A method and system for coordinated early warning of condensation and overheating in cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array.
By using a gradient moisture-conducting flexible temperature and humidity heterogeneous array and adaptive acquisition technology, the problems of adaptability to complex curved surfaces and temperature and humidity distribution in cable joint monitoring have been solved, enabling coordinated early warning of condensation and overheating, and improving the accuracy and reliability of monitoring.
Patent Information
- Application Number
- CN202611549851.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-08
- Publication Date
- 2026-11-17
AI Technical Summary
Existing cable joint monitoring technologies are ill-suited to complex curved surfaces and cannot accurately acquire the spatial distribution of local temperature and humidity anomalies. Humidity-sensitive units are prone to saturation and hysteresis under high humidity or condensation conditions. Temperature and humidity cross-sensitivity and long-term drift affect measurement accuracy, and the spatial coupling risks of condensation and overheating are difficult to identify in a coordinated manner.
A gradient moisture-conducting flexible temperature and humidity heterogeneous array is adopted, including a flexible insulating substrate, interdigitated moisture-sensitive electrodes, a gradient moisture-conducting moisture-sensitive layer, and a moisture-permeable liquid-resistant protective layer, to construct an axial-circumferential two-dimensional monitoring array. By combining adaptive acquisition, data preprocessing, spatiotemporal alignment, and multiple compensation, the accuracy and reliability of temperature and humidity data are achieved. Dew point margin, temperature and humidity change rate, and spatial gradient are integrated for collaborative early warning.
It enables integrated multi-node monitoring of temperature and humidity fields of cable joints, reduces the impact of liquid condensation on measurements, improves data accuracy and reliability, and can accurately identify condensation, overheating and their combined anomalies, providing node-level assessment and regional location-based hierarchical early warning.
Smart Images

Figure CN123111379A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of power cable accessory condition monitoring, flexible sensors, and power equipment fault early warning technology, specifically to a method and system for coordinated early warning of condensation and overheating of cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array. Background Technology
[0002] Cable joints are crucial components in power cable lines, used for conductor connection, electric field transition, insulation restoration, shielding continuity, and mechanical sealing. Because cable joints contain conductor connection structures, insulation restoration structures, stress control structures, and external shielding transition structures, they have numerous material interfaces and complex structures, making them susceptible to the combined effects of installation processes, load fluctuations, contact resistance, ambient temperature and humidity, sealing performance, and insulation material aging.
[0003] When the quality of conductor crimping or bolting at cable joints deteriorates, the contact resistance at the conductor connection increases, easily leading to localized temperature rise and persistent overheating. Localized overheating accelerates the thermal aging of insulation materials and may cause insulation layer deformation, interface debonding, and degradation of stress control material performance.
[0004] On the other hand, cable joints are often installed in environments with high humidity and limited ventilation, such as cable trenches, underground pipe corridors, switchgear, ring main units, or outdoor terminal boxes. When the local surface temperature of the joint is close to or lower than the dew point temperature corresponding to the surrounding moisture, condensation is likely to form in areas of insulation stress concentration, stress cone areas, outer shielding transition areas, or weak sealing areas. The water film formed by condensation reduces the surface resistance of the insulation, increasing the risk of surface leakage current and partial discharge.
[0005] In actual operation of cable joints, condensation and overheating are not entirely independent. Localized overheating can alter the temperature field and moisture migration process near the joint; when the load decreases or the ambient temperature changes rapidly, previously accumulated moisture may recondense in lower-temperature areas. Moisture intrusion and condensation can also cause corrosion at conductor connection points, further increasing contact resistance. Therefore, there is a certain spatiotemporal coupling relationship between condensation and overheating, requiring coordinated monitoring and comprehensive early warning.
[0006] Current methods for monitoring cable joint temperature typically employ infrared thermometry, thermocouples, rigid digital temperature sensors, fiber optic thermometry, or wireless passive temperature sensors. Infrared thermometry is susceptible to limitations imposed by field-of-view obstruction, surface emissivity, and manual inspection cycles. Rigid sensors struggle to achieve adequate contact with the cylindrical, conical, and irregular transition surfaces of cable joints, and air gaps between the sensor and the joint surface can affect measurement accuracy. Single-point temperature monitoring also struggles to capture the axial and circumferential temperature distribution of the cable joint, making it difficult to accurately identify localized hotspots.
[0007] Existing humidity monitoring typically involves installing ambient humidity sensors inside switchgear, cable trenches, or cable terminal boxes. These sensors measure the macroscopic ambient humidity around the equipment, making it difficult to characterize the localized humidity accumulation in the microenvironment of cable joint surfaces, and even more difficult to accurately determine the axial and circumferential locations where condensation occurs.
[0008] In addition, existing flexible or thin-film humidity sensors have the following problems in cable joint monitoring applications:
[0009] First, the pore structure and hydrophilicity distribution inside a typical humidity-sensitive layer are relatively uniform, and the direction of water vapor transport is unclear, which makes it difficult to balance response speed, sensitivity, and recovery speed.
[0010] Second, under high humidity and condensation conditions, liquid water may come into direct contact with the humidity-sensitive material and interdigitated electrodes, causing humidity response saturation, output drift, or even electrode short circuit.
[0011] Third, the resistance, capacitance, or impedance response of humidity sensors is usually affected by temperature, and the temperature rise caused by changes in cable load may be misinterpreted as a change in humidity.
[0012] Fourth, humidity-sensitive materials have inconsistent response paths during the moisture absorption and desiccation processes. Without hysteresis compensation, the same relative humidity may produce different outputs during the moisture absorption and desiccation stages.
[0013] Fifth, the temperature sensing resistor electrode of the flexible temperature sensor will generate self-heating under the action of excitation current, and the installation thermal resistance between the sensor encapsulation layer, adhesive layer and connector surface will also cause measurement deviation.
[0014] Sixth, after long-term operation, flexible sensing nodes may experience problems such as low-frequency drift, local debonding, signal saturation, communication loss, and decreased consistency between nodes. Using data from all nodes in a fixed manner can easily lead to false alarms or missed alarms.
[0015] Seventh, existing methods mostly use temperature and humidity thresholds to judge anomalies separately, without making full use of dew point margin, temperature and humidity spatial gradient, anomaly duration, anomaly area continuity, and temperature and humidity anomaly spatial correlation. It is difficult to distinguish between overall high humidity, local condensation, overall load temperature rise, local overheating, and condensation-overheating synergistic anomalies.
[0016] Therefore, it is necessary to provide a method and system that can adapt to the complex curved surface of cable joints and realize multi-point array monitoring of temperature and humidity, gradient moisture conduction, liquid water barrier, temperature and humidity coupling compensation, node reliability assessment, abnormal area location, and condensation and overheating coordinated early warning. Summary of the Invention
[0017] The present invention aims to provide a method and system for early warning of condensation and overheating of cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array, in order to solve the problems in existing cable joint monitoring, such as rigid sensors being difficult to fit complex curved surfaces, difficulty in obtaining the spatial distribution of local temperature and humidity anomalies, easy saturation and hysteresis of humidity-sensitive units under high humidity or condensation conditions, cross-sensitivity of temperature and humidity and long-term drift affecting measurement accuracy, and difficulty in coordinating the identification of condensation, overheating and the spatial coupling risks between the two.
[0018] To achieve the above objectives, the present invention adopts the following technical solution.
[0019] A method for coordinated early warning of condensation and overheating in cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array includes the following steps:
[0020] Step S1: Construct a flexible temperature and humidity heterogeneous array. The array includes multiple gradient moisture-conducting flexible humidity sensing units and multiple flexible surface temperature sensing units. At least one spatially associated humidity sensing unit and at least one surface temperature sensing unit are defined as a sensing node. Multiple sensing nodes are attached and arranged along the axial and circumferential directions of the cable joint to establish a spatial coordinate mapping relationship between the sensing nodes and the surface position of the cable joint. The gradient moisture-conducting flexible humidity sensing unit includes a flexible insulating substrate, interdigitated humidity-sensitive electrodes, a gradient moisture-conducting humidity-sensitive layer, a moisture-permeable liquid-resistant protective layer, and a flexible encapsulation layer. The flexible surface temperature sensing unit includes a serpentine temperature-sensing resistance electrode and an insulating and thermally conductive encapsulation layer. The moisture-permeable liquid-resistant protective layer allows water vapor to pass through and prevents droplets exceeding a preset size from directly entering the gradient moisture-conducting humidity-sensitive layer, thereby reducing the probability of liquid condensation directly wetting the humidity-sensitive layer and the interdigitated humidity-sensitive electrodes.
[0021] Multiple axial monitoring sections are set along the conductor connection area, stress control transition area, insulation recovery area, and sealing transition area of the cable joint. Multiple sensing nodes are set at circumferential intervals on each axial monitoring section to form a circumferential monitoring ring. Sensing nodes at corresponding circumferential positions of adjacent axial monitoring sections form an axial monitoring chain. According to the embodiment, at least one local environmental reference temperature node and / or at least one environmental reference temperature and humidity node are set; preferably, a local environmental reference temperature node that is thermally isolated from the surface of the cable joint is set on at least one axial monitoring section, and at least one environmental reference temperature and humidity node is set on the outside of the cable joint.
[0022] Step S2: Adaptive acquisition of temperature and humidity data. Based on the cable operating status and the initial sampling data during system startup, or the corrected surface temperature, corrected relative humidity, temperature and humidity change rate, and dew point margin obtained from the previous sampling cycle, determine the normal inspection mode, local alert mode, or high-frequency monitoring mode, and acquire the raw temperature signal and raw humidity response signal of each sensor node. In normal inspection mode, all node data are periodically acquired at the first sampling frequency. When the temperature change rate, humidity change rate, temperature and humidity spatial gradient, or dew point margin of any node in the previous sampling cycle reaches the alert condition, data of that node and its axial and circumferential adjacent nodes are acquired at a second sampling frequency higher than the first sampling frequency, while maintaining periodic inspection of other nodes. When the condensation risk index, overheating risk index, or synergistic risk index of any node or abnormal area in the previous sampling cycle reaches the high-frequency monitoring condition, data of the abnormal area nodes are acquired at a third sampling frequency higher than the second sampling frequency, and all nodes are scanned according to a preset cycle. When the risk index is lower than the recovery threshold and continues to reach the preset recovery time, the sampling mode level is reduced.
[0023] Step S3: Preprocess and spatiotemporally align the raw data. Abnormal pulse removal, data integrity marking, and filtering are performed on the raw temperature and humidity response signals. Time alignment and spatial registration are then performed based on the sampling timestamps and spatial coordinates to obtain spatiotemporally aligned temperature and humidity array data. Time alignment maps data from different sampling modes, frequencies, and times to a unified time series. Spatial registration maps the data from each node to the axial-circumferential two-dimensional space of the cable joint. Short-term missing data can be interpolated using historical data from the same node with interpolation markers retained. Continuously missing, communication-abnormal, or suspected open-circuit data are only marked for integrity and are not replaced by adjacent node data in this step. The first and last circumferential nodes are established as adjacencies according to a ring topology.
[0024] Step S4: Perform temperature and humidity compensation on the spatiotemporal aligned data. Implement temperature self-heating compensation, installation heat transfer deviation compensation, temperature and humidity cross-sensitivity compensation, moisture absorption-dehumidification hysteresis compensation, and low-frequency drift compensation sequentially to obtain the preliminary corrected surface temperature and preliminary corrected relative humidity of each sensing node. Calculate the self-heating temperature rise based on the excitation current, temperature measuring resistance, and calibrated equivalent self-heating thermal resistance, and determine the installation heat transfer correction amount by combining the installation thermal resistance, local ambient reference temperature, and dynamic thermal response parameters. Perform temperature and humidity cross-sensitivity compensation based on the normalized humidity response, humidity-sensitive layer temperature or its estimated value, and joint calibration coefficients. Determine the moisture absorption or dehumidification process based on the calibrated humidity response direction and perform hysteresis correction. Determine the low-frequency drift compensation term based on the historical baseline under stable operating conditions.
[0025] Step S5: Assess node credibility and reconstruct low-credibility nodes. Calculate node credibility based on signal quality, data integrity, drift credibility, and consistency of trends with neighboring nodes. Each credibility component ranges from 0 to 1, with higher values indicating higher credibility. For low-credibility nodes caused by communication loss, open circuits, or abnormal drift, perform spatial reconstruction with low-credibility and reconstruction markers based on neighboring high-credibility nodes and historical time-series data. When humidity response saturation is accompanied by decreased dew point margin, persistent high humidity, or slow abnormal recovery, retain the response and use it as a suspected liquid condensation indicator, without directly overwriting it with neighboring node data. Arrange measured values or marked reconstructed values according to spatial coordinates to form a corrected surface temperature field and a corrected relative humidity field. The reconstructed values have a lower weight in subsequent risk calculations than the high-credibility measured values.
[0026] Step S6: Extract condensation, overheating, and spatial correlation features. Calculate the dew point temperature based on the corrected relative humidity and the corresponding local environmental reference temperature, and calculate the dew point margin based on the difference between the corrected surface temperature and the dew point temperature. Extract the local temperature rise rate, local humidity increase rate, axial temperature and humidity gradient, circumferential temperature and humidity gradient, high humidity persistence feature, low dew point margin persistence feature, high temperature persistence feature, spatial continuity of abnormal areas, and spatial correlation of temperature and humidity anomalies. The high humidity persistence feature is the duration for which the corrected relative humidity is continuously higher than a preset threshold; the low dew point margin persistence feature is the duration for which the dew point margin is continuously lower than a preset threshold; and the high temperature persistence feature is the duration for which the corrected surface temperature or load-normalized temperature deviation is continuously higher than the corresponding threshold. The axial and circumferential temperature and humidity gradients are determined based on the corrected temperature and humidity differences between adjacent axial monitoring sections and adjacent nodes in the same circumferential monitoring ring, respectively. The spatial continuity of abnormal areas is used to characterize the degree of aggregation of abnormal nodes. The spatial correlation of temperature and humidity anomalies is determined based on the degree of overlap and / or proximity of temperature and humidity anomaly areas.
[0027] Step S7: Perform risk assessment, anomaly location, and graded early warning. Calculate the condensation risk index, overheating risk index, and condensation-overheating synergistic risk index for each node based on the aforementioned characteristics and node credibility. The condensation risk index is obtained by normalizing and weighting the dew point margin, dew point margin decrease rate, high humidity persistence characteristics, low dew point margin persistence characteristics, humidity growth rate, humidity spatial gradient, and humidity anomaly region continuity. The overheating risk index is obtained by normalizing and weighting the corrected surface temperature, local temperature rise, local temperature rise rate, load-normalized temperature deviation, temperature spatial gradient, high temperature persistence characteristics, and temperature anomaly region continuity. The synergistic risk index is obtained by weighting the condensation risk index, overheating risk index, their product coupling term, and the spatial correlation of temperature and humidity anomalies.
[0028] Condensation anomaly nodes, overheating anomaly nodes, or coordinated anomaly nodes that meet the regional identification criteria and are spatially continuous are classified into abnormal connected regions. Regional risk is determined based on the maximum regional risk, average regional risk, anomaly duration, and spatial expansion. Dynamic early warning thresholds are determined comprehensively based on cable load current, ambient temperature, ambient relative humidity, operating season, and historical health baselines. When overall ambient humidity changes but dew point margin remains within a safe range, humidity change trends at each node are consistent, and the spatial humidity gradient is small, the impact of overall environmental changes on local condensation early warning is suppressed. When local relative humidity abnormally increases relative to the ambient reference humidity, dew point margin continuously decreases, or the abnormal humidity area continuously expands, the sensitivity of the condensation risk judgment criteria for the corresponding node or region is increased.
[0029] A condensation warning signal is output when only the condensation risk index reaches the condensation warning threshold; an overheating warning signal is output when only the overheating risk index reaches the overheating warning threshold; a condensation warning signal and an overheating warning signal are output respectively when both the condensation risk index and the overheating risk index reach their respective thresholds but the spatial correlation of temperature and humidity anomalies does not reach the synergistic threshold; and a condensation-overheating synergistic warning signal is output when both the synergistic risk index and the spatial correlation of temperature and humidity anomalies reach the synergistic warning threshold. The anomaly level is determined based on the magnitude of the risk index exceeding the threshold, the duration of the anomaly, the number of nodes in the anomaly area, and the degree of spatial expansion. The anomaly location is determined based on the axial and circumferential angular coordinates of the anomaly nodes or anomaly area, and the anomaly type, anomaly level, anomaly location, and corresponding risk index are output.
[0030] This invention also provides a cable joint condensation and overheating coordinated early warning system for performing the above-described method, comprising a flexible temperature and humidity heterogeneous array, a signal acquisition and adaptive control module, a data preprocessing and spatiotemporal alignment module, a temperature and humidity compensation module, a node credibility and reconstruction module, a feature extraction and risk assessment module, a location and graded early warning module, and a communication module. The flexible temperature and humidity heterogeneous array is used to acquire raw temperature signals, raw humidity response signals, and environmental reference data from different spatial locations; the signal acquisition and adaptive control module is used to acquire signals and switch between three sampling modes; the data preprocessing and spatiotemporal alignment module is used to perform abnormal pulse rejection, integrity marking, filtering, time alignment, spatial registration, and ring adjacency relationship establishment; the temperature and humidity compensation module is used to implement various temperature and humidity compensations; the node credibility and reconstruction module is used to calculate node credibility and reconstruct low-credibility nodes that meet the conditions; the feature extraction and risk assessment module is used to extract temperature and humidity features and calculate node-level condensation, overheating, and coordinated risk indices; the location and graded early warning module is used to divide abnormal connectivity areas, determine the abnormality type, level, and location, and output corresponding early warning signals to the cable condition monitoring platform or maintenance terminal through the communication module.
[0031] This invention achieves integrated, multi-node, and spatially distributed monitoring of temperature and humidity fields by constructing an axial-circumferential two-dimensional flexible temperature and humidity heterogeneous array on the surface of cable joints. Through a gradient moisture-conducting and moisture-permeable liquid-resistant protective layer, it promotes the migration of water vapor to the humidity-sensitive detection area while reducing response saturation and measurement distortion caused by droplet intrusion. Adaptive sampling, data preprocessing, spatiotemporal alignment, multiple compensation, node reliability assessment, and labeled reconstruction improve the accuracy and reliability of temperature and humidity data. By integrating dew point margin, temperature and humidity change rate, spatial gradient, anomaly persistence characteristics, anomaly region continuity, and spatial correlation of temperature and humidity anomalies, it achieves node-level assessment, regional location, and graded early warning of condensation risk, overheating risk, and synergistic risks. Attached Figure Description
[0032] Figure 1 This is a flowchart of the overall early warning method for condensation and overheating corresponding to the present invention and the amended claims;
[0033] Figure 2 This is a top view schematic diagram of the gradient moisture-conducting flexible temperature and humidity sensing node of the present invention;
[0034] Figure 3 This is a schematic cross-sectional view of the gradient moisture-conducting flexible temperature and humidity sensing node of the present invention.
[0035] Figure 4 A schematic diagram showing the attachment and deployment of a flexible temperature and humidity heterogeneous array, a local environmental reference temperature node, and an environmental reference temperature and humidity node.
[0036] Figure 5 A block diagram of the condensation and overheating coordinated early warning system corresponding to the amended system claims;
[0037] Figure 6 A schematic diagram for correcting the surface temperature field, correcting the relative humidity field, spatial correlation of temperature and humidity anomalies, and determining the location of anomalies. Detailed Implementation
[0038] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. The embodiments are given for illustrative purposes only and should not be construed as limiting the present invention. The accompanying drawings are for reference and illustration only and do not constitute a limitation on the scope of patent protection of the present invention, because many changes can be made to the present invention without departing from the spirit and scope of the present invention.
[0039] A method for early warning of condensation and overheating of cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array includes the following steps.
[0040] Step S1: Fabrication of a gradient moisture-wicking flexible temperature and humidity sensing unit
[0041] An interdigitated humidity-sensitive electrode and a serpentine temperature-sensing resistor electrode are disposed on a flexible insulating substrate. A gradient moisture-conducting humidity-sensitive layer is formed on the interdigitated humidity-sensitive electrode, and a moisture-permeable liquid-resistant protective layer is disposed on the outside of the gradient moisture-conducting humidity-sensitive layer. A flexible temperature and humidity sensing node including a gradient moisture-conducting flexible humidity sensing unit and a flexible surface temperature sensing unit is prepared.
[0042] The flexible temperature and humidity sensing node includes a flexible insulating substrate, a patterned electrode layer, a gradient moisture-conducting and humidity-sensitive layer, a moisture-permeable and liquid-resistant protective layer, and a flexible encapsulation layer covering the non-detection area.
[0043] The patterned electrode layer includes at least:
[0044] Interdigitated humidity-sensitive electrodes used to detect changes in resistance, capacitance, or impedance of humidity-sensitive materials;
[0045] A serpentine resistance thermometer electrode used to detect the surface temperature of cable joints;
[0046] Flexible interconnecting wires used to connect interdigital humidity-sensitive electrodes and serpentine temperature-sensing resistor electrodes to signal acquisition circuits.
[0047] Preferably, an electrically insulating isolation zone or a thermally insulating groove is provided between the interdigitated humidity-sensitive electrode and the serpentine temperature-sensing resistor electrode to reduce the influence of the temperature-sensing resistor's self-heating and the moisture absorption process of the humidity-sensitive layer on the surface temperature measurement.
[0048] The flexible insulating substrate is made of polyimide, polyethylene terephthalate, thermoplastic polyurethane, polydimethylsiloxane, or a composite material thereof.
[0049] The humidity detection electrode is made of gold, platinum, copper, silver, nickel, conductive ink, silver nanowires, carbon nanotubes or graphene conductive materials; the temperature detection electrode is made of platinum, nickel, copper or conductive composite materials with temperature resistance effect.
[0050] The gradient moisture-conducting and humidity-sensitive layer has a pore size distribution and a hydrophilicity distribution that vary along the thickness direction. From the side near the moisture-permeable and liquid-blocking protective layer to the side near the humidity detection electrode, the average pore size of the gradient moisture-conducting and humidity-sensitive layer gradually decreases, while the density of hydrophilic groups or the content of hygroscopic material gradually increases, allowing water vapor to be transferred from the outside to the inside and forming a detectable adsorption response near the humidity detection electrode.
[0051] The gradient moisture-conducting and moisture-sensitive layer is made of a hydrophilic polymer matrix, a moisture-sensitive functional material, a conductive material, and a pore-forming material. The hydrophilic polymer matrix includes polyvinyl alcohol, cellulose, chitosan, polyacrylamide, polyethylene oxide, or combinations thereof; the moisture-sensitive functional material includes graphene oxide, MXene, cellulose nanofibers, silica, metal oxides, or combinations thereof; the conductive material includes carbon nanotubes, graphene, conductive carbon black, polyaniline, polypyrrole, or poly(ethylenedioxythiophene)-polystyrene sulfonate.
[0052] Preferably, the gradient moisture-conducting and moisture-sensitive layer includes at least two porous sublayers stacked along the thickness direction. Each porous sublayer has a different content of pore-forming particles of different sizes, different content of moisture-sensitive functional materials, or different degrees of crosslinking, thereby forming a gradient pore structure and a gradient hydrophilic structure.
[0053] The moisture-permeable and liquid-blocking protective layer is made of expanded polytetrafluoroethylene membrane, hydrophobic microporous polyurethane membrane, hydrophobic modified nanofiber membrane, or silicone rubber membrane with microporous structure. It is used to allow water vapor to pass through and prevent droplets larger than a preset size from directly entering the gradient moisture-conducting and humidity-sensitive layer, so as to reduce the probability of liquid condensation directly wetting the gradient moisture-conducting and humidity-sensitive layer and the interdigitated humidity-sensitive electrode.
[0054] Step S2: Construct a flexible temperature and humidity heterogeneous array for cable joints
[0055] Multiple sensing nodes, including gradient moisture-conducting flexible humidity sensing units and flexible surface temperature sensing units, are arranged along the axial and circumferential directions of the cable joint to form a flexible temperature and humidity heterogeneous array.
[0056] The cable joint is divided along the axial direction into at least two monitoring areas: a conductor connection area, a stress control transition area, an insulation recovery area, and a sealing transition area. Different node layout densities are set according to the fault sensitivity of each area.
[0057] Main monitoring nodes are set in the conductor connection area, stress control transition area, shielded break area and sealing transition area, and auxiliary monitoring nodes are set between adjacent main monitoring nodes.
[0058] The flexible temperature and humidity heterogeneous array includes:
[0059] A composite temperature and humidity node for detecting local humidity and cable joint surface temperature, and a local environmental reference temperature node set at at least one axial monitoring section and thermally isolated from the cable joint surface.
[0060] At least one environmental reference temperature and humidity node is installed outside the cable joint to detect the macroscopic ambient temperature and relative humidity.
[0061] The composite temperature and humidity node is attached to the surface of the cable joint; the local environmental reference temperature node is set on the outside of the composite temperature and humidity node or on the outer surface of the flexible encapsulation layer, and is separated from the surface of the cable joint by a heat insulation structure, so that it mainly reflects the micro-environmental air temperature near the joint.
[0062] The nodes in the flexible temperature and humidity heterogeneous array are numbered as node i, and the spatial coordinates of the node are represented as follows:
[0063]
[0064] Among them, z i θ represents the position coordinates of the node along the axial direction of the cable joint. i This indicates the angular coordinates of the node along the circumference of the cable joint.
[0065] The flexible temperature and humidity heterogeneous array is attached to the surface of the cable connector via a flexible strip substrate, a segmented wrap-around substrate, or a stretchable mesh substrate. Adjacent flexible substrates are connected by serpentine interconnects to accommodate cable connectors of different diameters and bending and stretching deformations during installation.
[0066] Step S3: Obtain heterogeneous array temperature and humidity data based on the state of the previous sampling period.
[0067] During the operation of the cable joint, humidity electrical response data, joint surface temperature data, local environmental reference temperature data, and external environmental reference temperature and humidity data are collected through a flexible temperature and humidity heterogeneous array, and at least one auxiliary operating condition data is obtained from the cable load current and operating status.
[0068] The system adopts an adaptive acquisition strategy based on a multi-state machine. Based on the initial sampling data during the startup phase or the corrected surface temperature, corrected relative humidity, temperature and humidity change rate, dew point margin, and risk index obtained in the previous sampling cycle, the system sets the flexible temperature and humidity heterogeneous array to normal inspection mode, local alert mode, or high-frequency monitoring mode.
[0069] In normal inspection mode, data from all nodes is collected periodically at the first sampling frequency.
[0070] When the temperature change rate, humidity change rate, temperature and humidity spatial gradient, or dew point margin of any node in the previous sampling period reaches the corresponding alarm condition, the local alarm mode is entered, and data of the node and its axial and circumferential adjacent nodes are collected at the second sampling frequency, while periodic inspection of the remaining nodes is maintained.
[0071] When the condensation risk index, overheating risk index, or condensation-overheating synergistic risk index of any node or abnormal area reaches the corresponding high-frequency monitoring conditions, the high-frequency monitoring mode is entered, and the abnormal area node data is collected at the third sampling frequency, and all nodes are scanned according to the preset cycle. When the corresponding risk index is lower than the recovery threshold and continues to reach the preset recovery time, the sampling mode level is reduced, where the third sampling frequency is higher than the second sampling frequency, and the second sampling frequency is higher than the first sampling frequency.
[0072] Step S4: Preprocess, time-align, and spatially register the heterogeneous array data.
[0073] First, abnormal pulses are removed from the raw data, data integrity is marked, filtered and limited. Then, based on the sampling timestamps of each node, the humidity electrical response data, surface temperature data, environmental reference data and auxiliary operating condition data are mapped to a unified time series.
[0074] For short-term missing data, linear interpolation, time-series interpolation with preservation, or interpolation based on the rate of change constraint are performed using historical time-series data of the same node, and the interpolation mark is retained; for continuously missing, communication abnormal, or suspected disconnected data, only data integrity mark is set, and it is left to be processed in the node credibility assessment and spatial reconstruction steps.
[0075] Based on the axial and circumferential positions of each node, the time-aligned data is mapped into an axial-circumferential two-dimensional array:
[0076]
[0077] Where m represents the axial monitoring section number, n represents the circumferential monitoring position number, and k represents the uniform sampling time.
[0078] Nodes located at the beginning and end of the circumference are connected to each other in a ring to avoid spatial topological breaks between the 0-degree and 360-degree positions. This step does not directly replace the original measurements of consecutively missing or abnormally drifting nodes with the data of adjacent nodes.
[0079] After preprocessing, time alignment, and spatial registration, spatiotemporally aligned temperature and humidity array data with data integrity markers and interpolation markers are obtained.
[0080] Step S5: Perform temperature and humidity coupling compensation, node reliability assessment, and spatial reconstruction.
[0081] The original resistance, capacitance, or impedance of the interdigitated humidity-sensitive electrode is converted into a normalized humidity electrical response X. i (k).
[0082] Preferably, the normalized electrical response of humidity is expressed as:
[0083]
[0084] Among them, Y i (k) represents the original humidity electrical response of node i at time k, Y i,min and Y i,max These represent the lower and upper limits of the response of node i within the calibration range, respectively.
[0085] Based on the excitation current, temperature sensing resistance, equivalent self-heating thermal resistance, installation thermal resistance, local ambient reference temperature, and dynamic thermal response parameters of the flexible surface temperature sensing unit, temperature sensing self-heating compensation and installation heat transfer deviation compensation are performed on the original surface temperature.
[0086]
[0087] Among them, T s,i (k) is for preliminary correction of surface temperature, T raw,i (k) represents the original temperature measurement value, R th,i,eq For the equivalent self-heating thermal resistance, ΔT install,i (k) is the signed installation heat transfer correction determined based on the installation thermal resistance and dynamic thermal response parameters.
[0088] Based on the normalized humidity electrical response X i (k) and humidity-sensitive layer temperature T h,i (k) or its estimated value is used to establish a basic compensation model for temperature and humidity coupling:
[0089]
[0090] Based on the calibrated humidity response direction coefficient The sign of the product of the humidity response change at adjacent sampling times determines whether the current node is in a moisture absorption or dehumidification process:
[0091]
[0092] Hysteresis correction terms η are set for both the moisture absorption and desiccation processes. i,s (k), and combined with the low-frequency drift estimate d i (k) Obtain the preliminary corrected relative humidity:
[0093]
[0094] Among them, H min The preset lower limit for humidity calculation is greater than zero. `clip(·)` restricts the calculation results to a range of H... min Within the range of %RH to 100%RH; the low-frequency drift estimate is determined based on historical baselines under stable operating conditions, periodic calibration data, and baseline change trends.
[0095] During low-frequency drift baseline updates, a baseline update timestamp, update amplitude limit, and pause update flag are set for each node to prevent abnormal states from being absorbed as normal baselines.
[0096] When a node enters a local alert mode or a high-frequency monitoring mode, or when the corresponding condensation risk, overheating risk, or synergistic risk reaches the conditions for pausing updates, the low-frequency drift baseline of the corresponding node is paused.
[0097] The stable operating condition should simultaneously meet the following basic conditions: the cable joint is not in a condensation risk state, an overheating risk state, or a condensation-overheating combined risk state; the cable load current change rate is lower than a preset load change rate threshold; the local surface temperature change rate is lower than a preset temperature change rate threshold; and the humidity response variance within the data window is lower than a preset response variance threshold. Preferably, the stable operating condition also satisfies that the humidity change trend of the sensing node is consistent with that of its axial and circumferential adjacent nodes. The low-frequency drift baseline of the corresponding sensing node is updated only after the stable operating condition has been maintained continuously for a preset stable duration.
[0098] After obtaining the preliminary corrected surface temperature and preliminary corrected relative humidity of each node, the node reliability is calculated based on the node's signal quality, data integrity, drift reliability, and consistency of the change trend of adjacent nodes.
[0099]
[0100] Among them, the value range of each confidence component is 0 to 1, and the larger the value, the higher the confidence. w1 to w4 are non-negative weights and their sum is 1. The consistency of adjacent nodes is mainly based on the trend of change, repeatability and cross-physical quantity verification relationship, and is not only determined by the deviation between the current measurement amplitude and the mean of the neighborhood.
[0101] When a node's confidence level is below a preset confidence threshold and the reason for the low confidence level is missing data, communication anomalies, or abnormal drift, spatial reconstruction with low confidence and reconstruction markers is performed based on the confidence level of adjacent high-confidence nodes, spatial distance, and historical time-series data. When the humidity response is saturated and accompanied by a decrease in dew point margin, persistent high humidity, or slow abnormal recovery, the original response is retained and a suspected liquid condensation indicator feature is generated. This yields the corrected surface temperature field and the corrected relative humidity field.
[0102] Step S6: Extract condensation features, overheating features, and spatial correlation features of temperature and humidity anomalies.
[0103] According to the corrected relative humidity H i (k) Local environmental reference temperature T corresponding to the node a,i (k) and corrected surface temperature T s,i (k) calculates the dew point temperature and dew point margin corresponding to the node.
[0104] Dew point temperature is calculated using the following formula:
[0105]
[0106]
[0107] Among them, T d,i (k) represents the dew point temperature; a and b are the dew point calculation parameters determined based on the target temperature range.
[0108] Dew point margin is calculated using the following formula:
[0109]
[0110] When the dew point margin is less than or equal to zero, it indicates that the node location meets the theoretical conditions for condensation formation; when the dew point margin is greater than zero but less than the preset safety margin, it indicates that the node location is in a state close to condensation.
[0111] Condensation characteristics include at least two of the following parameters:
[0112] (1) Dew point margin M d,i (k);
[0113] (2) Correction of relative humidity H i (k);
[0114] (3) Humidity change rate;
[0115] (4) Duration of high humidity;
[0116] (5) Dew point margin decrease rate and duration of low dew point margin;
[0117] (6) The spatial gradient of humidity between nodes and adjacent nodes and the continuity of humidity anomaly areas;
[0118] (7) Suspected liquid condensation indication characteristics determined based on humidity response saturation, dew point margin, abnormal duration and dehumidification recovery rate.
[0119] Overheating characteristics include at least two of the following parameters:
[0120] (1) Correcting surface temperature T s,i (k);
[0121] (2) Local temperature rise relative to the local ambient reference temperature;
[0122] (3) Surface temperature change rate or local temperature rise rate;
[0123] (4) Duration of high temperature or cumulative temperature rise within a preset time window;
[0124] (5) Spatial temperature gradient between nodes and adjacent nodes and continuity of temperature anomaly regions;
[0125] (6) Load-normalized temperature deviation relative to the historical health baseline under the same load conditions.
[0126] Local temperature rise is expressed as:
[0127]
[0128] The spatial temperature gradient of node i can be expressed as:
[0129]
[0130] The humidity spatial gradient at node i can be expressed as:
[0131]
[0132] Where, N i w represents the set of spatially adjacent nodes of node i; ij This represents the spatial weight determined by node distance, axial adjacency, and circumferential adjacency.
[0133] Based on the spatial location of condensation candidate areas and overheating candidate areas, the spatial correlation degree of temperature and humidity anomalies is calculated, and the spatial correlation degree includes the degree of regional overlap and / or the degree of regional proximity.
[0134]
[0135] Among them, O i (k) represents the spatial correlation degree of temperature and humidity anomalies, N i (k) represents the proximity degree determined based on the minimum distance between abnormal regions, adjacency relationship, or whether they are located in the same functional region. μ1 and μ2 are non-negative weights and their sum is 1.
[0136] The overlap of regions can be calculated based on the intersection and union areas of the condensation candidate regions and the overheating candidate regions; when the union of two candidate regions is empty, the overlap is defined as 0. For arrays with non-uniform node density, the overlap can be calculated by weighting based on the representative area of each node.
[0137] Step S7: Calculate the node-level condensation risk index, overheating risk index, and synergistic risk index.
[0138] The node condensation risk index is obtained by normalizing and weighting by dew point margin, dew point margin decline rate, high humidity persistence characteristics, low dew point margin persistence characteristics, humidity growth rate, humidity spatial gradient, suspected liquid condensation indicator characteristics, and continuity of humidity anomaly areas, and by node confidence.
[0139]
[0140] Among them, F c,p,i (k) represents the p-th condensation feature, φ p (·) represents the corresponding normalized mapping function, α p These are non-negative feature weights.
[0141] The node overheating risk index is obtained by normalizing and weighting the corrected surface temperature, local temperature rise, local temperature rise rate, load-normalized temperature deviation, temperature spatial gradient, high temperature persistence characteristics, and continuity of temperature anomaly regions, as well as node reliability.
[0142]
[0143] The nodal condensation-overheating synergistic risk index is calculated using the following formula:
[0144]
[0145] Among them, A i (k) represents the spatial correlation degree of temperature and humidity anomalies, λ c , λ h , λ ch and λ a The weights are non-negative and their sum is 1.
[0146] For nodes marked with a reconstruction flag, a risk weight lower than that of high-confidence measured nodes under the same conditions is adopted; for nodes with saturated humidity response but simultaneously meet the conditions of low dew point margin, persistent high humidity, or abnormally slow recovery, the risk contribution of suspected liquid condensation indication characteristics is retained.
[0147] The condensation anomaly nodes, overheating anomaly nodes, or collaborative anomaly nodes that meet the regional identification conditions and are spatially adjacent are divided into anomaly connected regions, and the first and last nodes in the circumferential direction are connected according to the ring spatial topology.
[0148] Regional risk can be determined by comprehensively considering the maximum value of node risk within the region, the confidence-weighted average, the duration of anomalies, and the degree of spatial expansion.
[0149]
[0150] Where ν1 to ν4 are non-negative weights and their sum is 1, R dur (k) and R exp (k) represents the normalized anomaly persistence risk and spatial expansion risk, respectively.
[0151] Condensation risk weight, overheating risk weight, synergistic risk weight, and regional risk weight can be dynamically adjusted based on cable load current, ambient temperature, ambient relative humidity, operating season, and the historical health status of cable joints.
[0152] Node-level risk index, regional risk, and dynamic early warning threshold are used together as the basis for determining anomaly type, anomaly level, and anomaly location.
[0153] Step S8: Locate the abnormal area and output a tiered early warning.
[0154] Nodes that meet the corresponding region identification criteria and are spatially adjacent, based on the condensation risk index, overheating risk index, or condensation-overheating synergistic risk index, are classified as corresponding abnormal connected regions.
[0155] For the circumferential first and last nodes of the cable joint, a connection relationship is established according to the annular spatial topology.
[0156] Calculate the risk-weighted center of anomalous connectivity regions:
[0157]
[0158] The circumferential angle center is calculated using a circumferential weighted method to avoid abrupt coordinate changes between the 0-degree and 360-degree positions.
[0159] Based on the condensation risk index, overheating risk index, condensation-overheating synergistic risk index, spatial correlation of temperature and humidity anomalies, duration of anomalies, and area of anomalies, the abnormal state is divided into at least three levels: normal state, condensation concern state, overheating concern state, concurrent condensation and overheating early warning state, synergistic early warning state, and severe alarm state.
[0160] Preferably, the warning levels include:
[0161] Level 1 Concern: The condensation risk index, overheating risk index, or condensation-overheating synergistic risk index exceeds the Level 1 threshold, but the duration does not exceed the Level 1 duration threshold.
[0162] Level 2 warning: The condensation risk index or overheating risk index exceeds the level 2 threshold, or the dew point margin is lower than the preset value and continues for a preset time, or the local temperature rise and temperature rise rate both exceed the corresponding thresholds; when both condensation risk and overheating risk exceed the corresponding thresholds but the spatial correlation of temperature and humidity anomalies does not reach the coordination threshold, condensation warning and overheating warning are output respectively.
[0163] Level 3 Alarm: The condensation-overheating synergistic risk index and the spatial correlation of temperature and humidity anomalies both exceed the corresponding Level 3 thresholds, and at least one of the abnormal area area, duration, or spatial expansion rate exceeds the advanced threshold.
[0164] The dynamic early warning threshold is determined comprehensively based on cable load current, ambient temperature, ambient relative humidity, operating season, and the historical health baseline of cable joints. For normal overall temperature rise caused by increased load current, when the temperature change trends of each sensing node are basically consistent, and the temperature spatial gradient and temperature differences between adjacent nodes do not reach abnormal conditions, the allowable range of the absolute temperature criterion is appropriately increased, while retaining the criteria for local temperature rise, temperature rise rate, temperature spatial gradient, and temperature deviation relative to the historical health baseline of the same load condition.
[0165] When the overall relative humidity of the environment increases, but the dew point margin of each sensing node is still within the preset safe range, the humidity change trend is basically the same, and the humidity spatial gradient is less than the preset humidity gradient threshold, the condensation warning threshold will not be reduced separately due to the increase in relative humidity of the environment.
[0166] When the local relative humidity of at least one sensing node is abnormally high relative to the ambient reference humidity, the dew point margin continues to decrease, the duration of low dew point margin increases, or the area of abnormal humidity continues to expand, the sensitivity of the condensation risk judgment criterion for the corresponding sensing node or abnormal area is increased.
[0167] The early warning output information should include at least the anomaly type, early warning level, anomaly node number, axial position, circumferential position, anomaly area range, current corrected surface temperature, current corrected relative humidity, dew point margin, spatial correlation of temperature and humidity anomalies, and corresponding risk index.
[0168] The present invention also provides a cable joint condensation and overheating coordinated early warning system based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array, for executing the above-mentioned early warning method.
[0169] The system includes: a flexible temperature and humidity heterogeneous array; a signal acquisition and adaptive control module; a data preprocessing and spatiotemporal alignment module; a temperature and humidity compensation module; a node reliability and reconstruction module; a feature extraction and risk assessment module; a positioning and hierarchical early warning module; and a communication and power supply module.
[0170] The flexible temperature and humidity heterogeneous array is used to collect cable joint surface temperature, local humidity, local ambient reference temperature, and external ambient reference temperature and humidity.
[0171] The signal acquisition and adaptive control module is used to excite, condition, convert analog to digital and output the humidity and temperature electrical signals, and switch the sampling mode according to the initial sampling data or the correction data of the previous sampling period and the risk status.
[0172] The data preprocessing and spatiotemporal alignment module is used for abnormal pulse removal, data integrity marking, filtering, timestamp alignment, spatial coordinate mapping, and establishment of circular adjacency relationships.
[0173] The temperature and humidity compensation module is used to implement temperature measurement self-heating compensation, installation heat transfer deviation compensation, temperature and humidity cross-sensitivity compensation, moisture absorption-dehumidification hysteresis compensation, and low-frequency drift compensation.
[0174] The node credibility and reconstruction module is used to calculate signal quality credibility, data integrity credibility, drift credibility, and trend consistency credibility, and to perform spatial reconstruction with low credibility markers and reconstruction markers on low credibility nodes that meet the reconstruction conditions.
[0175] The feature extraction and risk assessment module is used to calculate dew point temperature, dew point margin, local temperature rise, temperature and humidity change rate, temperature and humidity spatial gradient, high humidity persistence characteristics, low dew point margin persistence characteristics, high temperature persistence characteristics, spatial continuity of abnormal areas and spatial correlation of temperature and humidity anomalies, and to calculate node-level condensation risk index, overheating risk index and condensation-overheating synergistic risk index.
[0176] The positioning and hierarchical early warning module is used to identify condensation anomaly areas, overheating anomaly areas, and synergistic anomaly areas, determine the anomaly type, anomaly level, axial position, and circumferential position, and output condensation early warning signal, overheating early warning signal, or condensation-overheating synergistic early warning signal.
[0177] The communication and power supply module transmits data using wired communication, wireless communication, or a combination of both, and sends early warning information to the cable condition monitoring platform, power distribution automation platform, or maintenance terminal.
[0178] The power supply methods include external DC power supply, battery power supply, inductive energy harvesting, thermoelectric energy harvesting, or a combination thereof.
[0179] When both condensation risk and overheating risk meet the corresponding warning conditions, but the spatial correlation of temperature and humidity anomalies does not meet the coordination condition, the positioning and hierarchical warning module outputs condensation warning signals and overheating warning signals respectively; when both the coordination risk index and the spatial correlation of temperature and humidity anomalies meet the coordination condition, the module outputs a condensation-overheating coordinated warning signal.
[0180] After the risk is eliminated, the system uses recovery thresholds and recovery duration to control the cancellation of warnings and the downgrading of sampling modes in order to avoid frequent state switching.
[0181] The above modules can be implemented by a processor located in a field acquisition terminal, edge computing device, or cable condition monitoring platform, which executes the corresponding program instructions.
[0182] Compared with the prior art, the present invention has at least the following beneficial effects:
[0183] By employing a flexible insulating substrate and flexible interconnection structure, the temperature and humidity sensing unit can conform to the cylindrical, conical, and irregular transition surfaces of the cable joint, reducing the adhesion gap and improving the representativeness of local temperature and micro-environment humidity measurements. A moisture-sensitive layer with gradient aperture and gradient hydrophilicity guides water vapor migration to the humidity-sensitive detection area, and combined with a moisture-permeable and liquid-resistant protective layer, reduces response saturation and abnormal drift caused by droplet intrusion. By arranging heterogeneous arrays along the axial and circumferential directions of the cable joint and using differentiated node densities in key areas, the spatial distribution of temperature and humidity can be obtained, and condensation or overheating areas can be located. Compensation is provided for temperature cross-sensitivity, moisture absorption-dehumidification hysteresis, low-frequency drift, temperature self-heating, and installation heat transfer deviations to improve the accuracy of temperature and humidity data. By integrating dew point margin, local temperature rise, rate of change, spatial gradient, anomaly persistence characteristics, and spatial correlation of temperature and humidity anomalies, node-level and region-level risk indicators can be constructed, enabling the differentiation of condensation anomalies, overheating anomalies, concurrent anomalies, and synergistic anomalies. By employing node reliability assessment, labeled spatial reconstruction, and a mechanism for preserving suspected liquid condensation features, missing node data, open-circuit data, or abnormal drift data can be corrected, and the true condensation response can be prevented from being overwritten. Through multi-state adaptive acquisition, the sampling frequency is reduced under normal conditions and increased in local or full array sampling frequencies under abnormal conditions, balancing low power consumption and high temporal resolution.
[0184] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The following embodiments are only for illustrating the present invention and do not constitute a limitation on the scope of protection; equivalent substitutions or adjustments made without departing from the concept of the present invention are all within the scope of the technical concept of the present invention.
[0185] Example 1: Fabrication of a Gradient Moisture-Conducting Flexible Temperature and Humidity Sensing Unit
[0186] This embodiment provides a flexible temperature and humidity sensing unit suitable for attachment to the curved surface of a cable joint.
[0187] First, a polyimide film with a thickness of 25 μm to 200 μm is cleaned and surface-activated, serving as a flexible insulating substrate 1. A patterned electrode layer is formed on the surface of the flexible insulating substrate 1. The patterned electrode layer includes an interdigitated humidity-sensitive electrode 2, a serpentine temperature-sensing resistance electrode 3, and flexible interconnecting wires 7.
[0188] The interdigitated humidity-sensitive electrode 2 includes an alternating first interdigitated electrode and a second interdigitated electrode, forming a humidity-sensitive detection gap between the first and second interdigitated electrodes. A serpentine temperature-sensing resistance electrode 3 is disposed on one side or in the peripheral area of the interdigitated humidity-sensitive electrode 2, and the surface temperature of the cable joint is obtained by detecting changes in its resistance.
[0189] To reduce the mutual interference between the humidity detection area and the temperature detection area, an insulating isolation zone with a width of 0.2 mm to 2 mm is provided between the interdigitated humidity-sensitive electrode 2 and the serpentine temperature-sensing resistor electrode 3. When it is necessary to further reduce thermal conductivity coupling, a thermal isolation groove that penetrates or does not penetrate the flexible insulating substrate 1 can be fabricated within the insulating isolation zone.
[0190] Subsequently, a gradient moisture-conducting and moisture-sensitive layer 4 is fabricated on the surface of the interdigitated humidity-sensitive electrode 2. In this embodiment, the gradient moisture-conducting and moisture-sensitive layer 4 is divided into an outer moisture-conducting sublayer, an intermediate transport sublayer, and an inner sensing sublayer. The outer moisture-conducting sublayer has an average pore size of 80 μm to 200 μm, which is used to improve the water vapor ingress rate; the intermediate transport sublayer has an average pore size of 20 μm to 100 μm, which is used to guide water vapor transport along the thickness direction; the inner sensing sublayer has an average pore size of 5 μm to 50 μm and has a high content of moisture-sensitive functional materials, which is used to enhance the electrical response to humidity.
[0191] The outer moisture-conducting sublayer, the middle transport sublayer, and the inner sensitive sublayer can be formed by coating layer by layer after preparing slurries with different particle sizes and pore-forming particle contents, or they can be formed by gradient freeze-drying, gradient phase separation, continuous spraying, or layer-by-layer printing.
[0192] One specific preparation method is as follows:
[0193] Three types of moisture-sensitive composite slurries were prepared by using polyvinyl alcohol solution as the hydrophilic polymer matrix, graphene oxide and cellulose nanofibers as moisture-sensitive functional materials, and carbon nanotubes as conductive materials.
[0194] Sodium chloride particles or polymethyl methacrylate microspheres with different average particle sizes were added to three types of moisture-sensitive composite slurries as pore-forming particles.
[0195] Inner sensitive slurry, middle transport slurry and outer moisture-conducting slurry are sequentially coated on the interdigitated moisture-sensitive electrode 2; after curing or cross-linking, the pore-forming particles are removed by water washing or solvent etching, thereby forming a gradient moisture-conducting and moisture-sensitive layer 4 with an average pore size that gradually changes along the thickness direction.
[0196] An expanded polytetrafluoroethylene microporous membrane is bonded to the outside of the gradient moisture-conducting moisture-sensitive layer 4 to form a moisture-permeable and liquid-blocking protective layer 5. The moisture-permeable and liquid-blocking protective layer 5 allows water vapor to pass through and uses a hydrophobic microporous structure to prevent droplets larger than a preset size from directly entering the moisture-sensitive layer, thereby reducing the probability of liquid condensation directly wetting the moisture-sensitive layer and the interdigitated moisture-sensitive electrode.
[0197] A flexible encapsulation layer 6 is formed by covering the non-detection areas of the electrodes and interconnecting wires with thermoplastic polyurethane or polydimethylsiloxane flexible encapsulation material, while retaining the moisture-permeable window corresponding to the gradient moisture-conducting and moisture-sensitive layer 4.
[0198] The fabricated flexible temperature and humidity sensing unit can employ impedance, resistive, or capacitive humidity detection methods. When using impedance detection, the signal acquisition module applies AC excitation to the interdigital humidity-sensitive electrode 2 and measures the amplitude and phase response under different humidity conditions.
[0199] Example 2: Deployment of a Flexible Temperature and Humidity Heterogeneous Array
[0200] In this embodiment, multiple flexible temperature and humidity sensing units are arranged on the outer surface of medium-voltage or high-voltage cable joints.
[0201] With the cable joint axis as the z-axis, the cable joint is divided into a central area 11 of the connecting pipe, stress control transition areas on both sides 12, insulation recovery area 13, and sealing transition area 14. Five monitoring sections are set along the cable joint axis, and eight composite temperature and humidity nodes 8 are set at 45-degree intervals along the circumference of each monitoring section, forming a 5×8 axial-circumferential array.
[0202] The central area of the connecting pipe and the stress control transition zone are key monitoring areas. The circumferential interval can be reduced to 30 degrees or auxiliary nodes can be added between the existing nodes. The sealing transition zone is equipped with a main humidity monitoring node for monitoring the intrusion of external moisture.
[0203] Each node is connected via a flexible strip substrate. An insulating and thermally conductive adhesive layer is applied to the back of the flexible strip substrate, ensuring a tight fit between the nodes and the cable connector surface. For scenarios requiring detachable installation, a ring-shaped strap, flexible clip, or Velcro structure is used for fixation.
[0204] At least one local ambient temperature reference node 9 is provided at least one axial monitoring section. A heat insulation pad is provided between the local ambient temperature reference node 9 and the joint surface, so that it mainly reflects the micro-ambient air temperature near the joint, while the serpentine temperature measuring resistance electrode in the composite temperature and humidity node 8 is in direct thermal contact with the joint surface.
[0205] At least one environmental reference temperature and humidity node is set outside the cable joint to obtain the macroscopic ambient temperature and relative humidity. When an independent local environmental reference temperature node is not set at a certain axial monitoring section, the local environmental reference temperature node of the adjacent section, the low thermal coupling reference node, or the environmental reference temperature obtained by spatial interpolation can be used.
[0206] Example 3: Establishment of a joint temperature and humidity calibration and compensation model
[0207] The prepared flexible temperature and humidity sensing unit was placed in a controlled temperature and humidity test environment for joint calibration.
[0208] Multiple temperature calibration points and multiple relative humidity calibration points are set. Preferably, the temperature calibration range covers the actual operating environment of the cable joint, and the humidity calibration range covers normal humidity, high humidity, and near-saturation humidity conditions.
[0209] Under various temperature and humidity combinations, the resistance, capacitance, or impedance response of the interdigital humidity-sensitive electrode and the output of the serpentine thermoelectric electrode were recorded.
[0210] The coefficients a of the basic compensation model for temperature and humidity coupling at each node are fitted based on the calibration data. i0 To a i5 During the calibration process, humidity increment and humidity decrement processes were performed separately, and the hysteresis correction factor b was determined based on the response differences between the moisture absorption and desiccation processes. i,s and c i,s The calibration residuals at each node are used to determine the calibration reliability. The smaller the calibration residuals, the higher the calibration reliability.
[0211] After installation, when the cable joint is in a stable, non-condensing, and non-abnormal temperature rise state, a baseline data segment is collected to establish the node initial humidity baseline, surface temperature baseline, ambient reference temperature baseline, and adjacent node difference baseline.
[0212] During operation, the system identifies the status of each sensor node according to the aforementioned stable operating condition determination criteria. When a sensor node remains in a stable operating condition for a preset stable duration, its low-frequency drift baseline is slowly updated using a moving average, exponential smoothing, or restricted recursive update method, with each update not exceeding a preset baseline update limit. When the sensor node enters a local alert mode or high-frequency monitoring mode, or when the corresponding condensation risk index, overheating risk index, or condensation-overheating synergistic risk index reaches the pause update condition, the low-frequency drift baseline update of the sensor node is paused to prevent abnormal responses from being absorbed as normal baselines. Low-frequency drift baseline updates resume only after the sensor node recovers to a stable operating condition and maintains it continuously for a preset recovery time.
[0213] Example 4: Adaptive Acquisition and Data Alignment
[0214] After the system starts up, it first acquires initial sampling data at the first sampling frequency and enters normal inspection mode; after completing the initial temperature and humidity compensation and dew point calculation, the subsequent sampling mode is determined based on the correction data and risk status of the previous sampling cycle.
[0215] In normal inspection mode, all node data is collected every preset period. In local alert mode, the sampling frequency of abnormal nodes and their axial and circumferential adjacent nodes is increased, while periodic inspection of the remaining nodes is maintained. In high-frequency monitoring mode, node data in abnormal areas is collected at a third sampling frequency, while all nodes are scanned according to a preset period.
[0216] In one example, different operating modes can use the entry conditions, sampling methods, and exit conditions shown in Table 1.
[0217] Table 1 Different Working Modes
[0218] Work mode Entry conditions Sampling method Exit conditions Routine inspection mode Node risk is lower than level one Low-frequency polling of all nodes Any node triggers the initial threshold point-level condition Local vigilance mode Abnormal humidity change rate, temperature rise, or dew point margin Intermediate frequency acquisition abnormal nodes and adjacent nodes Risk recovery or further increase High-risk monitoring mode Collaborative risks exceed advanced thresholds High-frequency continuous acquisition of abnormal region nodes The risk continues to decline and the conditions for recovery are met.
[0219] When the relative humidity of a node increases rapidly, the dew point margin decreases, the surface temperature rises, or the spatial gradient of temperature and humidity reaches the alarm condition in the previous sampling period, the system switches the node and its axial and circumferential adjacent nodes to local alarm mode.
[0220] When the condensation risk index, overheating risk index, or condensation-overheating synergistic risk index of a node or abnormal area reaches the high-frequency monitoring conditions, the system enters the high-frequency monitoring mode; when the corresponding risk index is lower than the recovery threshold and continues to reach the preset recovery time, the sampling mode level is reduced.
[0221] The signal acquisition and adaptive control module adds node numbers and sampling timestamps to each frame of data. The data preprocessing and spatiotemporal alignment module establishes a target time axis, maps data with different sampling frequencies to a unified time series, and performs spatial registration according to axial and circumferential angular coordinates.
[0222] For short-term missing data, limited interpolation is performed using historical time-series data from the same node, and the interpolation marker is retained. For data that is continuously missing for more than a preset time, the data integrity reliability is reduced, and after compensation, the node reliability and reconstruction module performs labeled reconstruction based on adjacent high-reliability nodes and historical time-series data.
[0223] Example 5: Extraction of Condensation Risk Characteristics
[0224] Obtain the corrected relative humidity H at node i i Surface temperature T s,i and ambient reference temperature T a,i Then, calculate the dew point temperature T. d,i and dew point margin M d,i When M d,i When ≤0, the node is determined to satisfy the theoretical condensation condition; when 0 <M d,i <M safe When a node is determined to be in a near-condensation state, M is considered to be in such a state. safe This is a preset dew point safety margin.
[0225] In addition to the dew point margin, the rate of change of humidity is also calculated:
[0226]
[0227] Where L represents the calculation window length and Δt represents the sampling interval.
[0228] Calculate the duration of high humidity. When the relative humidity of a node is continuously higher than the high humidity threshold, the duration is accumulated; when the relative humidity returns to below the release threshold, it is reset to zero or decayed at a preset rate.
[0229] To identify humidity response saturation or abnormal recovery states that may be caused by localized condensation, a comprehensive analysis is performed on the degree of humidity response saturation, dew point margin, and desiccation recovery characteristics. Since the moisture-permeable, liquid-resistant protective layer is used to reduce the probability of droplets directly intruding into the gradient moisture-conducting humidity-sensitive layer and the interdigitated humidity-sensitive electrode, this embodiment does not directly interpret the output of the gradient moisture-conducting flexible humidity sensing unit as a liquid water contact signal.
[0230] A suspected liquid condensation indicator feature is generated when at least two of the following conditions are met simultaneously:
[0231] (1) The humidity electrical response reaches or approaches the preset saturation response threshold;
[0232] (2) The dew point margin of the corresponding sensing node is lower than the preset dew point margin threshold;
[0233] (3) After the relative humidity of the environment decreases or the dew point margin is restored, the humidity electrical response remains in a saturated state for a preset time.
[0234] (4) The recovery speed of the humidity electrical response is lower than the baseline recovery speed of the sensing node during normal dehumidification;
[0235] (5) The sensing node and at least one of its spatially adjacent nodes simultaneously exhibit high humidity, low dew point margin, or slow humidity response recovery.
[0236] The suspected liquid condensation indicator features are used as one of the input features of the condensation risk index, and together with dew point margin, high humidity persistence features, humidity growth rate, humidity spatial gradient and humidity anomaly area continuity, they are used for condensation risk assessment, but are not used alone as a definite criterion for the formation of liquid condensation.
[0237] Example 6: Extraction of Overheat Risk Features
[0238] Compute local temperature rise at the node:
[0239]
[0240] Calculate the surface temperature change rate:
[0241]
[0242] To differentiate between overall ambient temperature rise and localized joint overheating, the median surface temperature of each node within the same axial monitoring section was calculated, and the deviation of the node temperature from the median temperature of the section was also calculated.
[0243]
[0244] in, This represents the set of nodes located on the same axial monitoring section as node i.
[0245] When the temperature of all nodes rises simultaneously but the temperature difference between nodes remains small, the system tends to determine that the ambient temperature rises or the overall load increases; when the temperature of a few adjacent nodes is significantly higher than that of other nodes on the same cross section, the system tends to determine that the problem is a local contact anomaly or a local thermal defect.
[0246] Simultaneously, the duration of high temperature is accumulated; when the deviation of the corrected surface temperature or the load normalized temperature is continuously higher than the corresponding threshold, the duration is accumulated, and when it returns to below the release threshold, it is cleared or decayed according to the preset method.
[0247] Example 7: Node Trustworthiness and Collaboration Risk Assessment
[0248] The reliability of node signal quality is determined based on signal noise amplitude, saturation state, and abnormal pulse conditions; the reliability of node data integrity is determined based on packet loss rate and continuous missing time; the reliability of node drift is determined based on baseline drift under stable operating conditions; and the reliability of node change trend consistency is determined based on the short-term change trends, repeatability, and cross-physical quantity verification relationships between the node and its neighboring nodes. Each reliability component is weighted according to non-negative weights to obtain the node reliability.
[0249] When the humidity of a single node suddenly increases, but the humidity, dew point margin, and surface temperature of adjacent nodes do not change accordingly, and the noise level of the node increases or the data integrity decreases, the reliability of the signal quality, data integrity, or consistency of the trend of change of that node is reduced. When the anomaly occurs repeatedly in multiple consecutive sampling periods and is confirmed by dew point margin, temperature, or at least one adjacent node, the reliability is not directly reduced because its measurement amplitude deviates from the neighborhood mean.
[0250] When the humidity of a node and its adjacent nodes increases simultaneously, and the corrected surface temperature is close to or below the dew point temperature, the condensation risk index is increased; when the humidity response is saturated and accompanied by low dew point margin, persistent high humidity, or abnormally slow recovery, the risk contribution of suspected liquid condensation indicator characteristics is retained.
[0251] When the local temperature rise, temperature rise rate, temperature spatial gradient, or load normalized temperature deviation of a node and its adjacent nodes increases and continues to reach a preset time, the overheating risk index is increased.
[0252] When condensation risk and overheating risk coincide in the same location or are adjacent in the same functional area, the synergistic risk index is increased based on the spatial correlation and product coupling term of temperature and humidity anomalies; when both types of risks are high but the spatial correlation does not reach the synergistic threshold, they are identified as concurrent condensation anomalies and overheating anomalies, rather than synergistic anomalies.
[0253] For low-confidence nodes caused by communication loss, circuit break, or abnormal drift, labeled reconstruction is performed based on spatially adjacent high-confidence nodes and historical time-series data, and the direct weight of the reconstructed value in risk fusion is reduced; for response saturation nodes that may be caused by condensation, their original response and condensation indication characteristics are retained to avoid the spatial reconstruction from completely smoothing out the real anomaly.
[0254] Example 8: Anomaly Area Location and Hierarchical Early Warning
[0255] Nodes whose condensation risk index, overheating risk index, or condensation-overheating synergistic risk index reach the corresponding regional identification threshold are marked as candidate abnormal nodes.
[0256] Candidate anomalous nodes are divided into connected regions based on axial adjacency, circumferential adjacency, and diagonal adjacency. Nodes with circumferential angles near 0 degrees and 360 degrees are treated as adjacent nodes.
[0257] For each abnormally connected region, calculate the region's maximum risk, average risk, number of nodes, duration of the anomaly, axial expansion rate, and circumferential expansion rate.
[0258] Based on regional risk characteristics, anomalies are classified into the following types:
[0259] High humidity environment: Humidity increases in multiple areas at the same time, but the dew point margin is still within a safe range. The humidity change trend of each node is basically the same and the humidity spatial gradient is small, with no obvious local temperature rise.
[0260] Localized condensation type: The dew point margin at local nodes is close to or less than zero, the humidity spatial gradient is obvious, and the risk of condensation is concentrated in the sealing transition area or stress control transition area.
[0261] Localized overheating type: There is a significant local temperature rise, temperature rise rate or temperature spatial gradient, but there is no obvious abnormality in the corrected relative humidity and dew point margin;
[0262] Concurrent condensation and overheating type: Both condensation risk and overheating risk reach the corresponding warning conditions, but the spatial correlation of temperature and humidity anomalies does not reach the coordination threshold, so condensation warning and overheating warning are output respectively; Condensation-overheating coordination type: Both the condensation-overheating coordination risk index and the spatial correlation of temperature and humidity anomalies reach the corresponding coordination warning conditions.
[0263] Sensor anomaly type: Single node output is abnormal and node reliability is consistently low, while surrounding nodes and other physical quantities have no corresponding anomalies.
[0264] Early warning output can take the following forms:
[0265] "Level 2 Collaborative Early Warning; The abnormal area is located at the second axial section from the center of the joint to the left of the stress cone, with a circumferential angle of 90 to 135 degrees; the maximum surface temperature is the current temperature of the corresponding node; the maximum relative humidity is the current humidity of the corresponding node; the minimum dew point margin is the calculation result of the corresponding node; it is recommended to check the joint sealing status and conductor connection status."
[0266] Example 9: Operation process of the early warning system
[0267] After the system is powered on, it first reads the calibration parameters, spatial coordinates, historical baselines and warning thresholds of each node.
[0268] The signal acquisition and adaptive control module sequentially acquires data from the flexible temperature and humidity heterogeneous array, determines the sampling mode based on the initial sampling data or the state of the previous sampling period, and then sends the data to the data preprocessing and spatiotemporal alignment module.
[0269] The data preprocessing and spatiotemporal alignment module completes abnormal pulse removal, data integrity marking, time alignment, and spatial registration; the temperature and humidity compensation module sequentially completes temperature measurement self-heating compensation, installation heat transfer deviation compensation, temperature and humidity cross-sensitivity compensation, moisture absorption-dehumidification hysteresis correction, and low-frequency drift correction.
[0270] The node credibility and reconstruction module calculates the signal quality credibility, data integrity credibility, drift credibility, and trend consistency credibility of each node, and performs labeled spatial reconstruction on low-credibility nodes that meet the reconstruction conditions.
[0271] The feature extraction and risk assessment module calculates dew point temperature, dew point margin, local temperature rise, temperature and humidity change rate, temperature and humidity spatial gradient, high humidity duration, low dew point margin duration, high temperature duration, spatial continuity of abnormal areas, and spatial correlation of temperature and humidity anomalies.
[0272] The feature extraction and risk assessment module further calculates the node-level condensation risk index, overheating risk index, and condensation-overheating synergistic risk index, and forms corresponding abnormal candidate regions.
[0273] The positioning and hierarchical early warning module divides abnormal connected areas according to the spatial distribution of various risks, and determines the abnormal type, early warning level and abnormal location. When both condensation risk and overheating risk meet the early warning conditions but the spatial correlation is insufficient, two types of early warnings are output respectively. When both the coordinated risk index and spatial correlation meet the coordinated conditions, a condensation-overheating coordinated early warning is output and uploaded to the cable condition monitoring platform through the communication module.
[0274] After the risk is eliminated, the system does not immediately cancel the warning. Instead, it sets recovery thresholds and recovery durations corresponding to various risks. Only when the corresponding risk index is lower than the recovery threshold and continues to reach the preset time will the system reduce the warning level or sampling mode level to avoid frequent switching of the warning status.
[0275] The above description is merely a preferred embodiment of the present invention. Any equivalent substitutions or adjustments made by those skilled in the art to the flexible substrate material, humidity-sensitive material, conductive material, gradient structure formation method, number of array nodes, compensation model form, risk calculation method, and early warning threshold without departing from the principles and concept of the present invention should fall within the protection scope of the present invention.
Claims
1. A method for coordinated early warning of condensation and overheating in cable joints based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array, characterized in that, Includes the following steps: Step S1: Provide a flexible temperature and humidity heterogeneous array including multiple gradient moisture-conducting flexible humidity sensing units and multiple flexible surface temperature sensing units. Determine at least one spatially associated gradient moisture-conducting flexible humidity sensing unit and at least one flexible surface temperature sensing unit as a sensing node. Attach and arrange multiple sensing nodes along the axial and circumferential directions of the cable joint. Set at least one environmental reference temperature node and / or an environmental reference temperature and humidity node to establish a spatial coordinate mapping relationship between each sensing node and the surface position of the cable joint. Step S2: Based on the cable operating status and the initial sampling data during the system startup phase or the corrected surface temperature, corrected relative humidity, temperature and humidity change rate and dew point margin obtained in the previous sampling cycle, adaptively determine the sampling mode of the flexible temperature and humidity heterogeneous array, and collect the original temperature signal and original humidity response signal output by each sensing node. Step S3: Perform abnormal pulse removal, data integrity marking and filtering on the original temperature signal and the original humidity response signal, and perform time alignment and spatial registration according to the sampling timestamp and spatial coordinates to obtain spatiotemporally aligned temperature and humidity array data; Step S4: Perform temperature self-heating compensation, installation heat transfer deviation compensation, temperature and humidity cross-sensitivity compensation, moisture absorption-dehumidification hysteresis compensation and low frequency drift compensation on the spatiotemporally aligned temperature and humidity array data to obtain the preliminary corrected surface temperature and preliminary corrected relative humidity of each sensing node; Step S5: Determine the node reliability based on the signal quality, data integrity, drift reliability, and consistency of the change trend of adjacent nodes for each sensor node; For low-confidence nodes caused by communication loss, circuit breakage or abnormal drift, spatial reconstruction with confidence labeling is performed based on adjacent high-confidence nodes and historical time series data to obtain the corrected surface temperature field and corrected relative humidity field. Step S6: Calculate the dew point temperature based on the corrected relative humidity and its corresponding local environmental reference temperature, and calculate the dew point margin based on the difference between the corrected surface temperature and the dew point temperature; extract the local temperature rise rate, local humidity increase rate, axial temperature and humidity gradient, circumferential temperature and humidity gradient, high humidity persistence characteristics, low dew point margin persistence characteristics, high temperature persistence characteristics, spatial continuity of abnormal areas, and spatial correlation of temperature and humidity anomalies. Step S7: Calculate the condensation risk index, overheating risk index, and condensation-overheating synergistic risk index for each sensing node based on the aforementioned features and node reliability. Divide spatially continuous abnormal nodes into abnormal regions, and determine the abnormal type, abnormal level, and abnormal location based on the risk index and dynamic early warning threshold. When both condensation risk and overheating risk meet the early warning conditions but the spatial correlation of temperature and humidity anomalies does not meet the synergistic conditions, output condensation early warning signals and overheating early warning signals respectively. When both the condensation-overheating synergistic risk index and the spatial correlation of temperature and humidity anomalies meet the synergistic early warning conditions, output a condensation-overheating synergistic early warning signal.
2. The method according to claim 1, characterized in that: The gradient moisture-conducting flexible humidity sensing unit includes a flexible insulating substrate, an interdigitated humidity-sensitive electrode disposed on the flexible insulating substrate, a gradient moisture-conducting humidity-sensitive layer covering the interdigitated humidity-sensitive electrode, a moisture-permeable and liquid-resistant protective layer disposed outside the gradient moisture-conducting humidity-sensitive layer, and a flexible encapsulation layer covering the non-detection area. The flexible surface temperature sensing unit includes a serpentine temperature-sensing resistance electrode disposed on a flexible insulating substrate and an insulating and thermally conductive encapsulation layer covering the serpentine temperature-sensing resistance electrode. The interdigitated humidity-sensitive electrode and the serpentine temperature-sensing resistor electrode are disposed alternately on the same flexible insulating substrate, or are disposed on spatially adjacent flexible insulating substrates and are isolated from each other by an electrical insulation isolation structure.
3. The method according to claim 2, characterized in that: The gradient moisture-conducting and moisture-sensitive layer includes a moisture-absorbing polymer matrix and moisture-sensitive functional materials and / or conductive sensitive materials dispersed in the moisture-absorbing polymer matrix, and has a pore structure gradient and / or hydrophilicity gradient that varies from the moisture-permeable and liquid-blocking protective layer side toward the interdigitated moisture-sensitive electrode side. The pore structure gradient includes an average pore size that gradually decreases along the direction toward the interdigitated humidity-sensitive electrode, and the hydrophilic gradient includes a hydrophilic group density or hygroscopic material content that gradually increases along the direction toward the interdigitated humidity-sensitive electrode, so as to form a water vapor transport and adsorption potential gradient toward the humidity-sensitive detection area. The moisture-permeable and liquid-blocking protective layer allows water vapor to pass through and prevents droplets larger than a preset size from directly entering the gradient moisture-conducting and humidity-sensitive layer, thereby reducing the probability of liquid condensation directly wetting the gradient moisture-conducting and humidity-sensitive layer and the interdigitated humidity-sensitive electrode.
4. The method according to claim 1, characterized in that, The multiple sensing nodes are attached and arranged along the axial and circumferential directions of the cable joint, specifically including: Multiple axial monitoring sections are set along the conductor connection area, stress control transition area, insulation recovery area and sealing transition area of the cable joint; Multiple sensing nodes are set at circumferential intervals on each axial monitoring section to form a circumferential monitoring ring. The sensing nodes corresponding to the circumferential positions of adjacent axial monitoring sections are formed into an axial monitoring chain, so that multiple circumferential monitoring rings and multiple axial monitoring chains together constitute a circumferential-axial two-dimensional flexible temperature and humidity heterogeneous array. The position of each sensing node is represented as node coordinates including axial coordinates and circumferential angular coordinates. A local environmental reference temperature node is set at least one axial monitoring section that is thermally isolated from the surface of the cable joint. At least one environmental reference temperature and humidity node is set outside the cable joint.
5. The method according to claim 1, characterized in that, The sampling modes include normal inspection mode, local alert mode, and high-frequency monitoring mode; In normal inspection mode, data from all sensor nodes are periodically collected at the first sampling frequency; When the temperature change rate, humidity change rate, temperature and humidity spatial gradient, or dew point margin of any sensor node in the previous sampling period reaches the corresponding alarm condition, the local alarm mode is entered, and data of the sensor node and its axial and circumferential adjacent nodes are collected at a second sampling frequency higher than the first sampling frequency, while periodic inspection of the remaining sensor nodes is maintained. When the condensation risk index, overheating risk index, or condensation-overheating synergistic risk index of any sensing node or abnormal area reaches the corresponding high-frequency monitoring condition, the high-frequency monitoring mode is entered, and data of the abnormal area node is collected at a third sampling frequency higher than the second sampling frequency, and all sensing nodes are scanned according to a preset cycle. When the corresponding risk indicator falls below the recovery threshold and continues to reach the preset recovery time, the sampling mode level is reduced.
6. The method according to claim 1, characterized in that: The self-heating temperature rise is calculated based on the excitation current, temperature measuring resistance, and calibrated equivalent self-heating thermal resistance of the flexible surface temperature sensing unit. The marked installation heat transfer correction is determined based on the installation thermal resistance, local ambient reference temperature, and dynamic thermal response parameters to obtain the corrected surface temperature. The corrected relative humidity of the i-th sensing node at time k is calculated according to the following formula: Where, x i (k) represents the normalized raw humidity response signal, T h,i (k) is the temperature of the humidity-sensitive layer or its estimated value, a i0 To a i5 This refers to the joint calibration coefficient for temperature and humidity. d is the hysteresis correction term corresponding to the moisture absorption or desiccation process. i (k) is the low-frequency drift estimate, H min The preset lower limit for humidity calculation is greater than zero; the signs of the moisture absorption and dehumidification processes are determined by the product of the calibrated humidity response direction coefficient and the change in humidity response at adjacent sampling times.
7. The method according to claim 6, characterized in that, The dew point temperature and dew point margin of the i-th sensing node at time k are calculated according to the following formula: Among them, T a,i (k) represents the local environmental reference temperature corresponding to the i-th sensing node, T s,i (k) is for correcting surface temperature, T d,i (k) is the dew point temperature, M d,i (k) represents the dew point margin, and a and b are the dew point calculation coefficients; the high humidity duration feature is the duration during which the corrected relative humidity is continuously higher than the preset relative humidity threshold; the low dew point margin duration feature is the duration during which the dew point margin is continuously lower than the preset dew point margin threshold; and the high temperature duration feature is the duration during which the corrected surface temperature or load normalized temperature deviation is continuously higher than the corresponding threshold.
8. The method according to claim 1, characterized in that, Node credibility is calculated using the following formula: Where, q i,sig For signal quality reliability, q i,data For data integrity and reliability, q i,drift For drift confidence, q i,trend The reliability of the change trend between the sensing node and its axial and circumferential adjacent nodes is defined. The value of each reliability component ranges from 0 to 1, with larger values indicating higher reliability. w1 to w4 are non-negative weights, and their sum is 1. When the node reliability is lower than the preset reliability threshold and the reason for the low reliability is data loss, communication anomaly, or abnormal drift, the node is spatially reconstructed based on the reliability of adjacent high-reliability nodes, spatial distance, and historical time series data, and the low-reliability marker and reconstruction marker are retained. When the humidity response saturation is accompanied by a decrease in dew point margin, persistent high humidity, or slow abnormal recovery, the humidity response saturation is retained as a condensation risk feature.
9. The method according to claim 1, characterized in that: The condensation risk index is obtained by normalizing and weighting the dew point margin, the rate of decrease of the dew point margin, the characteristics of high humidity persistence, the characteristics of low dew point margin persistence, the rate of humidity increase, the humidity spatial gradient, and the continuity of humidity anomaly regions; the overheating risk index is obtained by normalizing and weighting the corrected surface temperature, local temperature rise, local temperature rise rate, load-normalized temperature deviation, temperature spatial gradient, high temperature persistence, and the continuity of temperature anomaly regions. The condensation-overheating synergistic risk index is calculated according to the following formula: Among them, R c,i (k) is the node condensation risk index, R h,i (k) is the node overheating risk index, A i (k) represents the spatial correlation degree of temperature and humidity anomalies determined based on the degree of overlap and / or proximity of temperature and humidity anomaly regions, λ c , λ h , λ ch and λ a The weights are non-negative and their sum is 1. A condensation warning is output when only the condensation risk index reaches the condensation warning threshold, and an overheating warning is output when only the overheating risk index reaches the overheating warning threshold. When both the condensation risk index and the overheating risk index reach their respective warning thresholds, but the spatial correlation of temperature and humidity anomalies does not reach the synergy threshold, a condensation warning and an overheating warning are output respectively. When both the condensation-overheating synergy risk index and the spatial correlation of temperature and humidity anomalies reach their respective synergy warning thresholds, a condensation-overheating synergy warning is output.
10. A cable joint condensation and overheating coordinated early warning system based on a gradient moisture-conducting flexible temperature and humidity heterogeneous array, characterized in that, include: The flexible temperature and humidity heterogeneous array includes multiple gradient moisture-conducting flexible humidity sensing units, multiple flexible surface temperature sensing units, and at least one environmental reference temperature node and / or environmental reference temperature and humidity node arranged along the axial and circumferential directions of the cable joint. The signal acquisition and adaptive control module is used to acquire raw temperature signals and raw humidity response signals, and switch the sampling mode according to the initial sampling data or the risk status of the previous sampling period. The data preprocessing and spatiotemporal alignment module is used for outlier removal, data integrity marking, temporal alignment, spatial registration, and establishment of circular adjacency relationships. The temperature and humidity compensation module is used to implement temperature measurement self-heating compensation, installation heat transfer deviation compensation, temperature and humidity cross-sensitivity compensation, moisture absorption-dehumidification hysteresis compensation and low frequency drift compensation; The node credibility and reconstruction module is used to calculate node credibility and perform spatial reconstruction with reconstruction tags on low-credibility nodes that meet the reconstruction conditions. The feature extraction and risk assessment module is used to calculate dew point temperature, dew point margin, temperature and humidity change rate, temperature and humidity spatial gradient, anomaly persistence characteristics, spatial continuity of anomaly area and spatial correlation of temperature and humidity anomalies, and to calculate node-level condensation risk index, overheating risk index and condensation-overheating synergistic risk index. The positioning and hierarchical early warning module is used to divide abnormal connected areas, determine the abnormal type, abnormal level, axial position and circumferential position, and output condensation early warning signal, overheating early warning signal or condensation-overheating combined early warning signal; The communication module is used to send early warning information to the cable condition monitoring platform or maintenance terminal.
