connections, or magnetic components in the corresponding area.

[0249] After completing the above fusion analysis, this step finally outputs four types of diagnostic results, including: false alarm suppression, sealing warning, moisture warning, and overheating location.

[0250] Example 2: Based on the same inventive concept as the multi-physical parameter analysis-based power distribution equipment defect diagnosis method in the foregoing examples, this application also provides a multi-physical parameter analysis-based power distribution equipment defect diagnosis system. Please refer to the appendix. Figure 2 The system includes:

[0251] The multi-parameter synchronous acquisition module 11 is used to deploy a multi-parameter synchronous sensor group through an internal and external differential distribution point method to acquire the original time series data of the power distribution equipment and construct an internal and external multi-dimensional original time series database. The multi-parameter synchronous sensor group includes an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group.

[0252] The thermal and humidity environment correction module 12 is used to perform data cleaning, key derived feature calculation, thermal balance correction and drying force correction on the internal and external multidimensional original time series database to obtain a standardized feature database.

[0253] The device dynamic response fingerprint feature construction module 13 is used to extract dynamic response feature parameters of the power distribution equipment shell based on a standardized feature database and construct a device dynamic response fingerprint feature vector. The shell dynamic response feature parameters include thermal response time constant, wet response time constant and damping ratio.

[0254] The wall phase change and fluid feature construction module 14 is used to perform condensation and evaporation relaxation analysis and thermal plume extraction based on a standardized feature database, and to construct a wall phase change and fluid feature vector.

[0255] The fusion diagnostic module 15 is used to establish a dynamic health baseline library, combine the device dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, perform three-level fusion diagnosis, and output diagnostic results. The diagnostic results include false alarm suppression, sealing warning, moisture warning and overheating location.

[0256] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0257] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for defect diagnosis of power distribution equipment based on multi-physical parameter analysis, characterized in that, Includes the following steps: By deploying multi-parameter synchronous sensor groups using an internal-external differential distribution point method, the original time series data of the power distribution equipment is obtained, and an internal and external multi-dimensional original time series database is constructed. The multi-parameter synchronous sensor groups include an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group. Data cleaning, key derived feature calculation, thermal balance correction and drying force correction are performed on the internal and external multidimensional original time series databases to obtain a standardized feature database; Based on a standardized feature database, dynamic response feature parameters of the power distribution equipment casing are extracted, and a dynamic response fingerprint feature vector of the equipment is constructed. The dynamic response feature parameters of the casing include thermal response time constant, wet response time constant, and damping ratio. Based on a standardized feature database, condensation and evaporation relaxation analysis and thermal plume extraction are performed to construct wall phase change and fluid feature vectors. A dynamic health baseline database is established. By combining the device's dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, a three-level fusion diagnosis is performed, and the diagnosis results are output. The diagnosis results include false alarm suppression, sealing warning, moisture warning, and overheating location.

2. The method for defect diagnosis of power distribution equipment based on multi-physical parameter analysis as described in claim 1, characterized in that, The construction of the internal and external multidimensional original time series database includes: Basic meteorological parameters are collected by an external environment sensor group, which includes an ambient temperature sensor and an ambient relative humidity sensor. The basic meteorological parameters include the ambient temperature and the ambient relative humidity. Environmental correction parameters are obtained by an external correction sensor group, which includes a wind speed sensor and a solar irradiance sensor. The environmental correction parameters include wind speed and solar irradiance. Internal air condition information and shell temperature response information are collected by an internal response sensor group. The internal response sensor group includes an internal air temperature sensor, an internal relative humidity sensor, and three wall temperature sensors. The internal air condition information includes internal air temperature and internal relative humidity, and the shell temperature response information is the wall temperature. Information on local air disturbances is collected by an internal disturbance sensor group, which includes a polymer thin film capacitive condensation sensor and a near-wall micro-airflow monitoring unit. The information on local air disturbances is a near-wall micro-airflow velocity sequence. By integrating external ambient temperature, external relative humidity, wind speed, solar irradiance, internal air temperature, internal relative humidity, wall temperature, and near-wall micro-airflow velocity sequences, a multi-dimensional original time-series database is constructed.

3. The method for defect diagnosis of power distribution equipment based on multi-physical parameter analysis as described in claim 1, characterized in that, The acquisition of the standardized feature database includes: Obtain internal and external multidimensional raw time series databases, and perform data cleaning on all raw time series data in the database. The data cleaning includes integrity checks and smoothing processes. The internal air dew point temperature is calculated based on the internal air temperature and relative humidity after data cleaning. The specific calculation formula is as follows: ； Among them, T d T is the internal air dew point temperature. in RH (Relative Tolerance) is the internal air temperature. in Let be the internal relative humidity, and a and b be empirical coefficients of the Magnus formula, where a is a dimensionless coefficient and b is a temperature-dimensional coefficient, a = 17.62, b = 243.12℃. As an intermediate variable; The wall subcooling at the corresponding location is calculated based on the wall temperature and the internal air dew point temperature. The specific calculation formula is as follows: ； in, T represents the wall subcooling. s T represents the corresponding wall temperature. d This refers to the internal air dew point temperature. A convective heat transfer coefficient is set, and a heat balance correction is performed based on the convective heat transfer coefficient to obtain the equivalent external temperature. The specific calculation formula is as follows: ； Among them, T eq For equivalent external temperature, T out The external ambient temperature is represented by Solar, and the solar irradiance is represented by h. conv The convective heat transfer coefficient is... The heat absorption coefficient of the shell surface; A dryness index is constructed, and dryness is corrected by comprehensively considering wind speed, external ambient temperature, and external relative humidity. The specific calculation formula is as follows: ； Where DI is the drying power index, and V wind For wind speed, RH out For external relative humidity, T out External ambient temperature; The raw time-series data after data cleaning, along with the calculated internal air dew point temperature, wall subcooling, equivalent external temperature, and drying index, constitute a standardized feature database.

4. The method for diagnosing defects in power distribution equipment based on multi-physical parameter analysis as described in claim 3, characterized in that, The set convective heat transfer coefficient includes: The average wall temperature is calculated based on the wall surface temperature. The specific calculation formula is as follows: ； in, The average wall temperature, The wall temperatures at three measuring points on the wall are shown respectively. The wall temperature difference is calculated based on the average wall temperature and the ambient temperature. The specific calculation formula is as follows: ； in, Due to the temperature difference in the wall environment, The average wall temperature, External ambient temperature; The characteristic length L is determined based on the geometric parameters of the power distribution equipment housing; The qualitative temperature is calculated based on the average wall temperature and the external ambient temperature. The specific calculation formula is as follows: ； in, For qualitative temperature, The average wall temperature, The external ambient temperature is represented by the constant 273.15, which is the conversion constant between the Celsius scale and the thermodynamic temperature scale. The thermal conductivity and kinematic viscosity of air at the qualitative temperature are calculated based on the air's thermophysical properties. The specific calculation formulas are as follows: ； in, The thermal conductivity of air, The constant 293.15 is the Kelvin temperature corresponding to 20℃, the constant 0.0242 is the fitting intercept of the thermal conductivity of air, and the constant 0.000072 is the temperature sensitivity coefficient of the thermal conductivity of air. ； in, For air kinematic viscosity, The qualitative temperature is constant 1.48 × 10⁻⁶. -5 The constant 293.15 is the reference value for the kinematic viscosity of air at 20℃, the constant 110.4 is the Kelvin temperature corresponding to 20℃, and the constant 293.15 is the Sutherland constant for air. The Reynolds number is calculated based on wind speed and characteristic length, using the following formula: ； in, Let V be the Reynolds number. wind For wind speed, Where L is the viscosity of air motion, and L is the characteristic length of the power distribution equipment; Determine the airflow state on the shell surface based on the Reynolds number. When Re < 5 × 10⁻⁶ 5 The flow is determined to be laminar. The proportional gain of the Nusselt number correlation is set to 0.664, and the exponential term of the Reynolds number is set to 0.

5. When Re ≥ 5 × 10⁻⁶, the flow is considered laminar. 5 The flow is determined to be in a turbulent state. The proportionality coefficient of the Nusselt number correlation is set to 0.037, and the exponential term of the Reynolds number is set to 0.

8. The formula for calculating the forced convection heat transfer coefficient component is as follows: ； in, For the forced convection heat transfer coefficient component, C is the proportionality coefficient of the Nusselt number correlation, Re is the Reynolds number, m is the exponential term of the Reynolds number, Pr is the Prandtl number, λ is the thermal conductivity of air, and L is the characteristic length; Calculate the natural convection heat transfer coefficient components: ； in, This represents the natural convection heat transfer coefficient component. The temperature difference in the wall environment is denoted by L, the characteristic length is L, and the constant 1.42 is an empirical coefficient obtained by fitting experimental data of natural convection on a vertical wall. The convective heat transfer coefficient is obtained by setting the forced convection heat transfer coefficient component and the natural convection heat transfer coefficient component. The specific calculation formula is as follows: ； in, The convective heat transfer coefficient is... For the forced convection heat transfer coefficient component, This represents the natural convection heat transfer coefficient component.

5. The method for defect diagnosis of power distribution equipment based on multi-physical parameter analysis as described in claim 1, characterized in that, The construction device dynamically responds to the fingerprint feature vector, including: Obtain the equivalent external temperature, internal air temperature, internal relative humidity, external relative humidity, and wind speed from the standardized feature database; The thermal response process of the shell is equivalent to a first-order RC network, and a continuous thermal response model is constructed, the specific expression of which is: ； in, For equivalent external temperature, The thermal response time constant, This refers to the internal air temperature. The continuous thermal response model is discretized using the forward Euler difference method to obtain the discrete thermal response model, the specific expression of which is: ； in, Let K be the internal air temperature at the k-th sampling point. Let K be the internal air temperature at the (k-1)th sampling point. The equivalent external temperature of the (k-1)th sampling point For thermal response recursion coefficients, This is the thermal excitation gain coefficient; The thermal response recursion coefficients are updated in real time using a recursive least squares algorithm. Then, based on the correspondence between the discrete model and the continuous model, the thermal response time constant under the current operating condition is calculated. The specific calculation formula is as follows: ； in, The thermal response time constant, The sampling period is The recursive coefficients of the thermal response obtained through recursive identification; The wind speed correction factor is constructed based on the wind speed, and the specific formula is as follows: ； in, V is the wind speed correction factor. wind Where β is the wind speed and β is the wind speed correction factor; A continuous wet response model is established, with the following formula: ； in, The wet response time constant is This refers to the internal relative humidity. External relative humidity, This is the wind speed correction factor; The wet response continuous model is discretized using the same forward Euler difference method as the thermal response continuous model, and the parameters of the wet response discrete model are identified online using the recursive least squares algorithm to obtain the wet response time constant under the current operating conditions. The specific formula for calculating the dynamic damping ratio of the equipment is as follows: ； in, For the damping ratio, A in A represents the amplitude of the internal response signal fluctuation. out This corresponds to the fluctuation amplitude of the external excitation signal; By integrating thermal response time constant, wet response time constant, and damping ratio, a dynamic response fingerprint feature vector of the equipment is constructed.

6. The method for defect diagnosis of power distribution equipment based on multi-physical parameter analysis as described in claim 1, characterized in that, The construction of the wall phase transition and fluid feature vector includes: Obtain wall subcooling, drying force index, and near-wall micro-airflow velocity sequences from a standardized feature database; Based on the wall supercooling and combined with the status of the polymer thin film capacitive condensation sensor, the original evaporation relaxation time corresponding to this condensation event is obtained. A standardized correction model for evaporation time is established using the drying power index to calculate the standardized evaporation relaxation time. The specific calculation formula is as follows: ； in, To standardize the evaporation relaxation time, This is the original evaporation relaxation time. The drying index corresponds to the current operating environment. Standard drying power index; Based on the near-wall micro-airflow velocity sequence, the average micro-airflow velocity at each measuring point is calculated using the following formula: ； in, Let M be the average micro-airflow velocity at the i-th measurement point within the current analysis window, M be the number of sampling points within the current analysis window, and k be the index number of the sampling point. The near-wall micro-airflow velocity at the k-th sample point of the i-th measurement point; Further calculation of the micro-airflow pulsation coefficient at the corresponding measuring point is performed using the following formula: ； in, Let be the micro-airflow pulsation coefficient at the i-th measuring point. The near-wall micro-airflow velocity at the k-th sampling point is the velocity of the airflow at the i-th measurement point. The average micro-airflow velocity corresponds to the measurement point, M is the number of sampling points in the current analysis window, and k is the index number of the sampling point; By integrating standardized evaporation relaxation time and micro-airflow pulsation coefficients at each measuring point, a wall phase change and fluid characteristic vector are constructed.

7. A defect diagnosis system for power distribution equipment based on multi-physical parameter analysis, characterized in that, The system is used to implement the method for diagnosing defects in power distribution equipment based on multi-physical parameter analysis as described in any one of claims 1 to 6, and the system comprises: The multi-parameter synchronous acquisition module is used to deploy a multi-parameter synchronous sensor group through an internal and external differential distribution point method to acquire the original time series data of the power distribution equipment and construct an internal and external multi-dimensional original time series database. The multi-parameter synchronous sensor group includes an external environment sensor group, an external correction sensor group, an internal response sensor group, and an internal disturbance sensor group. The thermal and humidity environment correction module is used to perform data cleaning, key derived feature calculation, heat balance correction and drying force correction on the internal and external multidimensional original time series databases to obtain a standardized feature database. The device dynamic response fingerprint feature construction module is used to extract dynamic response feature parameters of the power distribution equipment shell based on a standardized feature database, and construct a device dynamic response fingerprint feature vector. The shell dynamic response feature parameters include thermal response time constant, wet response time constant and damping ratio. A wall phase change and fluid feature construction module is used to perform condensation and evaporation relaxation analysis and thermal plume extraction based on a standardized feature database, and to construct a wall phase change and fluid feature vector. The fusion diagnostic module is used to establish a dynamic health baseline library, combine the device's dynamic response fingerprint feature vector with the wall phase change and fluid feature vector, perform three-level fusion diagnostics, and output diagnostic results, including false alarm suppression, sealing warning, moisture warning, and overheating location.

Images