Cable intermediate head underground environment comprehensive monitoring method and system

Abstract

The invention relates to the technical field of cable environment monitoring, and discloses a cable intermediate head underground environment comprehensive monitoring method and system, and the method comprises the following steps: S101, obtaining the time and energy sequence of a partial discharge event; step S102, calculating to obtain an original full width and half height; step S103, calculating to obtain a real full width and half height; step S104, calculating to obtain the axial thermal dispersion full width at half maximum of the joint; step S105, calculating to obtain a weighted full width at half maximum; and S106, generating a cable health index. According to the method, an instrument space measurement fuzzy effect is stripped through a point spread function convolution model, scale normalization of thermal parameters under different ventilation working conditions is completed in combination with an outflow cylinder forced convection model, and the real axial thermal dispersion characteristics of the joint can be accurately obtained; therefore, normal fluctuation and early degradation are effectively distinguished, and the accuracy and timeliness of underground cable intermediate head state monitoring are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of cable environmental monitoring technology, and more specifically, to a method and system for comprehensive monitoring of the underground environment at the cable joint. Background Technology

[0002] Cables play a crucial role in underground power transmission systems, and cable joints, due to their relatively weak insulation, are prone to failure in the entire transmission chain. The underground environment, characterized by strong electromagnetic interference and dynamic fluctuations in temperature, humidity, and wind speed, presents numerous obstacles to monitoring the condition of cable joints.

[0003] In traditional monitoring methods, the sampling rates of temperature profiles obtained from distributed fiber optic temperature measurement and discharge event sequences obtained from UHF acoustic emission partial discharge monitoring differ significantly, and the time references for various data are inconsistent, making direct multi-source data fusion analysis impossible. Furthermore, distributed fiber optic temperature measurement equipment suffers from inherent spatial measurement ambiguity, distorting the peak shape parameters of the measured temperature rise profile and failing to reflect the true thermal dissipation state of the joint. Underground ventilation intensity varies significantly with production shifts and seasons, causing fluctuations in the convective heat transfer coefficient of the cable surface. The lack of a unified comparison benchmark for thermal dissipation parameters obtained under different operating conditions makes cross-time period state assessment impossible. In addition, partial discharge events are discrete instantaneous pulse signals, while joint thermal dissipation is a continuous, slowly varying process. Current technologies have not established a time-domain correlation between these two characteristics, failing to reflect the impact of discharge energy on the thermal state.

[0004] Existing monitoring systems mostly monitor single parameters and lack unified quantitative health indicators. Maintenance personnel can only rely on scattered data to judge the status, making it difficult to identify early deterioration trends. This can easily lead to serious accidents such as cable short circuits due to untimely fault warnings, threatening the safety of underground power supply and the continuity of production operations. Summary of the Invention

[0005] This invention provides a method and system for comprehensive monitoring of the underground environment at cable joints, solving the technical problems mentioned in the background.

[0006] This invention provides a method for comprehensive monitoring of the underground environment at cable joints, comprising the following steps: Step S101: Obtain a temperature profile through distributed optical fiber temperature measurement, combine it with the ambient temperature to form a temperature rise profile, collect the time and energy sequence of partial discharge events, and resample to form a unified time grid. Step S102: Calculate the original full width and half height based on the temperature rise profile under a unified time grid. Step S103: Convert the original full width half height to the true full width half height based on the point spread function and the instrument space half height. Step S104: The outer surface of the cable is taken as the object of forced convection of the outflow cylinder. The air physical parameters are determined according to the wind speed, relative humidity and ambient temperature. The convective heat transfer coefficient is calculated by the correlation formula of forced convection of the outflow cylinder. Based on this, the true full width half height is normalized to obtain the axial heat dissipation half height of the joint. Step S105: The partial discharge event time and energy sequence is low-pass processed by smoothing kernel and weighted window to obtain partial discharge energy low-pass sequence, and the weighted half width of the joint axial thermal dispersion is time-weighted using this sequence as weight to obtain weighted half width of the joint. Step S106: Select a reference subset based on the partial discharge energy low-pass sequence within the rolling statistical interval, determine the relative bandwidth by taking the median of the joint axial thermal dissipation half-width within the reference subset and the absolute deviation of the median, and generate the cable health index through exponential mapping.

[0007] The beneficial effects of this invention are as follows: This invention eliminates the spatial measurement ambiguity effect of instruments through a point spread function convolution model, and combines it with an outflow cylindrical forced convection model to achieve scale normalization of thermal parameters under different ventilation conditions, which is beneficial for accurately obtaining the true axial heat dissipation characteristics of the joint. Simultaneously, the system performs time-domain weighted correlation between partial discharge energy and heat dissipation parameters, and generates a quantitative cable health index based on rolling statistical intervals and robust statistical methods. This helps distinguish between normal fluctuations and early degradation, improves the accuracy and timeliness of monitoring the condition of underground cable joints, provides maintenance personnel with intuitive and reliable health status data, reduces the probability of underground power supply failures, and ensures the stable operation of the underground power transmission system. Attached Figure Description

[0008] Figure 1 This is a flowchart of a method for comprehensive monitoring of the underground environment at the cable mid-joint of the present invention; Figure 2 This is a schematic diagram of an integrated monitoring system for the underground environment of a cable joint according to the present invention; Figure 3 This is a schematic diagram of the real temperature rise profile reconstruction based on deconvolution according to the present invention; Figure 4 This is a schematic diagram verifying the environmental ventilation interference suppression effect of the present invention; Figure 5 This is a schematic diagram illustrating the evolution trend of the health index of the cable joint according to the present invention.

[0009] In the diagram: Unified Time Grid Module 201, Original Full Width Half Height Calculation Module 202, True Full Width Half Height Calculation Module 203, Axial Thermal Dispersion Half Height Calculation Module 204, Weighted Half Height Calculation Module 205, Cable Health Index Generation Module 206. Detailed Implementation

[0010] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0011] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of the present invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in one or more embodiments of the present invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" indicate that the element or object preceding the term encompasses the elements or objects listed following the term and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0012] like Figures 1-5 As shown, the method for comprehensive monitoring of the downhole environment at the cable joint provided by the embodiments of this application includes the following steps: Step S101: Obtain a temperature profile through distributed optical fiber temperature measurement, combine it with the ambient temperature to form a temperature rise profile, collect the time and energy sequence of partial discharge events, and resample to form a unified time grid. Step S102: Calculate the original full width and half height based on the temperature rise profile under a unified time grid. Step S103: Convert the original full width half height to the true full width half height based on the point spread function and the instrument space half height. Step S104: The outer surface of the cable is taken as the object of forced convection of the outflow cylinder. The air physical parameters are determined according to the wind speed, relative humidity and ambient temperature. The convective heat transfer coefficient is calculated by the correlation formula of forced convection of the outflow cylinder. Based on this, the true full width half height is normalized to obtain the axial heat dissipation half height of the joint. Step S105: The partial discharge event time and energy sequence is low-pass processed by smoothing kernel and weighted window to obtain partial discharge energy low-pass sequence, and the weighted half width of the joint axial thermal dispersion is time-weighted using this sequence as weight to obtain weighted half width of the joint. Step S106: Select a reference subset based on the partial discharge energy low-pass sequence within the rolling statistical interval, determine the relative bandwidth by taking the median of the joint axial thermal dissipation half-width within the reference subset and the absolute deviation of the median, and generate the cable health index through exponential mapping.

[0013] It should be noted that the sampling rates of different monitoring data vary significantly, and the time bases are not unified, making it impossible to directly conduct multi-source data fusion analysis; distributed fiber optic temperature measurement equipment has an inherent spatial measurement ambiguity effect, resulting in distortion of the temperature rise profile peak parameters, making it difficult to reflect the true thermal dispersion state of the joint; dynamic fluctuations in underground ventilation conditions cause changes in the convective heat transfer coefficient of the cable surface, and there is a lack of a unified comparison benchmark for thermal dispersion parameters under different conditions; partial discharge events and joint thermal dispersion characteristics have not been correlated in the time domain, making it impossible to reflect the impact of discharge energy on the thermal state; existing monitoring lacks quantitative health indicators, making it difficult to distinguish between normal fluctuations and early degradation, and fault warnings are not timely. This invention first obtains a temperature profile through distributed optical fiber temperature measurement, constructs a temperature rise profile by combining it with ambient temperature, collects and resamples partial discharge-related sequences to form a unified time grid, and solves the problem of inconsistent data time references. Based on the temperature rise profile under the unified time grid, the original full width at half height (WHM) is calculated, providing a foundation for subsequent acquisition of real thermal dispersion parameters. The original WHM is converted to the real WHM using a point spread function convolution model, eliminating the spatial ambiguity effect of the instrument. The convective heat transfer coefficient is calculated through an outflow cylindrical forced convection model, and the real WHM is scaled to eliminate the influence of ventilation conditions. The temporal correlation between partial discharge energy and thermal dispersion parameters is established through smoothing kernel and weighted window processing to obtain the weighted WHM width. A reference subset is selected within the rolling statistical interval, and the relative bandwidth is determined through robust statistical methods. An exponential mapping is used to generate a cable health index, realizing a quantitative assessment of the health status. Ultimately, the system achieves a unified time reference for multi-source monitoring data, laying the foundation for fusion analysis; it eliminates the spatial measurement ambiguity effect of instruments, restoring the true thermal dispersion characteristics of the joints; it eliminates the interference of ventilation condition fluctuations on thermal parameters, making thermal dispersion parameters comparable across different time periods; it establishes a temporal correlation between partial discharge energy and thermal dispersion characteristics, highlighting the impact of discharge energy on thermal status; and it generates a quantitative cable health index, clearly presenting the cable health status and degradation trend, improving the timeliness and reliability of condition monitoring, and providing an intuitive basis for operation and maintenance decisions.

[0014] It should be noted that the acquisition of the time and energy sequence of partial discharge events through ultra-high frequency (UHF) partial discharge monitoring and acoustic emission (AE) partial discharge monitoring is a composite monitoring technology for partial discharge phenomena in cables, integrating the two technical approaches. UHF partial discharge monitoring uses dedicated sensors to capture the ultra-high frequency electromagnetic wave signals generated when partial discharge occurs inside the cable, thereby determining the time and corresponding energy of the discharge. Acoustic emission partial discharge monitoring uses acoustic sensors to receive the mechanical stress waves, i.e., acoustic emission signals, generated during partial discharge, thus capturing the discharge event. While this technology can acquire the time and energy sequence of cable partial discharge events, in downhole scenarios, its sampling rate differs significantly from that of distributed fiber optic temperature measurement data, and the time references for various data types are inconsistent, making direct multi-source data fusion analysis impossible.

[0015] It should be noted that the inherent spatial measurement ambiguity effect is a characteristic of distributed fiber optic temperature measurement equipment, determined by factors such as the equipment's structure and performance limitations. This characteristic restricts the spatial resolution of the cable temperature rise profile data measured by the equipment. The temperature distribution corresponding to heat sources that were originally concentrated in a specific area is blurred by the equipment's own characteristics, resulting in distortion of the peak-related parameters of the temperature rise profile. This makes it impossible to objectively reflect the true heat dissipation state of the cable joint, hindering the accurate determination of the thermal state of the cable joint in underground mines.

[0016] In one embodiment of the present invention, a temperature profile is obtained by distributed optical fiber temperature measurement, which is combined with the ambient temperature to form a temperature rise profile. The time and energy sequence of partial discharge events are collected and resampled to form a unified time grid. The process includes: constructing a set of discrete time points generated by recursion of the start time and time interval as a unified time grid; and resampling the temperature profile obtained by distributed optical fiber temperature measurement and the ambient temperature to the unified time grid according to the nearest timestamp and interpolation weight. The temperature rise profile is constructed by calculating the difference between the resampled temperature profile and the ambient temperature. The partial discharge event time and energy sequence is converted into pulse energy density, and the pulse energy density is integrated and aggregated in the time domain within the time interval defined by a unified time grid to form a partial discharge energy sequence aligned with the unified time grid.

[0017] It should be noted that in traditional monitoring, there are significant differences in the sampling rates of temperature profiles, ambient temperatures, and partial discharge event sequences from distributed fiber optic temperature measurement, resulting in inconsistent time bases and making it impossible to directly fuse and analyze multi-source data. Furthermore, the temperature profiles do not eliminate interference from ambient temperatures, making it difficult to reflect the true temperature rise of the cable joint itself. Partial discharge events are discrete instantaneous signals and lack a continuous energy characterization form that is time-aligned with other monitoring data, thus making them unsuitable for subsequent multi-source feature correlation analysis. First, a set of discrete time points recursively derived from the start time and fixed time intervals is constructed as a unified time grid, providing a unified time reference standard for all monitoring data. Next, based on adjacent timestamps and interpolation weights, temperature profiles at different sampling rates and ambient temperatures are resampled to this unified time grid to eliminate time reference differences. Then, by calculating the difference between the resampled temperature profile and the ambient temperature, ambient temperature interference is eliminated, forming a temperature rise profile that reflects the self-heating of the cable joint. Finally, the discrete partial discharge event times and energy sequences are converted into pulse energy densities, and time-domain integration and aggregation are performed within the time intervals of the unified time grid to obtain a partial discharge energy sequence time-aligned with other data. This completes the time unification and effective characterization of multi-source data, laying the foundation for subsequent calculations. Ultimately, the time reference of multi-source monitoring data is unified, enabling temperature profiles, ambient temperature, and partial discharge-related data to be analyzed collaboratively within the same time frame; the influence of ambient temperature on cable temperature is removed, resulting in a temperature rise profile that reflects the true heating state of the cable joint; discrete partial discharge events are transformed into continuous energy sequences aligned with a unified time grid, providing a consistent data foundation for subsequent correlation calculations of thermal parameters and partial discharge energy, and ensuring the feasibility of multi-source data fusion analysis.

[0018] It should be noted that the start time represents the first time node of the unified time grid, serving as the initial anchor point for the entire time reference and used to determine the starting point of the time grid; the time interval represents the fixed time length between two adjacent discrete time points in the unified time grid, used to ensure the uniformity of the time grid. The set of discrete time points represents a set of equally spaced time nodes recursively generated from the start time and time interval. The nearest timestamp represents the original data timestamp in the distributed fiber optic temperature measurement original temperature profile or original ambient temperature data that is closest to the target time point of the unified time grid, serving as the basic reference point for resampling. The interpolation weight represents the weight value assigned to the original data corresponding to each nearest timestamp during the resampling process. Its function is to ensure the continuity and accuracy of the data after resampling, and it must meet the normalization requirement that the sum of the weights equals one. The distributed fiber optic temperature measurement temperature profile represents the set of temperature data at different original timestamps at different locations along the cable axis, containing temperature information in both spatial and temporal dimensions.

[0019] It should be noted that ambient temperature represents the temperature of the environmental medium within the monitoring area obtained through environmental monitoring equipment, serving as the baseline data to eliminate ambient temperature interference. Temperature rise profile represents the set of data showing the difference between the cable's temperature along the cable after resampling and the ambient temperature at the corresponding time point, reflecting the actual temperature rise of the cable itself. Partial discharge event time represents the specific time point at which the partial discharge phenomenon occurs. Partial discharge event energy represents the energy magnitude of each corresponding partial discharge event, reflecting the intensity of a single partial discharge. Pulse energy density represents the energy distribution characteristic of converting discrete partial discharge events into a continuous time domain, used to represent the distribution state of partial discharge energy over time. The unified time grid time interval is the aforementioned time interval, which is the time length of adjacent time points in the unified time grid, used to define the time range for partial discharge energy aggregation. The unified time grid aligned partial discharge energy sequence represents the energy data set obtained after integrating and aggregating the pulse energy density within each time interval of the unified time grid, with its time points completely consistent with the unified time grid.

[0020] Specifically, partial discharge pulse energy density The calculation formula is as follows: Where N represents the total number of partial discharge events within a uniform time grid. This represents the energy of the i-th partial discharge event. Represents the Dirac function, t represents the occurrence time of the i-th partial discharge event, and t represents the current time within the unified time grid.

[0021] Specifically, partial discharge energy sequences aligned with a uniform time grid. The calculation formula is as follows: in Indicates the time interval of a uniform time grid. Indicates the first Partial discharge pulse energy density at a given moment. Let represent the integral variable, and t represent the current time within the unified time grid.

[0022] It should be noted that the interpolation weights are calculated as follows: First, the time difference between the target time point and two adjacent time stamps is determined. The reciprocal of the two time differences is used as the basic weight. Then, the basic weights are normalized. That is, the interpolation weight of each adjacent time stamp is equal to its corresponding basic weight divided by the sum of the two basic weights. Finally, the sum of the interpolation weights of two adjacent time stamps is guaranteed to be one. If multiple adjacent time stamps are involved, the reciprocal of the time difference between each time stamp and the target time point is used as the basic weight, and then normalized. The pulse energy density is calculated based on the occurrence time of the partial discharge event. At the time of the event, the pulse energy density value is equal to the energy of the event. At other times, the pulse energy density value is zero, thus forming a discrete pulsed energy density distribution. The specific calculation method for the partial discharge energy sequence aligned with the unified time grid is as follows: for each time point of the unified time grid, take the duration of half the time interval of the unified time grid before and after the time point as the center, and calculate the cumulative value of the pulse energy density within the interval. This cumulative value is the partial discharge energy sequence value corresponding to that time point.

[0023] It should be noted that in the underground cable well environment, the original sampling rate of distributed fiber optic temperature measurement data differs significantly from that of partial discharge and ambient temperature. Using a combination of nearest timestamps and normalized interpolation weights allows for time axis alignment with other monitoring data without disrupting the spatial characteristics of the temperature profile, while also meeting the data loss compensation requirements of the complex electromagnetic environment underground. Furthermore, during resampling, the two nearest timestamps before and after the target time point are selected first. The time difference between the target time point and these two nearest timestamps is calculated, and the reciprocal of this time difference is used as the base weight. The two base weights are then divided by their sum to obtain the interpolation weights corresponding to the two nearest timestamps. This weight naturally satisfies the normalization requirement. If there are missing original data, this can be extended to selecting three nearest timestamps before and after the target time point, similarly using the reciprocal of the time difference as the base weight and normalizing.

[0024] In one embodiment of the present invention, the original full width half height is calculated based on the temperature rise profile under a unified time grid, including: selecting the current moment within the unified time grid, taking the joint center position as a reference, and extracting temperature rise profile data within the analysis half window of a preset length. The search analysis identifies the maximum temperature rise within the half-window range as the peak temperature rise, and half of the peak temperature rise is taken as the half-height line height.

[0025] It should be noted that the temperature rise profile analysis lacks a clear definition of its effective range, making it prone to including interference data from areas unrelated to the joint; the lack of a unified half-height line determination standard makes it impossible to stably extract parameters characterizing the width of the joint temperature rise peak; and the key reference benchmarks for the temperature rise peak shape are inconsistent at different times or under different operating conditions, making it difficult to form comparable thermal state characteristic parameters, affecting the accurate assessment of subsequent thermal dispersion state. First, the current time is selected within a unified time grid to ensure that the analysis time benchmark is consistent with other monitoring data; a preset length of analysis half-window is set with the joint center position as the core to limit the effective analysis range of the temperature rise profile, ensuring that the extracted data all revolve around the joint's thermally affected area and avoiding interference from data in unrelated areas; then, the maximum temperature rise value in the temperature rise data is searched within this analysis half-window range and determined as the peak temperature rise, directly reflecting the highest heating intensity of the joint at that moment; finally, the peak temperature rise value is halved to obtain the half-height line height, establishing a unified and stable determination standard for subsequently locating the intersection of the left and right half-heights of the temperature rise peak shape and calculating the original full width half-height, completing the basic preparatory work for the calculation of the original full width half-height. By setting a pre-defined half-window for analysis, the effective temperature rise data around the center of the joint is focused, and interference from irrelevant areas is eliminated to ensure the relevance of the analysis object. A half-height line with half of the peak temperature rise as the standard is established to form a unified judgment benchmark and ensure the consistency of peak shape parameter extraction. The basic parameters characterizing the peak width of the joint temperature rise at the current moment are successfully obtained, providing a reliable premise for the subsequent complete calculation of the original full width and half height, so that the temperature rise peak shape characteristics have a basis for cross-time comparison.

[0026] It should be noted that the current moment represents any time segment within the unified time grid used to calculate the corresponding original full width and half height; the joint center position represents the axial spatial coordinates corresponding to the geometric center of the cable joint. The preset length of the analysis half-window represents the axial spatial analysis range formed by extending a fixed length to the left and right sides based on the joint center position, used to define the effective analysis interval of the temperature rise profile. The peak temperature rise represents the maximum value in the temperature rise profile data within the analysis half-window range. The half-height line height represents half of the peak temperature rise value and is the elevation reference for locating the half-height position of the temperature rise profile. The preset length of the analysis half-window needs to be determined in conjunction with the length of the cable joint body and the spatial resolution of the distributed fiber optic temperature measurement. It is usually taken as 1.5 to 2 times the length of the joint body, and not less than 5 times the spatial resolution of the distributed fiber optic temperature measurement, to ensure that the analysis half-window can completely cover the heat-affected zone of the joint.

[0027] In one embodiment of the present invention, the original full width half height is calculated based on the temperature rise profile under a unified time grid, including: determining two adjacent sampling positions to the left of the joint center position, where the temperature rise value is greater than and less than the half height line height respectively; constructing a linear relationship between the coordinates of these two adjacent sampling positions and the temperature rise value; and calculating the spatial coordinates when the temperature rise value is equal to the half height line height as the left half height intersection position. To the right of the center of the joint, determine two adjacent sampling positions where the temperature rise is greater than and less than the half-height line height, respectively. Based on the coordinates of these two adjacent sampling positions and the temperature rise value, establish a linear relationship and calculate the spatial coordinates when the temperature rise value is equal to the half-height line height as the right half-height intersection point. Calculate the absolute value of the difference between the intersection point of the right half-height and the intersection point of the left half-height to obtain the original full width and half-height at the current moment.

[0028] It should be noted that the sampling locations of the temperature rise profile are discretely distributed, and there are no sampling points where the temperature rise value is exactly equal to the half-height line height, making it impossible to directly obtain the half-height intersection location; there is a lack of a unified half-height intersection location method, and the judgment criteria for the intersection points on the left and right sides are inconsistent, making the calculation results of the parameters characterizing the temperature rise peak width unstable; it is impossible to accurately convert discrete sampling data into continuous peak width characteristics, affecting the consistency of subsequent thermal dispersion state assessment. Based on the half-height line height determined above, the temperature rise profile data on the left and right sides are analyzed separately, with the center of the joint as the dividing line. In the discrete sampling positions on the left, two adjacent sampling positions with temperature rise values ​​greater than and less than the half-height line height are selected. Using the coordinates of these two positions and the corresponding temperature rise values, a linear relationship between the temperature rise value and the spatial coordinates is constructed. This relationship is used to deduce the spatial coordinates when the temperature rise value equals the half-height line height, which is the left half-height intersection point. The same method is used on the right side, selecting adjacent sampling positions and constructing a linear relationship to calculate the right half-height intersection point. Finally, by calculating the absolute value of the difference between the right and left half-height intersection points, the discrete sampling data is transformed into a continuous peak width parameter, ultimately obtaining the original full-width half-height at the current moment. Finally, a unified half-height intersection positioning method was established to accurately calculate the positions of the half-height intersections on the left and right sides; stable and physically meaningful original full-width half-height parameters were obtained to accurately characterize the peak width of the joint temperature rise profile at the current moment; reliable basic data were provided for subsequent removal of instrument spatial ambiguity effects and acquisition of real heat dissipation characteristics, ensuring the continuity and consistency of thermal parameter calculations throughout the entire monitoring process.

[0029] It should be noted that the adjacent sampling positions to the left of the joint center represent two spatially adjacent temperature rise sampling points to the left of the joint center, with temperature rise values ​​greater than and less than the half-height line height, respectively. These are the basic reference points for calculating the left half-height intersection point. The left half-height intersection point represents the axial spatial coordinate corresponding to the temperature rise value exactly equal to the half-height line height after establishing a linear relationship through the adjacent sampling positions to the left. This is the left boundary point for calculating the original full-width half-height. The adjacent sampling positions to the right of the joint center represent two spatially adjacent temperature rise sampling points to the right of the joint center, with temperature rise values ​​greater than and less than the half-height line height, respectively. These are the basic reference points for calculating the right half-height intersection point. The right half-height intersection point represents the axial spatial coordinate corresponding to the temperature rise value exactly equal to the half-height line height after establishing a linear relationship through the adjacent sampling positions to the right. This is the right boundary point for calculating the original full-width half-height. The original full-width half-height represents the absolute value of the spatial distance between the left and right half-height intersection points, and is a shape parameter characterizing the peak width of the joint temperature rise profile at that moment.

[0030] Specifically, the location of the intersection of the temperature rise profile and the left half-height line. The calculation formula is as follows: in and Indicates the adjacent sampling position to the left of the connector center (satisfying) , (The center position of the connector). Indicates the half-height line height. and They represent and Temperature rise at time t.

[0031] It should be noted that when selecting adjacent sampling locations, priority should be given to selecting two adjacent sampling points that are closest to the center of the joint on one side and whose temperature rise values ​​exactly span the half-height line height. If multiple sampling point pairs meet the conditions, the pair closest to the joint center should be selected for calculation. The linear relationship is constructed with the sampling location as the abscissa and the corresponding temperature rise value as the ordinate. Based on the geometric principle that two points determine a straight line, a correlation model is established for the linear change of temperature rise value with spatial coordinates. Then, the spatial coordinates corresponding to the half-height line height are deduced from this model, which is the half-height intersection point.

[0032] In one embodiment of the present invention, step S103, converting the original full width half-height to the true full width half-height based on the point spread function and the instrument space half-height width, includes: establishing a convolution model with the point spread function as the convolution kernel, wherein the spatial scale parameter of the point spread function is determined by the instrument space half-height width. The spread function at the set point conforms to a Gaussian distribution. The difference between the square of the original full width and half height and the square of the instrument space half height is calculated, and the arithmetic square root of this difference is performed to obtain the true full width and half height.

[0033] It should be noted that distributed fiber optic temperature measurement equipment suffers from an inherent spatial measurement ambiguity effect. This effect distorts the measured temperature rise profile peak parameters, causing the original full width at half maximum (FWHM) to fail to objectively reflect the true thermal dispersion state of the cable joint. This, in turn, interferes with subsequent cable condition assessments based on thermal characteristic parameters, failing to provide a reliable basis for accurately judging cable health status. First, a convolution model using the point spread function as the convolution kernel is established, clarifying that the spatial scale parameter of the point spread function is determined by the instrument's spatial FWHM, thus quantifying the spatial measurement ambiguity effect of the equipment into the model. Next, the point spread function is set to conform to a Gaussian distribution, which is suitable for the spatial response characteristics of distributed fiber optic temperature measurement equipment, enabling reasonable conversion of the FWHM. Finally, through mathematical calculations, the difference between the square of the original full width at half maximum (FWHM) and the square of the instrument's spatial FWHM is calculated, and then the arithmetic square root of this difference is applied to offset the influence of the instrument ambiguity effect, ultimately obtaining the true full width at half maximum (FWHM) that reflects the true thermal dispersion state of the cable joint. Finally, the spatial measurement ambiguity effect of the distributed fiber optic temperature measurement equipment was eliminated, and the interference of the equipment's own characteristics on the thermal parameter measurement was removed, so as to obtain the true full width and half height that can truly characterize the thermal dispersion state of the cable joint. This parameter provides accurate basic data for the subsequent thermal parameter scale normalization under different ventilation conditions, ensuring the reliability of thermal characteristic analysis in the entire monitoring process, and making the subsequent evaluation of cable condition more in line with the actual situation.

[0034] It should be noted that the point spread function (PSF) represents the temperature response distribution pattern of a distributed fiber optic temperature measurement device to a single point heat source in space; the convolution kernel represents the core function used in the convolution model for convolution with the true temperature rise profile; the spatial scale parameter of the PSF represents the key parameter determining the spatial distribution range of the PSF, and its value is uniquely determined by the instrument's spatial half-width at half-maximum (HWHM), directly related to the spatial measurement resolution of the device. The instrument spatial HWHM represents the inherent spatial HWHM index of the distributed fiber optic temperature measurement device itself, a fixed parameter obtained from factory or field calibration, characterizing the degree of spatial measurement ambiguity of the device. The Gaussian distribution shape represents the specific mathematical distribution form adopted by the PSF. This shape has symmetrical and continuous characteristics, adapting to the spatial response characteristics of the distributed fiber optic temperature measurement device, and enabling accurate HWHM conversion. The true full-width HWHM represents the true full-width HWHM parameter of the temperature rise profile in the cable joint area after removing the instrument spatial ambiguity effect, and is a physical quantity reflecting the true heat dissipation characteristics of the joint.

[0035] Specifically, the temperature rise profile along the cable measured by the instrument. The calculation formula is as follows: in This represents the actual temperature rise profile along the cable route. Represents the integral variable. Represents the diffusion function. The value represents the half-height and width of the instrument space, and x represents the axial coordinate of the cable.

[0036] It should be noted that the underground cable monitoring environment is subject to strong electromagnetic interference and spatial limitations. The spatial measurement ambiguity effect of distributed fiber optic temperature measurement equipment can obscure the true thermal dispersion characteristics of the joint. The point spread function convolution model can quantify and remove the inherent ambiguity effect from the measured data, ensuring the accuracy of the true thermal characteristic parameters. The spatial scale parameter of the point spread function is equal to half the instrument's spatial half-width. That is, once the instrument's spatial half-width is determined, halving its value yields the spatial scale parameter of the point spread function. The instrument's spatial half-width needs to be obtained through on-site calibration. Specifically, a uniform heat source segment of known length is selected in the non-joint section of the cable. The half-width of this heat source segment is measured using the distributed fiber optic temperature measurement equipment. This half-width is then calculated by combining it with the actual length of the heat source segment. After calibration, the value is permanently stored in the system. The convolution model is only applicable to scenarios where the spatial resolution of the distributed fiber optic temperature measurement equipment is stable and there is no significant electromagnetic interference in the measurement area causing data distortion. If the equipment resolution fluctuates during monitoring, the instrument's spatial half-width needs to be recalibrated and the convolution model parameters updated.

[0037] In one embodiment of the present invention, the outer surface of the cable is used as the object of forced convection in the outflow cylinder, and the air physical properties are determined according to the wind speed, relative humidity and ambient temperature, including: determining the air density, air dynamic viscosity, air thermal conductivity and air specific heat at constant pressure according to the ambient temperature and relative humidity, and calculating the air kinematic viscosity, air thermal diffusivity and Prandtl number accordingly.

[0038] It should be noted that the temperature and humidity of the underground environment fluctuate dynamically with factors such as production shifts and seasons, causing changes in air physical properties. If fixed air physical properties are used for convective heat transfer calculations, the impact of environmental changes on the thermophysical properties of the air will be ignored, resulting in deviations in the subsequent calculations of convective heat transfer coefficients and heat dissipation parameters. This will affect the consistency and accuracy of heat dissipation characteristic assessments under different operating conditions, and will fail to provide reliable data support for cross-time period condition monitoring. First, it is clear that the dynamic fluctuations in temperature and humidity in the downhole environment directly alter the physical properties of the air. Fixed physical property parameters cannot adapt to actual working conditions, and parameters must be determined based on real-time environmental data. Real-time ambient temperature and relative humidity are obtained through environmental monitoring equipment. Based on the laws of thermophysical characteristics, fundamental physical property parameters such as air density, aerodynamic viscosity, air thermal conductivity, and air specific heat at constant pressure are determined from these two key environmental parameters. Then, based on the inherent mathematical relationship between the fundamental physical property parameters, the ratio of aerodynamic viscosity to air density is taken as the kinematic viscosity of air, the ratio of air thermal conductivity to the product of air density and specific heat at constant pressure is taken as the thermal diffusivity of air, and the ratio of kinematic viscosity to thermal diffusivity is taken as the Prandtl number. The derivative physical property parameters are calculated to provide complete and accurate parameter support for the subsequent solution of the convective heat transfer coefficient. Ultimately, the system achieves dynamic and accurate acquisition of air physical parameters, fully adapting to the fluctuating temperature and humidity characteristics of the downhole environment. The determined basic physical parameters and derived parameters provide a reliable basis for the accurate calculation of Reynolds number, Nusselt number, and convective heat transfer coefficient. It ensures the scientific nature of convective heat transfer related calculations, creates conditions for the scaling of thermal parameters under different ventilation conditions, and provides a basis for cross-condition comparison of heat dissipation characteristic parameters.

[0039] It should be noted that air density represents the mass of a unit volume of air under corresponding ambient temperature and humidity, and is a fundamental parameter characterizing air properties; aerodynamic viscosity represents the coefficient of internal friction that hinders the relative motion of adjacent air layers, and is a physical property parameter measuring air viscosity; air thermal conductivity represents a quantitative index of air's ability to conduct heat, with a higher value indicating stronger thermal conductivity; air specific heat at constant pressure represents the amount of heat required to raise the temperature per unit mass of air at a constant pressure; air kinematic viscosity represents the ratio of air dynamic viscosity to air density; air thermal diffusivity represents the ratio of air thermal conductivity to the product of air density and air specific heat at constant pressure, reflecting the rate of heat diffusion in air; Prandtl number represents the ratio of air kinematic viscosity to thermal diffusivity, and is a dimensionless number characterizing the relative relationship between air momentum diffusion and heat diffusion.

[0040] In one embodiment of the present invention, the convective heat transfer coefficient is calculated using the external cylindrical forced convection correlation formula, and the true full width at half height (WHM) is normalized accordingly to obtain the axial heat dissipation WHM of the joint. This includes: using the outer surface of the cable as the boundary of the external cylindrical forced convection model, calculating the Reynolds number based on air density, wind speed, cable outer diameter, and aerodynamic viscosity; calculating the Nusselt number based on the Reynolds number and Prandtl number using the external cylindrical forced convection correlation formula containing empirical coefficients; and calculating the convective heat transfer coefficient at the current moment based on the Nusselt number, air thermal conductivity, and cable outer diameter. Set a reference convective heat transfer coefficient, calculate the ratio of the reference convective heat transfer coefficient to the convective heat transfer coefficient at the current moment to form a scale factor; calculate the product of the scale factor and the true full width at half height to obtain the axial heat dissipation half width at half height of the joint.

[0041] It should be noted that the intensity of underground ventilation varies greatly with factors such as production shifts and seasons, causing dynamic fluctuations in the convective heat transfer coefficient of the cable surface. The true full width and half height under different operating conditions are affected by the convective heat transfer efficiency, lacking a unified comparison benchmark, making it impossible to effectively compare the heat dissipation characteristics across time periods and operating conditions. Directly using the true full width and half height for condition assessment can easily misjudge parameter fluctuations caused by ventilation changes as cable degradation signals, interfering with the accuracy of condition monitoring. First, the outer surface of the cable is used as the boundary of the external cylindrical forced convection model. Combining the determined air density, aerodynamic viscosity, measured wind speed, and cable outer diameter, the Reynolds number is calculated using fluid dynamics formulas to characterize the flow state of air around the cable. Then, combined with the Prandtl number obtained earlier, and based on empirical coefficients matching the Reynolds number range, the Nusselt number is calculated using the external cylindrical forced convection correlation formula to quantify the convective heat transfer intensity. Subsequently, based on the Nusselt number, air thermal conductivity, and cable outer diameter, the convective heat transfer coefficient at the current moment is calculated. A reference convective heat transfer coefficient is set as a unified benchmark, and its ratio with the current convective heat transfer coefficient is calculated to obtain the scaling factor. Finally, the scaling factor is multiplied by the true full width at half maximum (FWHM) to complete the scaling normalization of the thermal parameters, ultimately obtaining the axial heat dissipation FWHM of the joint. Finally, the convective heat transfer coefficient under the current operating conditions is calculated to achieve scale normalization of the true full width and half height; the interference of ventilation condition fluctuations on heat dispersion parameters is eliminated, and the axial heat dispersion half height and width of the joint with a unified comparison benchmark are obtained; the heat dispersion characteristic parameters under different ventilation conditions at different times can be directly compared, providing stable and reliable basic data for subsequent health status assessment, and improving the consistency and credibility of condition monitoring.

[0042] Specifically, Nusselt numbers The calculation formula is as follows: Where C, m, and n represent empirical coefficients that match the Reynolds number interval. Represents the Reynolds number. This represents the Prandtl number.

[0043] Specifically, convective heat transfer coefficient The calculation formula is as follows: in This represents the thermal conductivity of air at the current moment. This represents the Nusselt number at the current moment, and D represents the outer diameter of the cable.

[0044] It should be noted that the external cylindrical forced convection model represents a physical model that equates the outer surface of the cable to a cylindrical boundary, describing the convective heat transfer law when external airflow passes over the cable surface. This model is suitable for ventilation and heat transfer scenarios in cable wells. The cable outer diameter represents the overall external diameter of the cable; the Reynolds number represents the flow state when air flows around the cable, used to determine whether the flow is laminar or turbulent. The empirical coefficient represents a constant matching the Reynolds number range and is a key parameter in the external cylindrical forced convection correlation equation; its value is obtained by fitting a large amount of experimental data. The Nusselt number represents the dimensionless number of the convective heat transfer intensity on the cable surface; a larger value indicates stronger convective heat transfer capacity. The convective heat transfer coefficient represents the heat transferred per unit area per unit time between the air and the cable surface under a unit temperature difference, and is a key indicator for quantifying the intensity of convective heat transfer. The reference convective heat transfer coefficient represents a fixed reference value set during system deployment, serving as a benchmark to eliminate differences in convective heat transfer under different operating conditions and ensuring uniformity in scale. The scaling factor represents the ratio of the reference convective heat transfer coefficient to the current convective heat transfer coefficient, used to eliminate the influence of ventilation. The axial heat dissipation half-width of the joint represents the true full width and half-height after convective condition scaling.

[0045] It should be noted that underground cables are located in narrow, confined spaces, and the ventilation airflow is mostly a forced flow that skims parallel across the cable surface. Equating this to an outward-flowing cylinder accurately characterizes the heat transfer boundary between the airflow and the cable. Furthermore, this model can integrate underground monitoring data such as ambient temperature, humidity, and wind speed to achieve real-time quantification of heat transfer intensity. The dynamic fluctuations in underground ambient temperature and humidity directly affect air properties. Through the derivation of a complete property parameter chain, dynamic correction of the convective heat transfer coefficient can be achieved, avoiding calculation deviations caused by fixed properties and ensuring the accuracy of heat dissipation characteristic parameters. In addition, the underground ventilation intensity varies significantly with production shifts and seasons, leading to fluctuations in the convective heat transfer coefficient on the cable surface, which in turn interferes with the determination of the heat dissipation half-width at half-maximum (HWHM). By constructing a scaling factor based on the convective heat transfer coefficient, the true HWHM under different operating conditions can be normalized to a unified benchmark, enabling direct comparison of characteristic parameters across time periods and operating conditions.

[0046] It should be noted that the values ​​of the empirical coefficients must strictly match the Reynolds number range. When the Reynolds number is between 1,000 and 10,000, the coefficient C is 0.683, the coefficient m is 0.466, and the coefficient n is 0.333; when the Reynolds number is between 10,000 and 100,000, the coefficient C is 0.076, the coefficient m is 0.7, and the coefficient n is 0.333. The system must first calculate the Reynolds number and then automatically match the corresponding coefficients. In addition, the reference convective heat transfer coefficient needs to be calibrated on-site at the beginning of system deployment. The specific method is to select a period of stable downhole ventilation, continuously monitor the convective heat transfer coefficient for one week, and take the median of the convective heat transfer coefficient during this period as the reference convective heat transfer coefficient. After calibration, the value is fixed in the system and only needs to be recalibrated after the downhole ventilation system is modified.

[0047] In one embodiment of the present invention, a partial discharge energy low-pass sequence is obtained by performing low-pass processing on the partial discharge event time and energy sequence using a smoothing kernel and a weighted window, and a weighted half-width at half maximum (WHM) of the joint axial thermal dispersion is obtained by using this sequence as a weight. The process includes: obtaining the partial discharge energy sequence and the WHM of the joint axial thermal dispersion under a uniform time grid; and performing a time-domain convolution operation on the partial discharge energy sequence using a smoothing kernel that satisfies the conditions of nonnegativity and integral normalization to generate a partial discharge energy low-pass sequence. Within a weighted window of a preset duration, the time integral of the product of the joint axial thermal dispersion half-width and the partial discharge energy low-pass sequence is calculated and used as the numerator. Within the weighted window, the time integral value of the partial discharge energy low-pass sequence is calculated and added to a preset minimum positive number as the denominator; the ratio of the numerator to the denominator is calculated to obtain the weighted half-width.

[0048] It should be noted that partial discharge events are discrete instantaneous pulse signals, while joint thermal dispersion is a continuous, slowly varying process. Existing methods do not establish a time-domain correlation between the two types of characteristics, and cannot reflect the impact of discharge energy on the thermal state. The discrete partial discharge energy sequence cannot be directly used as a weight for collaborative analysis with the continuous joint axial thermal dispersion half-width. The lack of effective data processing methods makes it difficult to highlight the abnormal characteristics of thermal dispersion related to discharge energy, affecting the accurate capture of changes in the state of cable joints. First, aligned partial discharge energy sequences and the full width at half maximum (FWHM) of joint axial thermal dispersion under a unified time grid are obtained to ensure temporal consistency between the two types of data. A smoothing kernel satisfying non-negativity and integral normalization conditions is selected to perform temporal convolution on the discrete partial discharge energy sequences, filtering out instantaneous pulse interference and generating a continuous partial discharge energy low-pass sequence that reflects energy change trends. A weighted window of preset duration is set, and the time integral value of the product of the joint axial thermal dispersion FWHM and the partial discharge energy low-pass sequence is calculated within the window, serving as the numerator. Simultaneously, the time integral value of the partial discharge energy low-pass sequence is calculated within the same window and added to a preset minimum positive number as the denominator to avoid a zero denominator. Finally, by calculating the ratio of the numerator to the denominator, the weighted FWHM with discharge energy as the weight is obtained, completing the temporal fusion of the two types of features. Finally, a time-domain correlation between partial discharge energy and joint thermal dispersion characteristics was established, transforming discrete discharge signals into a continuous low-pass sequence of partial discharge energy. Through time-weighted calculation, a weighted half-width at half maximum (WHM) that reflects the impact of discharge energy was obtained, highlighting thermal dispersion anomalies related to discharge. This provides key parameters that integrate the two core characteristics for the subsequent generation of cable health index, improving the sensitivity and correlation of condition monitoring to early degradation.

[0049] It should be noted that the partial discharge pulse energy density represents the pulsed energy distribution characteristic transformed from discrete partial discharge events under a unified time grid. A smoothing kernel is used for temporal convolution of the partial discharge pulse energy density. Non-negativity requires that the smoothing kernel value is not less than zero throughout the entire time domain. The integral normalization condition requires that the integral value of the smoothing kernel be one throughout the entire time domain, ensuring the conservation of the total partial discharge energy before and after processing and preventing the introduction of additional energy deviation. The temporal convolution operation is used to convolve the smoothing kernel and the partial discharge pulse energy density in the time dimension. The partial discharge energy low-pass sequence represents the continuous partial discharge energy sequence obtained after temporal convolution with the smoothing kernel, which can be used as a weight signal for weighting the joint axial thermal dispersion half-width at half-maximum. The preset duration weighting window represents a fixed time length used to limit the time weighting range, serving as the time boundary for realizing the temporal correlation between partial discharge energy and thermal dispersion characteristics. The numerator represents the time integral value of the product of the joint axial thermal dispersion half-width at half-maximum and the partial discharge energy low-pass sequence within the weighting window. The operational denominator represents the sum of the time integral of the partial discharge energy low-pass sequence within the weighted window and the minimum positive number. The minimum positive number is a small constant preset to prevent the operational denominator from being zero. The weighted half-width at half-maximum (WHM) represents the characteristic parameter obtained by time-weighting the axial thermal dispersion WHM of the joint using the partial discharge energy low-pass sequence as the weight.

[0050] Specifically, weighted half-width The calculation formula is as follows: in Indicates the weighted window duration. Indicates the first The axial heat dissipation half-width at the joint at any given moment. Indicates the first Partial discharge energy low-pass sequence at each time point Represents the integral variable. It represents a very small positive number.

[0051] It should be noted that partial discharge events are discrete instantaneous pulses, while joint thermal dispersion is a continuous, slowly varying process. By using a smoothing kernel low-pass processing method, the discrete discharge energy can be converted into continuous weights that match the time sequence of thermal dispersion characteristics, achieving accurate temporal correlation between the two types of characteristics. Simultaneously, non-negativity and integral normalization conditions ensure the accuracy of the energy data. There is a lag correlation between the changes in thermal dispersion characteristics of the cable joint and the injection of partial discharge energy. A fixed-duration weighting window can limit the time range of this lag correlation, enabling targeted weighting of discharge energy on thermal dispersion characteristics, highlighting thermal dispersion anomalies related to discharge, and improving the sensitivity of characteristic parameters to early degradation.

[0052] It should be noted that the smoothing kernel should preferably be the commonly used exponential smoothing kernel or rectangular smoothing kernel in engineering. The exponential smoothing kernel is suitable for scenarios where the intervals between downhole discharge events are uneven, and can assign higher weights to recent discharges. The rectangular smoothing kernel is suitable for scenarios where the discharge events are evenly distributed, and can achieve uniform weighting over the time range. Both types of kernel functions must be verified in advance to meet the conditions of non-negativity and integral normalization. The duration of the preset weighting window needs to be determined in conjunction with the thermal response time of the cable joint. In downhole scenarios, the thermal response time of cables is usually 10 to 30 minutes, so the weighting window duration can be set to 15 minutes. If differences in thermal response time are detected on-site, the window duration can be adjusted to 1.2 times the actual thermal response time. The value of the smallest positive number should be much smaller than the conventional partial discharge energy integral value within the weighting window. In engineering, it can be set to one-thousandth of the historical minimum energy integral value within the weighting window, and the value should not be greater than the order of ten to the power of negative six, so as to achieve the purpose of preventing zero denominator without interfering with the normal weighting calculation results.

[0053] In one embodiment of the present invention, selecting a reference subset based on the partial discharge energy low-pass sequence within a rolling statistical interval includes: within a unified time grid, taking the current time as the endpoint and a preset window width as the length as a rolling statistical interval, calculating the lower quartile values ​​of the partial discharge energy low-pass sequence within the rolling statistical interval, and taking the set of times when the partial discharge energy low-pass sequence values ​​within the rolling statistical interval are not higher than the lower quartile values ​​as a reference subset.

[0054] It should be noted that existing reference benchmark selection lacks a clear method, and the intermittent fluctuations in partial discharge energy result in the selection of reference periods including high discharge disturbances, failing to reflect the steady-state thermal dispersion characteristics of the cable joint. Furthermore, the lack of dynamically adjustable statistical intervals makes it difficult to adapt to changes in downhole operating conditions, leading to inconsistent benchmarks for subsequent health status assessments and affecting the stability and reliability of the assessment results. Firstly, within a unified time grid, a rolling statistical interval is defined with the current time as the endpoint, combined with a preset window width, ensuring sufficient and timely data samples within the interval to adapt to dynamic changes in cable status. Next, the lower quartile values ​​of the partial discharge energy low-pass sequence within this interval are statistically analyzed. These values ​​effectively distinguish between high and low discharge periods, serving as a quantitative standard for selecting steady-state periods. Finally, all times within the rolling statistical interval where the partial discharge energy low-pass sequence value is not higher than the lower quartile value are integrated into a reference subset, ensuring that the periods within this subset are in a low-discharge, low-disturbance state, accurately reflecting the steady-state thermal dispersion characteristics of the cable joint and laying a stable foundation for the calculation of relevant parameters in subsequent health assessments. Ultimately, the reference benchmark is dynamically updated through rolling statistical intervals to adapt to real-time changes in downhole operating conditions; the lower quartiles are used to accurately screen out steady-state periods with low discharge and low disturbance, forming a reliable reference subset; this provides a stable data foundation for subsequent calculations of the median and dispersion of thermal dispersion parameters, ensures the consistency of health index calculations, makes the condition assessment results more consistent with the actual operating conditions of the cable, and provides a reliable premise for distinguishing between normal fluctuations and early deterioration.

[0055] It should be noted that the rolling statistical interval represents a dynamic time interval extending backward by a preset window width from the current time, serving as the time range benchmark for rolling health status assessment. The preset window width represents a fixed time value defining the length of the rolling statistical interval, a key parameter ensuring sufficient statistical data and adapting to the cycle of changes in the downhole cable's condition. The lower quartile value represents the value at the 25th percentile after sorting the partial discharge energy low-pass sequence data within the rolling statistical interval from smallest to largest, serving as the quantitative threshold for selecting low-discharge steady-state periods. The reference subset represents the set of times within the rolling statistical interval where the partial discharge energy low-pass sequence value is no higher than the lower quartile value, representing an effective data subset for extracting the cable's steady-state thermal dissipation characteristics.

[0056] In one embodiment of the present invention, the relative bandwidth is determined by the median of the joint axial heat dispersion half-width within a reference subset and the absolute deviation of the median, and a cable health index is generated by exponential mapping. This includes: obtaining the joint axial heat dispersion half-width at each time point within the reference subset and calculating its median; calculating the median of the absolute value of the difference between the joint axial heat dispersion half-width and the median within the reference subset and multiplying the median of the absolute value by a preset scaling factor to obtain the dispersion. Calculate the ratio of the dispersion to the median, and select the larger value between this ratio and a preset fixed lower limit constant as the relative bandwidth; Calculate the ratio of the weighted half-width to the median, subtract one from the ratio to obtain the difference, and select the larger value between the difference and zero as the non-negative deviation; calculate the ratio of the natural logarithm of the second to the relative bandwidth, calculate the negative of the product of the ratio and the non-negative deviation, use the negative as the exponent to calculate the natural exponential function value, and multiply the natural exponential function value by one hundred to obtain the cable health index.

[0057] It should be noted that existing monitoring lacks unified and quantitative cable health indicators, making it difficult for maintenance personnel to intuitively judge the condition of cable joints; steady-state heat dissipation parameters exhibit natural fluctuations, and the lack of scientific methods for characterizing dispersion makes it easy to confuse normal fluctuations with abnormal degradation; relative bandwidth may be abnormally small due to small data fluctuations, leading to distorted health index calculations; and no reasonable quantitative method has been established for heat dissipation characteristics deviating from steady-state benchmarks, making it impossible to accurately reflect the degree of cable condition changes. First, obtain the axial thermal dispersion half-width at each moment within the reference subset, and calculate the median as the benchmark value for steady-state thermal dispersion characteristics. Next, calculate the absolute value of the difference between the thermal dispersion half-width and the median for each unit within the reference subset, take the median of these absolute values ​​and multiply it by a preset scaling factor to obtain the dispersion characterizing the degree of steady-state fluctuation. Then, calculate the ratio of the dispersion to the median, and select the larger value between this ratio and a preset fixed lower limit constant as the relative bandwidth to ensure calculation stability. Next, calculate the ratio of the weighted half-width to the median, subtract one, and take the larger value between this ratio and zero as the non-negative deviation to quantify the degree of deviation of the current thermal characteristics. Finally, calculate the ratio of the natural logarithm of the second value to the relative bandwidth, multiply it by the non-negative deviation, take the opposite number, use this value as the exponent to calculate the natural exponential function value, and multiply it by one hundred to finally obtain the cable health index. The final result is a quantitative cable health index ranging from 0 to 100, providing an intuitive and easy-to-understand basis for operation and maintenance. The dispersion of steady-state thermal dispersion parameters is accurately characterized through robust statistical methods to avoid interference from outliers. A fixed lower limit constant is set to ensure relative bandwidth stability and prevent abnormal jumps in the health index. The non-negative deviation of the weighted half-width relative to the steady-state benchmark is accurately quantified to effectively distinguish between normal fluctuations and early degradation, thereby improving the reliability of the condition assessment.

[0058] It should be noted that the proportionality coefficient represents a fixed constant used to convert the absolute deviation of the median into an approximate dispersion. Dispersion represents a statistical measure based on the absolute deviation of the median of the thermal dispersion half-width within the reference subset, and is a key parameter characterizing the degree of fluctuation in thermal dispersion characteristics under steady-state conditions. Relative bandwidth represents the larger of the ratio of dispersion to the median and the fixed lower limit constant, and is a scale parameter ensuring the numerical stability of the health index calculation. The fixed lower limit constant is a minimal constant preset to avoid abnormal jumps in the health index due to excessively small relative bandwidth. Non-negative deviation represents a non-negative quantified value of the deviation of the weighted half-width from the median, and is an indicator measuring the magnitude of the deviation of the current thermal dispersion characteristics from the steady-state benchmark. The cable health index represents a quantified value in the 0-100 range, transformed by exponential mapping of the non-negative deviation.

[0059] Specifically, cable health index The calculation formula is as follows: in Represents the natural logarithm of 2. Indicates relative bandwidth. represents the non-negative deviation of the weighted half-width at half-maximum relative to the steady-state baseline, and exp represents the exponential function with the natural constant as the base.

[0060] It should be noted that the preset window width needs to be determined based on the change cycle of the thermal and discharge states of the downhole cable joint. The effective statistical cycle for the downhole cable state is typically 7 to 30 days. Therefore, a preset window width of 15 days can be set. If the state change cycle is shortened during field monitoring, the window width can be adjusted to 1.5 times the actual cycle to ensure sufficient statistical sample size and timely performance. The preset proportional coefficient should match the conversion standard between robust statistics and conventional dispersion. In engineering, this coefficient is fixed at 1.4826. This value can accurately convert the absolute deviation of the median into a dispersion approximating the conventional standard deviation without additional adjustment, and it meets the statistical needs of most downhole monitoring data. The fixed lower limit constant should be much smaller than the relative bandwidth under normal operating conditions. In engineering, it can be set to 0.05, or 5%. This value can avoid abnormal jumps in the health index due to excessively small relative bandwidth, and it will not interfere with the index calculation under normal operating conditions, ensuring the stability and rationality of the output results. In addition, the partial discharge of downhole cables exhibits intermittent fluctuations. The rolling statistical interval can ensure that the benchmark parameters are dynamically updated according to the operating conditions. Furthermore, the method of selecting the reference subset using the lower quartile can accurately extract the steady-state period of low discharge and low disturbance. The health index calculated based on this benchmark can effectively distinguish between normal fluctuations and abnormal deterioration, thereby improving the accuracy of early fault identification.

[0061] It should be noted that the cable health index ranges from 0 to 100, and its value is positively correlated with the cable's health level. A higher cable health index indicates a smaller non-negative deviation of the cable joint's thermal dispersion characteristics from the steady-state baseline, a lower degree of disturbance of the cable's thermal state by partial discharge, and a better overall cable health. Conversely, a lower cable health index indicates a larger non-negative deviation of the cable joint's thermal dispersion characteristics from the steady-state baseline, a higher degree of disturbance of the cable's thermal state by partial discharge, and a worse overall cable health.

[0062] Specifically, a cable health index range of 80 to 100 corresponds to a high health level. Within this range, the thermal dispersion characteristics of the cable joint deviate very little from the steady-state baseline, partial discharge energy is at a low level with no significant abnormal fluctuations, and the cable as a whole is in a stable and reliable operating state. A cable health index range of 40 to 79 corresponds to a medium health level. Within this range, the thermal dispersion characteristics of the cable joint deviate to some extent from the steady-state baseline, partial discharge energy occasionally increases slightly, and the cable shows slight signs of degradation but has not affected current normal power supply. A cable health index range of 0 to 39 corresponds to a low health level. Within this range, the thermal dispersion characteristics of the cable joint deviate significantly from the steady-state baseline, partial discharge energy is frequently at a high level, the cable shows a clear tendency to deteriorate, and there is a potential risk of failure.

[0063] Specifically, the maintenance strategy for a high health level involves performing routine periodic inspections, conducting visual inspections and data verification of cable joints and surrounding monitoring equipment according to the established inspection frequency for downhole equipment. The maintenance strategy for a medium health level involves increasing monitoring frequency, halving the original monitoring cycle, and increasing the tracking and analysis of partial discharge energy trends and thermal dispersion half-width at half-maximum fluctuations. The maintenance strategy for a low health level involves immediately initiating a special investigation, organizing technical personnel to conduct a comprehensive on-site inspection of the cable joints, and reducing the cable's operating load to minimize the possibility of faults; details will not be elaborated further here.

[0064] The cable joint well environment integrated monitoring system provided by the embodiments of this application includes a unified time grid module 201, an original full width half height calculation module 202, a true full width half height calculation module 203, an axial heat dissipation half height width calculation module 204, a weighted half height width calculation module 205, and a cable health index generation module 206.

[0065] Among them, the unified time grid module 201 is used to acquire temperature profiles through distributed fiber optic temperature measurement, combine them with ambient temperature to form a temperature rise profile, collect partial discharge event time and energy sequences, and resample to form a unified time grid; the original full width half height calculation module 202 is used to calculate the original full width half height based on the temperature rise profile under the unified time grid; the true full width half height calculation module 203 is used to convert the original full width half height into the true full width half height based on the point spread function and the instrument space half height width; the axial thermal dissipation half height width calculation module 204 is used to treat the outer surface of the cable as the object of forced convection by the outflow cylinder, determine the air physical parameters based on wind speed, relative humidity and ambient temperature, and force convection through the outflow cylinder. The convective heat transfer coefficient is calculated using the convection correlation formula, and the true full width at half height (WHM) is normalized accordingly to obtain the axial thermal dispersion WHM of the joint. The weighted WHM calculation module 205 is used to perform low-pass processing on the partial discharge event time and energy sequence through a smoothing kernel and a weighting window to obtain the partial discharge energy low-pass sequence, and uses this sequence as a weight to perform time-weighted WHM of the joint axial thermal dispersion to obtain the weighted WHM. The cable health index generation module 206 is used to select a reference subset based on the partial discharge energy low-pass sequence within the rolling statistical interval, take the median of the joint axial thermal dispersion WHM in the reference subset and the absolute deviation of the median to determine the relative bandwidth, and generate the cable health index through index mapping.

[0066] It should be noted that cable circulating current refers to the induced current generated by electromagnetic induction in parts such as the metal sheath and shielding layer of underground cables during operation. In the underground environment, problems such as aging insulation at cable joints and loose connections can cause abnormal changes in the amplitude and frequency of the circulating current. Circulating current monitoring needs to capture the temporal characteristics and spatial distribution of this current in real time to determine the insulation integrity and connection reliability of the cable joints and the line. Similarly, the circulating current monitoring of cables can be implemented based on the concept of this invention, specifically including the following steps: 1. Constructing a unified time grid: Acquire cable circulation time series data through Rogowski coil sensors, and combine it with downhole ambient temperature and humidity data to construct a unified time grid generated by a fixed start time and time interval. Resample the circulation data and environmental parameters to this grid by interpolation according to the nearest timestamp, and at the same time, convert the discrete circulation pulse signal into a continuous circulation density sequence and align it with the time grid.

[0067] 2. Calculate the original peak circulating current width: Within a unified time grid, using the cable mid-joint position as a reference, define an analysis interval of a preset length, extract the circulating current density data within this interval, determine the maximum circulating current amplitude as the peak circulating current, take half of the peak circulating current as the reference value, locate the two positions to the left and right of the circulating current density equal to the reference value within the analysis interval through linear interpolation, calculate the absolute value of the difference between the two positions, and obtain the original peak circulating current width.

[0068] 3. Calculate the true peak circulation width: Establish a point spread function convolution model with the instrument's inherent error as the parameter, set the point spread function to conform to a Gaussian distribution, subtract the square of the instrument error parameter from the square of the original peak circulation width, take the arithmetic square root of the difference, remove the instrument measurement interference, and obtain the true peak circulation width.

[0069] 4. Circulation parameter scaling: Determine the air conductivity parameters based on ambient temperature and humidity, calculate the equivalent coefficient affecting circulation transmission in conjunction with cable laying spacing, set a reference equivalent coefficient, calculate its ratio with the current equivalent coefficient as the scaling coefficient, multiply the actual circulation peak width by the scaling coefficient to obtain the normalized axial circulation characteristic width.

[0070] 5. Calculate the weighted circulation characteristic width: The partial discharge energy sequence is low-pass processed by a smoothing kernel to obtain a continuous partial discharge energy low-pass sequence. Within a preset weighted window, the time integral of the product of the axial circulation characteristic width and the low-pass sequence is used as the numerator, and the time integral of the low-pass sequence is calculated and a very small positive number is added as the denominator. The ratio of the numerator to the denominator is the weighted circulation characteristic width.

[0071] 6. Generate circulation-related health index: Within the rolling statistical interval, the time corresponding to the quartile of the partial discharge energy low-pass sequence is selected as the reference subset. The median and dispersion of the axial circulation characteristic width within the reference subset are calculated to determine the relative bandwidth. Then, by weighting the deviation of the circulation characteristic width from the median, the circulation-related health index in the range of 0 to 100 is generated through index mapping.

[0072] The above embodiments follow the multi-source data fusion and robust statistical approach of this invention to achieve accurate monitoring of underground cable circulating current. By using a unified time grid, the time reference difference between circulating current data and environmental and partial discharge data is resolved; instrument measurement interference is removed to restore the true circulating current characteristics; the influence of environmental parameter fluctuations on circulating current monitoring is eliminated, making circulating current parameters comparable under different operating conditions; a time-domain correlation between circulating current and partial discharge energy is established, highlighting the impact of potential faults on current distribution; and a quantitative health index is generated to intuitively reflect the degree of cable circulating current anomalies. This monitoring supplements the original dimensions of thermal parameter and partial discharge monitoring, enabling a comprehensive assessment of the insulation and connection status of cable joints, further improving the comprehensiveness and timeliness of underground cable fault early warning, and ensuring the stable operation of the power transmission system.

[0073] It should be noted that, as Figure 2As shown, in downhole cable monitoring, due to the inherent spatial resolution limitation (half-width at half maximum) of the distributed fiber optic temperature measurement (DTS) equipment, the directly measured temperature rise data often suffers from spatial measurement ambiguity. This invention introduces a point spread function (PSF) convolution model that conforms to a Gaussian distribution to perform inverse deconvolution processing on the original data, thereby restoring the true temperature rise profile of the cable joint.

[0074] It should be noted that, as Figure 3 As shown, the wind speed in underground roadways fluctuates significantly due to production shifts and seasons. This forced convection environment directly alters the heat dissipation efficiency of the cable surface, leading to non-fault-related changes in the temperature rise distribution pattern of the joint. This invention utilizes an external cylindrical forced convection model to calculate the Reynolds number and Nusselt number using real-time monitored parameters such as wind speed and temperature, thereby obtaining the dynamic convective heat transfer coefficient and achieving scale normalization of the true full width at half height. The processed joint axial heat dissipation half-height eliminates the influence of wind speed fluctuations, and the curve tends to be stable. This effectively distinguishes between normal fluctuations caused by the environment and abnormalities caused by equipment failures, achieving accurate condition monitoring across operating conditions.

[0075] It should be noted that, as Figure 4 As shown, the simulation depicts the entire process of a cable joint from healthy operation to early degradation and increased risk of failure. On day 10, early insulation defects (degradation begins) are introduced into the simulation. This invention can capture abnormal pulses in the partial discharge energy sequence and find that its temporal correlation with thermal dispersion parameters is enhanced, helping maintenance personnel to identify degradation trends in the early stages of a fault, thereby avoiding serious power outages.

[0076] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0077] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A method for comprehensive monitoring of the underground environment at cable joints, characterized in that, Includes the following steps: Step S101: Obtain a temperature profile through distributed optical fiber temperature measurement, combine it with the ambient temperature to form a temperature rise profile, collect the time and energy sequence of partial discharge events, and resample to form a unified time grid. Step S102: Calculate the original full width and half height based on the temperature rise profile under a unified time grid. Step S103: Convert the original full width half height to the true full width half height based on the point spread function and the instrument space half height. Step S104: The outer surface of the cable is taken as the object of forced convection of the outflow cylinder. The air physical parameters are determined according to the wind speed, relative humidity and ambient temperature. The convective heat transfer coefficient is calculated by the correlation formula of forced convection of the outflow cylinder. Based on this, the true full width half height is normalized to obtain the axial heat dissipation half height of the joint. Step S105: The partial discharge event time and energy sequence is low-pass processed by smoothing kernel and weighted window to obtain partial discharge energy low-pass sequence, and the weighted half width of the joint axial thermal dispersion is time-weighted using this sequence as weight to obtain weighted half width of the joint. Step S106: Select a reference subset based on the partial discharge energy low-pass sequence within the rolling statistical interval, determine the relative bandwidth by taking the median of the joint axial thermal dissipation half-width within the reference subset and the absolute deviation of the median, and generate the cable health index through exponential mapping.

2. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, Step S101 involves acquiring a temperature profile through distributed fiber optic temperature measurement, combining it with the ambient temperature to form a temperature rise profile, collecting the time and energy sequences of partial discharge events, and resampling them to form a unified time grid, including: A set of discrete time points generated recursively from the start time and time interval is constructed as a unified time grid; the temperature profile obtained by distributed fiber optic temperature measurement and the ambient temperature are resampled to the unified time grid according to the nearest timestamp and interpolation weight. The temperature rise profile is constructed by calculating the difference between the resampled temperature profile and the ambient temperature. The partial discharge event time and energy sequence is converted into pulse energy density, and the pulse energy density is integrated and aggregated in the time domain within the time interval defined by a unified time grid to form a partial discharge energy sequence aligned with the unified time grid.

3. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, Step S102, which calculates the original full width and half height based on the temperature rise profile under a unified time grid, includes: Select the current moment within a unified time grid, and extract temperature rise profile data within a preset length of the analysis half-window, using the center position of the joint as a reference. The search analysis identifies the maximum temperature rise within the half-window range as the peak temperature rise, and half of the peak temperature rise is taken as the half-height line height.

4. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 3, characterized in that, Step S102, which calculates the original full width and half height based on the temperature rise profile under a unified time grid, includes: To the left of the center of the joint, determine two adjacent sampling positions where the temperature rise is greater than and less than the half-height line height, respectively. Based on the coordinates of these two adjacent sampling positions and the temperature rise value, establish a linear relationship and calculate the spatial coordinates when the temperature rise value is equal to the half-height line height as the left half-height intersection point. To the right of the center of the joint, determine two adjacent sampling positions where the temperature rise is greater than and less than the half-height line height, respectively. Based on the coordinates of these two adjacent sampling positions and the temperature rise value, establish a linear relationship and calculate the spatial coordinates when the temperature rise value is equal to the half-height line height as the right half-height intersection point. Calculate the absolute value of the difference between the intersection point of the right half-height and the intersection point of the left half-height to obtain the original full width and half-height at the current moment.

5. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, Step S103, which converts the original full width half-height to the true full width half-height based on the point spread function and the instrument space half-width, includes: A convolution model with a point spread function as the convolution kernel is established, where the spatial scale parameter of the point spread function is determined by the half-width at half-maximum of the instrument space. The spread function at the set point conforms to a Gaussian distribution. The difference between the square of the original full width and half height and the square of the instrument space half height is calculated, and the arithmetic square root of this difference is performed to obtain the true full width and half height.

6. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, In step S104, the outer surface of the cable is taken as the object of forced convection in the outflow cylinder. Air properties are determined based on wind speed, relative humidity, and ambient temperature. The convective heat transfer coefficient is calculated using the outflow cylinder forced convection correlation formula. Based on this, the true full width at half height is normalized to obtain the axial heat dissipation half-height of the joint, including: The air density, aerodynamic viscosity, air thermal conductivity, and air specific heat at constant pressure are determined based on ambient temperature and relative humidity, and the air kinematic viscosity, air thermal diffusivity, and Prandtl number are calculated accordingly.

7. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 6, characterized in that, In step S104, the outer surface of the cable is taken as the object of forced convection in the outflow cylinder. Air properties are determined based on wind speed, relative humidity, and ambient temperature. The convective heat transfer coefficient is calculated using the outflow cylinder forced convection correlation formula. Based on this, the true full width at half height is normalized to obtain the axial heat dissipation half-height of the joint, including: The outer surface of the cable is used as the boundary of the external cylindrical forced convection model. The Reynolds number is calculated based on air density, wind speed, cable outer diameter, and aerodynamic viscosity. The Nusselt number is calculated based on the Reynolds number and Prandtl number using the external cylindrical forced convection correlation formula containing empirical coefficients. The convective heat transfer coefficient at the current moment is calculated based on the Nusselt number, air thermal conductivity, and cable outer diameter. Set a reference convective heat transfer coefficient, calculate the ratio of the reference convective heat transfer coefficient to the convective heat transfer coefficient at the current moment to form a scale factor; calculate the product of the scale factor and the true full width at half height to obtain the axial heat dissipation half width at half height of the joint.

8. The method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, Step S105 involves low-pass processing of the partial discharge event time and energy sequence using a smoothing kernel and a weighted window to obtain a partial discharge energy low-pass sequence, and then using this sequence as a weight to perform time-weighted multiplication of the joint axial thermal dispersion full width at half maximum (FWHM) to obtain a weighted FWHM, including: Obtain the partial discharge energy sequence and the full width at half maximum (FWHM) of the joint axial thermal dissipation under a unified time grid; perform temporal convolution operation on the partial discharge energy sequence using a smoothing kernel that satisfies the conditions of nonnegativity and integral normalization to generate a low-pass partial discharge energy sequence. Within a weighted window of a preset duration, the time integral of the product of the joint axial thermal dispersion half-width and the partial discharge energy low-pass sequence is calculated and used as the numerator. Within the weighted window, the time integral value of the partial discharge energy low-pass sequence is calculated and added to a preset minimum positive number as the denominator; the ratio of the numerator to the denominator is calculated to obtain the weighted half-width.

9. A method for comprehensive monitoring of the underground environment at a cable joint as described in claim 1, characterized in that, In step S106, a reference subset is selected based on the partial discharge energy low-pass sequence within the rolling statistical interval. The relative bandwidth is determined by the median of the joint axial thermal dissipation half-width at half-maximum within the reference subset and its absolute deviation from the median. An exponential mapping is then used to generate a cable health index, including: Within a unified time grid, a rolling statistical interval is defined with the current time as the endpoint and a preset window width as the length. The lower quartile values ​​of the partial discharge energy low-pass sequence within this rolling statistical interval are counted, and the set of times when the partial discharge energy low-pass sequence values ​​within this rolling statistical interval are not higher than the lower quartile values ​​is used as a reference subset. Obtain the joint axial thermal dispersion half-width at each time point within the reference subset and calculate its median; calculate the median of the absolute value of the difference between the joint axial thermal dispersion half-width at each time point and the median, and multiply the median of the absolute value by a preset scaling factor to obtain the dispersion. Calculate the ratio of the dispersion to the median, and select the larger value between this ratio and a preset fixed lower limit constant as the relative bandwidth; Calculate the ratio of the weighted half-width to the median, subtract one from the ratio to obtain the difference, and select the larger value between the difference and zero as the non-negative deviation; calculate the ratio of the natural logarithm of the second to the relative bandwidth, calculate the negative of the product of the ratio and the non-negative deviation, use the negative as the exponent to calculate the natural exponential function value, and multiply the natural exponential function value by one hundred to obtain the cable health index.

10. A comprehensive monitoring system for the underground environment of a cable joint, characterized in that, The method for comprehensive monitoring of the underground environment at a cable joint as described in any one of claims 1 to 9 includes: The unified time grid module is used to obtain temperature profiles through distributed fiber optic temperature measurement, combine them with ambient temperature to form a temperature rise profile, collect the time and energy sequences of partial discharge events, and resample them to form a unified time grid. The original full width half height calculation module is used to calculate the original full width half height based on the temperature rise profile under a uniform time grid. The true full width half height calculation module is used to convert the original full width half height into the true full width half height based on the point spread function and the instrument space half height. The axial heat dissipation half-height and width calculation module is used to treat the outer surface of the cable as the object of forced convection of the outflow cylinder. It determines the air physical parameters based on wind speed, relative humidity and ambient temperature, calculates the convective heat transfer coefficient through the outflow cylinder forced convection correlation formula, and performs scale normalization on the true full width and half-height to obtain the axial heat dissipation half-height and width of the joint. The weighted half-width at half-maximum (WHM) calculation module is used to perform low-pass processing on the time and energy sequences of partial discharge events through a smoothing kernel and a weighting window to obtain a low-pass sequence of partial discharge energy, and then use this sequence as a weight to perform time-weighted calculation on the axial thermal dispersion half-width of the joint to obtain the weighted half-width at half-maximum. The cable health index generation module is used to select a reference subset based on the partial discharge energy low-pass sequence within the rolling statistical interval, determine the relative bandwidth by taking the median of the joint axial thermal dissipation half-width within the reference subset and the absolute deviation of the median, and generate the cable health index through index mapping.

Images