Anti-loosening monitoring method for electric power fitting wire clamp

Abstract

The invention relates to an anti-loose monitoring method for electric power fitting wire clamps, which comprises the following steps of: injecting a high-frequency probe signal into a preset conductor of a power transmission line serving as a sensing medium from a detection end, and receiving a reflected signal sequence generated by a plurality of electric power fitting wire clamps along the line due to discontinuous impedance; the round-trip time of each reflected signal in the sequence is analyzed to accurately position each wire clamp, and the reflection characteristics of the wire clamps are extracted to represent the real-time connection state. And comparing the real-time reflection characteristic with a pre-stored baseline reflection characteristic of the wire clamp in a fastening state, and determining that the wire clamp at the corresponding position is loosened when the difference between the real-time reflection characteristic and the pre-stored baseline reflection characteristic exceeds a preset threshold value. According to the invention, remote, passive and full-coverage monitoring of the fastening state of a plurality of electric power fitting wire clamps on a power transmission line can be realized, and the technical problems of low manual inspection efficiency and strong subjectivity, and high cost, difficult power supply, complex maintenance and the like of an online monitoring device in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of power fitting operation monitoring technology, and in particular to a method for monitoring the loosening of power fitting clamps. Background Technology

[0002] Overhead transmission lines, as the primary carriers of power transmission, are crucial for safe and stable operation. Within the transmission line structure, power clamps, as key components connecting conductors to insulator strings and securing conductors to towers, are numerous and widely distributed. However, these clamps are constantly exposed to the elements, enduring complex environments such as wind, rain, snow, alternating temperatures, electromagnetic forces, and mechanical vibrations, making them highly susceptible to problems like loose bolts and oxidation of contact surfaces. Loosening of clamps not only reduces mechanical connection strength, potentially leading to serious accidents like conductor detachment in severe weather, but also causes significant overheating under high current due to increased contact resistance, resulting in overheating defects. This can range from energy loss to burning out clamps and conductors, causing widespread power outages and posing a serious threat to power grid safety. Therefore, effectively monitoring the tightness of power clamps and promptly identifying and addressing potential loosening is an indispensable part of ensuring the safe operation of transmission lines.

[0003] Currently, there are two main technical approaches for monitoring the loosening of power fitting clamps. The first is the traditional manual inspection method. Maintenance personnel need to periodically inspect the lines using binoculars, infrared thermometers, or drones equipped with high-definition and infrared cameras. While this method is intuitive, its drawbacks are significant: firstly, it is inefficient and labor-intensive, making it difficult to achieve high-frequency, all-weather coverage for lines that are often tens or hundreds of kilometers long; secondly, it is severely constrained by weather conditions, making operation impossible in strong winds, rain, snow, or fog; and thirdly, the inspection results rely heavily on the experience and responsibility of the personnel, making them highly subjective. Furthermore, early, internal, or non-heat-related minor loosenings are difficult to detect effectively based solely on appearance or temperature, posing a high risk of missed detections.

[0004] To overcome the shortcomings of manual inspection, a second technological approach has emerged in recent years: directly installing online monitoring sensors on the clamps. For example, installing bolt loosening monitoring devices using strain gauges or angular displacement sensors, or installing temperature sensors for overheat monitoring. These devices transmit data back to the monitoring center via wireless communication modules (such as GPRS, 4G, LoRa, etc.). While this approach achieves online and automated monitoring, it has exposed new and more challenging problems in large-scale applications. First, there is the power supply challenge. Tens of thousands of clamps are distributed along high-voltage conductors, making independent power supply extremely difficult. Battery power has limited lifespan and requires significant replacement and maintenance work; using line induction or solar power increases the complexity and cost of the device, and reliability is difficult to guarantee during low loads or continuous rainy days. Second, there are communication and networking challenges. Each clamp requires a wireless communication unit, and networking a massive number of nodes, channel conflicts, and communication reliability under high-voltage and strong electromagnetic environments present significant technical challenges. Third, there are cost and maintenance challenges. Each clamp requires a complete set of sensing, processing, communication and power supply units, resulting in extremely high initial investment costs. Furthermore, the large number of electronic devices operate in harsh environments for extended periods, leading to high failure rates and an overwhelming burden of maintenance in the later stages.

[0005] In summary, existing technologies, whether relying on traditional manual inspections or online monitoring using distributed sensors, have inherent limitations and cannot simultaneously meet the requirements of low cost, maintenance-free operation, full coverage, and high reliability. Therefore, there is an urgent need in this field for a technological innovation that can break free from the constraints of installing equipment on each clamp and achieve remote centralized monitoring of the tightness status of all hardware clamps along the entire transmission line in a more economical, efficient, and reliable manner. Summary of the Invention

[0006] To address the technical problems of the prior art, this invention provides a method for monitoring the loosening of power fitting clamps. This method enables remote, passive, and full-coverage monitoring of the tightness of multiple power fitting clamps on transmission lines, overcoming the technical difficulties of low efficiency and strong subjectivity of manual inspections, as well as the high cost, power supply difficulties, and complex maintenance of online monitoring devices in the prior art.

[0007] This invention discloses a method and system for monitoring the loosening of electrical fitting clamps, comprising the following steps:

[0008] First, in step 1, a high-frequency probe signal is injected from a probe end into a pre-designed conductor of the transmission line, which serves as the sensing medium. This step is the starting point and physical basis of the entire monitoring process; it actively applies a known excitation signal to the system under test (i.e., the transmission line). The pre-designed conductor, such as the metal sheath of an optical fiber composite overhead ground wire or a regular overhead ground wire, is innovatively used as a signal propagation channel because of its stable potential and lack of interference with the main transmission circuit, transforming the entire transmission line into a vast distributed sensing network.

[0009] Subsequently, in step 2, the probe end receives reflected signals generated by impedance discontinuities when the probe signal encounters multiple power fitting clamps during its propagation along the line, thus forming a sequence of reflected signals. This step works closely with step 1 to form a complete "transmission-reflection-reception" detection loop. Each power fitting clamp, due to its physical structure and connection interface with the conductor, constitutes an impedance point different from the uniform conductor body. When the high-frequency probe signal propagates to this point, some energy will inevitably be reflected. Therefore, the received sequence of reflected signals, like a one-dimensional "radar image," objectively maps the spatial distribution and physical state of all fitting clamps along the line.

[0010] Next, in step 3, the received reflected signal sequence is processed. The position of the corresponding electrical fitting clamp is determined based on the round-trip time of each reflected signal, and real-time reflection features characterizing its connection status are extracted. This analysis step is the core of the localization and qualitative analysis, transforming the raw electrical signals obtained in step 2 into valuable physical information. On one hand, based on the principle that signals propagate at approximately the speed of light, by accurately measuring the round-trip time of each reflected signal from transmission to reception, the distance between the clamp generating the reflection and the detection end can be accurately calculated, achieving complete localization of all clamps along the entire line. On the other hand, by analyzing the waveform details of the reflected signals, such as amplitude, energy, pulse width, or phase, one or a set of real-time feature vectors that quantify the current mechanical and electrical connection status of the clamps can be extracted.

[0011] Furthermore, in step 4, the real-time reflection characteristics of each power fitting clamp are compared with the pre-stored baseline reflection characteristics of power fitting clamps in the same location under a tightened state. This step introduces a diagnostic benchmark, transforming monitoring from a one-time absolute measurement to a relative diagnosis based on state changes. The baseline reflection characteristics are collected and stored under a known healthy line condition (all clamps are tightened), representing the ideal reflection pattern that each clamp "should" have.

[0012] Finally, in step 5, when the difference generated by the comparison in step 4 exceeds a preset looseness judgment threshold, it is determined that the corresponding power fitting clamp is loose. This judgment step is the decision output link of the entire method. A tight clamp should maintain stable reflection characteristics over a long period of time; however, once loosening occurs, its contact interface will undergo slight changes, leading to a change in its impedance characteristics. This change will be directly and significantly reflected in its real-time reflection characteristics, thereby increasing the difference from the baseline characteristics (e.g., obtained by calculating Mahalanobis distance or other vector distances). By setting a reasonable threshold, this significant difference can be identified as a loosening defect, and an alarm can be output. These five steps are interconnected, from signal excitation, data acquisition, information processing, to benchmark comparison and final diagnosis, forming a logically rigorous and fully functional remote monitoring system.

[0013] Further, the high-frequency probe signal is a Gaussian pulse or raised cosine pulse with a pulse width of 2ns to 10ns for time-domain reflectometry (TDR) mode; or, the high-frequency probe signal is a linear frequency modulated (LFM) signal with a scanning bandwidth of 50MHz to 500MHz for frequency-domain reflectometry (FDR) mode. Specifically, this scheme defines two preferred forms of high-frequency probe signals: nanosecond-level pulse signals for time-domain reflectometry (TDR) mode and broadband LFM signals for frequency-domain reflectometry (FDR) mode. In a preferred embodiment, a Gaussian pulse or raised cosine pulse with a pulse width of 2ns to 10ns is selected as the probe signal. This pulse width directly determines the spatial resolution of the system, i.e., the minimum distance at which two adjacent reflection points can be distinguished along the line. Theoretically, spatial resolution is proportional to the product of pulse width; therefore, a narrower pulse (e.g., 2ns to 10ns) can achieve meter-level spatial resolution, which is sufficient to clearly distinguish power fitting clamps at different locations or in different phases on the same base tower. Gaussian pulses or raised cosine pulses are chosen because these pulse waveforms have concentrated energy in the time domain and low sidelobes in the frequency domain, reducing signal spectral leakage, thereby improving signal quality and reducing potential interference to other communication frequency bands. In another preferred approach, a linear frequency modulated signal with a scanning bandwidth of 50MHz to 500MHz is used. In this frequency domain reflection mode, spatial resolution is inversely proportional to the signal's scanning bandwidth. Therefore, a wider scanning bandwidth (e.g., 50MHz to 500MHz) can also achieve meter-level or sub-meter-level high spatial resolution. By clearly defining the key parameters of the probe signal, this invention ensures that the method has sufficiently high accuracy when performing positioning tasks, which is the basis for subsequent accurate state determination. Its technical effect is that by using probe signals with specific time-domain or frequency-domain parameters, high-resolution spatial positioning of densely distributed power fitting clamps on transmission lines is achieved, ensuring that the system can accurately correspond each reflected signal to a unique physical clamp, avoiding position confusion and providing a reliable prerequisite for subsequent accurate single-point state comparison.

[0014] Further, in step 3, processing the reflected signal sequence includes: before extracting the real-time reflection features, performing point-by-point alignment and coherent accumulation on multiple consecutive reflected signal sequences synchronously acquired from the detection end, and then performing digital bandpass filtering on the accumulated signal. Specifically, this scheme aims to improve signal quality. Before extracting the real-time reflection features, signal enhancement processing is required on multiple consecutive reflected signal sequences synchronously acquired from the detection end. The specific processing includes two core operations: first, point-by-point alignment and coherent accumulation, and then performing digital bandpass filtering on the accumulated signal. Because the reflected signals generated by power fitting clamps, especially those from distant locations or clamps with only slight loosening, are usually very weak and easily drowned out by background noise, the coherent accumulation technique utilizes the difference in statistical characteristics between signal and noise: the target reflected signal is deterministic, maintaining consistent shape and phase in each measurement, while background noise is random. By precisely aligning and summing multiple measured signal sequences point by point in time, the amplitude of the target signal increases linearly by N times, while the power of random noise only increases by N times (amplitude increases by sqrt(N) times). Therefore, the signal-to-noise ratio (SNR) after accumulation can theoretically be improved by sqrt(N) times. After coherent accumulation significantly improves the SNR, a digital bandpass filter, whose passband range matches the spectrum of the injected probe signal, can further filter out residual out-of-band noise and deterministic interference such as power frequency. The logic of this series of processes is to first "extract" the weak signal from random noise through accumulation, and then "purify" the signal through filtering. The technical effect is that the combined application of coherent accumulation and digital filtering greatly improves the SNR of the reflected signal sequence, enabling the clear detection of weak reflected signals from distant or early minor loosening defects that were previously submerged by noise. This significantly enhances the sensitivity and maximum monitoring distance of the monitoring method and reduces the possibility of missed detection.

[0015] Furthermore, prior to the comparison in step 4, a time axis calibration step is included. This step involves: selecting stable reflection peaks in amplitude and waveform generated by the tower body from the baseline reflection characteristics as anchor points; determining the peak position by calculating the cross-correlation function between the real-time reflection signal sequence and the anchor points to calculate the overall time axis shift; and using the shift to compensate and correct the time axis of the entire real-time reflection signal sequence. Specifically, this scheme proposes a pre-comparison calibration method to eliminate interference from environmental factors. Since the physical length of transmission lines expands and contracts with changes in ambient temperature, even small changes can cause an overall shift in the time axis of the entire reflection signal sequence. Without correction, this time drift can be incorrectly interpreted as a change in clamp status, leading to false alarms. Therefore, this method introduces a time axis calibration step. This step selects the most stable physical structures in the line with reflection signal characteristics (such as amplitude and waveform), such as the tower body, as reference "anchor points." The reflection characteristics of these anchor points are pre-stored in the baseline database. When processing real-time acquired reflection signal sequences, the system calculates the cross-correlation function between anchor point reflections in the real-time signal and anchor point reflections in the baseline to accurately determine the time delay between them. This delay is the overall shift of the entire time axis. Once this shift is determined, the system performs uniform compensation and correction on the time axis of the entire real-time reflection signal sequence to ensure precise temporal alignment with the baseline data. The technical advantage lies in effectively compensating for time reference drift caused by changes in line length due to variations in ambient temperature by introducing a time axis calibration mechanism based on stable anchor points and cross-correlation calculations. This ensures that subsequent comparisons between real-time features and baseline features are performed in an aligned and reliable coordinate system, thereby separating feature distortions caused by real changes in physical state from spurious changes caused by environmental factors. This significantly improves the accuracy of diagnostic results and fundamentally reduces the false alarm rate caused by environmental interference.

[0016] Further, in step 3, the process of extracting the real-time reflection features includes: for each located power fitting clamp, extracting one or more feature parameters of its reflected signal waveform to form a real-time feature vector; the feature parameters are selected from: peak amplitude, pulse energy obtained by integrating the pulse waveform, pulse full width at half maximum (FWHM), pulse rise time, pulse fall time, pulse symmetry coefficient, and the ringing factor defined by the energy ratio of subsequent small peaks to the main peak; in step 4, the comparison step specifically involves: retrieving the baseline feature vector corresponding to the position of the power fitting clamp from a pre-stored database; using the inverse matrix of the covariance matrix of the feature vector obtained statistically when the power fitting clamp is in a tightened state, calculating the Mahalanobis distance between the real-time feature vector and the baseline feature vector as the difference. Specifically, firstly, this scheme clarifies the refinement of state representation, that is, it no longer relies on a single parameter, but instead extracts multiple feature parameters of the reflected signal waveform to construct a multi-dimensional "feature vector". These parameters, such as peak amplitude, pulse energy, pulse width, rise / fall time, symmetry, and ringing factor, collectively depict the complete shape of the reflected waveform from different perspectives. A loose clamp can not only cause amplitude changes but also waveform broadening or subsequent oscillations; this multi-dimensional characterization can more comprehensively capture subtle signs of a fault. Secondly, this scheme specifies a more advanced and reliable method for measuring difference—the Mahalanobis distance. Unlike simple Euclidean distance, Mahalanobis distance utilizes the pre-statistically obtained covariance matrix of each characteristic parameter under healthy conditions when calculating difference. This means it considers not only the normal fluctuation range of each characteristic itself but also the correlation between characteristics. For example, amplitude and energy are inherently highly correlated; Mahalanobis distance can eliminate the influence of this correlation, thus providing a fairer and more statistically significant measure of "abnormality." Its technical advantages lie in the fact that by constructing multi-dimensional feature vectors, it achieves a comprehensive and detailed description of the connection status of wire clamps, thereby improving the ability to perceive complex fault modes. At the same time, by using Mahalanobis distance as a comparison criterion, it establishes a more statistically rigorous and reliable difference quantification system. This system can effectively avoid the interference caused by the different dimensions and inherent correlations of different feature parameters, making the threshold setting for loosening judgment more scientific, and ultimately improving the sensitivity and accuracy of fault diagnosis.

[0017] Furthermore, a passive impedance characteristic element with a preset high-frequency resonant characteristic is pre-fixed on each power fitting clamp. This characteristic element is either a spiral resonant ring or a stub array. The baseline reflection characteristics and real-time reflection characteristics extracted in step 3 are features of the resonant reflection signal with a specific morphology generated by the characteristic element. In step 5, the loosening is determined based on the characteristic distortion caused by the change in the quality factor Q value of the resonant reflection signal. Specifically, this scheme pre-fixes a passive impedance characteristic element with specific high-frequency resonant characteristics, such as a spiral resonant ring or a stub array composed of metal strips of a specific length, on each power fitting clamp. These elements themselves do not affect the power frequency current, but they generate an extremely clear and stable resonant reflection signal for the injected high-frequency probe signal. In this way, the monitoring target shifts from the relatively blurry, weak, and inconsistent natural reflection of the clamp itself to monitoring this clear, strong, and highly standardized "impedance fingerprint" signal. The clamp's tightness directly determines the quality of its mechanical and electrical connection with the characteristic element. Once the clamp becomes loose, it introduces minute contact resistance and gap capacitance between itself and the feature element, severely damaging the quality factor (Q value) of the resonant circuit. A sharp drop in Q value directly leads to drastic distortion of the characteristics of its resonant reflected signal, such as a significant attenuation of the reflection amplitude and waveform broadening in the time domain, or a shallower and wider resonance peak in the frequency domain. The technical advantage lies in introducing a high-Q resonant characteristic "out of thin air," transforming a difficult problem of detecting changes in weak natural signals into a simple problem of detecting significant distortions in strong characteristic signals. This not only greatly improves the signal-to-noise ratio of the signal under test, making detection almost unaffected by background noise, but also significantly amplifies the fault characteristics, allowing even extremely small early loosenings to trigger easily identifiable signal changes. Ultimately, this solution fundamentally improves the sensitivity, reliability, and anti-interference capability of the monitoring method.

[0018] Furthermore, the execution of step 1 is triggered by environmental disturbance data. The triggering steps include: a linkage step: acquiring real-time disturbance data characterizing the environmental state of the line from an external environmental monitoring system, the disturbance data including wind speed or line vibration data; a triggering step: when the disturbance data exceeds a preset disturbance threshold, triggering the detection end to continuously execute steps 1 to 3 multiple times at a preset high-density frequency to obtain a set of real-time reflection characteristics that change over time for each power fitting clamp; a dynamic analysis step: in step 5, calculating the standard deviation or peak-to-peak value of a certain characteristic parameter in the set of real-time reflection characteristics as a function of time, and when the calculation result exceeds a dynamic jitter threshold, determining that the corresponding power fitting clamp is loose. This scheme intelligently links monitoring activities with the real-time environmental conditions of the power line. Specifically, the monitoring system no longer performs detection at fixed time intervals, but instead acquires disturbance data characterizing the environmental conditions of the power line in real time from external environmental monitoring systems (such as weather stations, anemometers, or distributed vibration sensing systems integrated into the OPGW) through a linkage step. This data can specifically be wind speed or line vibration data. When the system determines that these real-time disturbance data, such as instantaneous wind speed, exceed a preset disturbance threshold, it triggers the detection end to continuously inject, receive, and process probe signals at a preset high-density frequency, such as several to dozens of times per second, thereby obtaining a set of real-time reflection characteristic sequences that change rapidly over time for each power fitting clamp position. When judging looseness, this method introduces a dynamic analysis step, which no longer simply compares the static difference between a single measurement result and the baseline, but calculates the standard deviation or peak-to-peak value of a specific characteristic parameter (such as the amplitude or energy of the reflection peak) in the set of real-time reflection characteristic sequences obtained during the disturbance period. When the calculated result representing the degree of "dynamic jitter" exceeds a preset dynamic jitter threshold, the system determines that the corresponding power fitting clamp is loose. The technical advantage of this solution lies in its clever use of natural environmental forces as a free "dynamic stress test" applied to the line. A mechanically secure clamp should maintain highly stable electrical contact characteristics even under disturbances such as strong winds; therefore, its reflection characteristic sequence will exhibit extremely low temporal fluctuations. Conversely, a clamp with early or slight loosening may have acceptable contact under static conditions and be difficult to detect. However, under the influence of environmental forces, its contact interface will experience minute, high-frequency displacement and vibration, leading to drastic, instantaneous changes in its contact resistance and impedance characteristics. This change directly modulates the reflected signal, making it exhibit significant "jitter." This method, by conducting high-density acquisition and analyzing its dynamic characteristics during periods of high disturbance, successfully transforms a difficult-to-detect static micro-defect into a easily quantifiable dynamic characteristic with drastic fluctuations in signal amplitude or energy over time.This greatly improves the sensitivity and reliability of detecting intermittent or early loosening defects that do not show obvious abnormalities under static conditions.

[0019] Furthermore, the method includes an additional step for diagnosing overheating defects. This additional step comprises: Step a, cold-state feature acquisition: In response to a line energizing command, when the line is in a cold state at ambient temperature, steps 1 to 3 are executed to acquire the cold-state reflection characteristics of each power fitting clamp; Step b, hot-state feature acquisition: After the line bears load and reaches a thermodynamically stable state, steps 1 to 3 are executed again to acquire the hot-state reflection characteristics of each power fitting clamp; Step c, overheating defect judgment: The characteristic evolution variable from the cold-state reflection characteristics to the hot-state reflection characteristics is calculated. When the value of this characteristic evolution variable exceeds a preset overheating defect threshold, the corresponding power fitting clamp is determined to have an overheating defect. Specifically, this method treats the process of the line itself going from a power outage to bearing load as a predictable, endogenous "thermodynamic disturbance." The specific implementation includes the following steps: First, in response to the line's energizing command, while the line as a whole is still in a "cold state" at ambient temperature, the system quickly performs a complete reflection characteristic measurement to obtain the cold-state reflection characteristics of each power fitting clamp. This characteristic serves as the starting point for subsequent comparisons. Then, after the line has continuously carried the load for a sufficient period, its temperature rises due to the thermal effect of the current and reaches a thermodynamically stable state (i.e., a "hot state"), the system performs another complete measurement to obtain the hot-state reflection characteristics of each power fitting clamp. The core of the diagnosis lies in the comparative analysis of these two measurement results, that is, calculating the characteristic evolution variable from cold-state reflection characteristics to hot-state reflection characteristics. When the value of the characteristic evolution variable at a certain clamp location, such as the drift distance of the characteristic vector in multidimensional space, exceeds a preset overheating defect threshold, the system determines that the clamp has an overheating defect. The technical effect of this solution is that it upgrades the monitoring target from simply mechanical structural loosening to directly diagnosing electrical overheating problems caused by loosening, which pose a greater practical operational hazard. The principle behind this method is that a well-fitting clamp has extremely low contact resistance. When energized, its temperature rise is synchronous and gradual with the conductor itself. Therefore, the evolution of its reflection characteristics from a cold to a hot state is minute, predictable, and consistent with other healthy points along the entire line. However, a loose clamp, with its increased contact resistance, becomes a "heat spot" under load, generating a localized temperature rise far exceeding normal levels. This drastic localized temperature change causes a significant, non-linear alteration in the impedance characteristics of that point, resulting in a much larger variation in its reflection characteristics compared to other healthy points. By capturing this differentiated "thermal drift," this method can effectively distinguish defective points with overheating risks from numerous normal points, enabling remote, non-contact diagnosis of the heating state of line connection points. This allows for more direct and accurate early warning of major safety hazards that could lead to serious accidents such as line burnout.

[0020] Further, in the additional steps, steps a and b are applied simultaneously to the A, B, and C phase conductors of the transmission line to obtain the cold-state reflection characteristics and hot-state reflection characteristics of each power fitting clamp on the three phases. Furthermore, the overheating defect judgment process in step c further includes: c1, differential mode calculation: for each power fitting clamp in the same location, based on its characteristic evolution variables on the A, B, and C phases, the differential mode characteristic evolution variable of each phase is calculated; c2, differential mode diagnosis: when the value of the differential mode characteristic evolution variable of any phase exceeds a preset differential mode imbalance threshold, it is determined that the power fitting clamp of that phase has an overheating defect. Specifically, this method requires the detection system to simultaneously perform the aforementioned cold and hot state characteristic acquisition steps on the A, B, and C phase conductors of the transmission line, thereby obtaining the complete characteristic evolution process of each power fitting clamp in the same location on the three-phase line. The overheating defect judgment process introduces the concept of differential mode calculation. Specifically, for each clamp in the same location, the system first calculates the differential-mode characteristic evolution variable of each phase based on its respective characteristic evolution variables on phases A, B, and C. This differential mode can be understood as the deviation between the actual evolution variable of each phase and its average evolution variable across the three phases. Subsequently, in the differential-mode diagnostic step, when the system detects that the value of the differential-mode characteristic evolution variable of any phase exceeds a preset differential-mode imbalance threshold, it can determine that the corresponding power fitting clamp has an overheating defect. The technical advantage of this solution lies in its ability to significantly improve the signal-to-noise ratio of the diagnosis and the ability to suppress common-mode interference by introducing inter-phase lateral comparison. In actual operation, factors such as overall changes in ambient temperature, changes in solar radiation intensity, or fluctuations in the total line load will have approximately the same impact on phases A, B, and C, causing a "common-mode drift" in the reflection characteristics of all clamps. In single-phase analysis, this common-mode drift may mask the weak "differential-mode drift" caused by a single defect. This method, by calculating the differential-mode evolution variable, can effectively cancel out these common-mode effects acting on the three phases. In a healthy, three-phase load-balanced system, the characteristic evolution trajectories of the three phases should be highly consistent, with the differential mode approaching zero. However, a defect present in only one phase will break this symmetry, generating a significant, isolated differential mode signal. Therefore, this method can accurately extract weak single-phase fault signals from strong common-mode background noise, significantly improving the detection sensitivity and diagnostic certainty of early, minor overheating defects, and fundamentally reducing the possibility of false alarms caused by changes in environmental factors.

[0021] Further, in step 4, the pre-stored baseline reflection features are obtained through the following steps: S41, Peak detection: Apply an adaptive threshold algorithm to the reflection signal sequence obtained when all power fitting clamps are in a known tight state to identify all reflection peaks; S42, Precise positioning: Apply Gaussian fitting or parabolic fitting to each identified reflection peak to calculate the peak time with sub-sampling point accuracy, and calculate the precise position of each reflection peak accordingly; S43, Identity binding and feature extraction: Associate the positions of all calculated reflection peaks with the as-built drawing data of the line to clarify the corresponding physical identity of each reflection peak, and extract the feature parameters of the reflection peak to store it in the database as the baseline reflection features. Specifically, this method ensures the accuracy of all subsequent comparison work. Its specific steps include, firstly, under the premise that all power fitting clamps along the entire line are in a known tight state, applying an adaptive threshold peak detection algorithm to the obtained original reflection signal sequence to automatically and efficiently identify all potential reflection peaks. Secondly, to obtain precise location information, this method applies numerical fitting algorithms such as Gaussian fitting or parabolic fitting to each initially identified reflection peak. These algorithms can calculate the peak time with sub-sampling point precision by curve fitting the peak point and its neighboring sampling points, thereby solving for the precise physical location of each reflector (i.e., clamp). Its positioning accuracy exceeds the sampling interval limitations of the data acquisition system. Finally, and crucially, the precise locations of all calculated reflection peaks are associated with the as-built drawings of the line. This "identity binding" process clarifies the corresponding physical identity of each abstract reflection peak, such as "Tower No. XX - Phase A - Tension Clamp". After identity binding, the system extracts the complete feature parameters of the reflection peak and stores them in the database as the healthy baseline reflection characteristics of the physical clamp. The comprehensive technical effect of this solution is that it can establish a high-fidelity, high spatial resolution "health fingerprint" database with clear physical meaning. First, the sub-sampling-point precision positioning algorithm ensures that the position of each clamp is accurately determined, avoiding the possibility of confusing adjacent clamps due to positioning ambiguity. This is a prerequisite for achieving accurate single-point diagnosis. Second, the association with as-built drawings transforms pure electrical signal data into engineering information that maintenance personnel can directly understand and execute. This allows diagnostic results to be directly translated into specific maintenance instructions, such as "Please check clamp XX on tower XX." Finally, this entire systematic baseline establishment process ensures that all subsequent online monitoring comparisons have an extremely stable, reliable, and accurate "reference point," thereby guaranteeing the accuracy and reliability of the entire monitoring system's diagnostic results from the source. This provides a solid foundation for achieving highly reliable automated remote monitoring.

[0022] The technical effects of the present invention regarding the anti-loosening monitoring method and system for power fitting clamps include: First, the method centrally deploys monitoring equipment at a detection station at one end of the line, eliminating the need for any active or passive sensors, power supplies, or communication modules to be installed on the power fitting clamps along the line. This fundamentally solves a series of bottleneck problems associated with traditional online monitoring solutions, such as high costs, complex deployment, battery replacement, and data communication caused by a large number of end devices, significantly reducing the system's total lifecycle cost and maintenance complexity. Second, by utilizing the transmission line's own conductor as the sensing medium, a single detection can cover all fitting clamps along the entire line, achieving a leap from traditional "point-based" inspection or monitoring to "line-based" remote sensing, resulting in an order-of-magnitude improvement in monitoring efficiency and coverage. Third, by establishing baseline reflection characteristics of the health status and performing differential comparisons, common-mode interference caused by environmental changes and other factors can be effectively filtered out, and characteristic distortions caused by changes in the clamp's own state can be accurately identified, thereby significantly improving the reliability and accuracy of diagnosis. Finally, this method can accurately locate the loosening defects of the clamps, providing operation and maintenance personnel with clear fault point information, enabling them to carry out precise and efficient maintenance, avoiding blind troubleshooting, and ensuring the safe and stable operation of the power system. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a flowchart of the monitoring method according to Embodiment 1 of the present invention;

[0025] Figure 2 This is a schematic diagram of the monitoring system according to Embodiment 1 of the present invention;

[0026] Figure 3 This is a baseline reflection signal characteristic diagram of the fitting clamp in a healthy state, as described in Embodiment 1 of the present invention;

[0027] Figure 4 This is a baseline reflection signal characteristic diagram of the hardware clamp in a loose state in Embodiment 1 of the present invention;

[0028] Figure 5 This is a characteristic diagram of the reflection amplitude of the fastening clamp under dynamic excitation in Embodiment 3 of the present invention;

[0029] Figure 6 This is a characteristic diagram of the reflection amplitude of a loosened clamp under dynamic excitation in Embodiment 3 of the present invention.

[0030] Figure label:

[0031] none. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0033] Example 1

[0034] refer to Figure 1 and Figure 2 As shown, this embodiment provides a method for monitoring the loosening of power fitting clamps. The implementation of this method relies on a line condition detection system deployed at one end of the line, using a specific conductor of the transmission line as a passive sensing medium. In a preferred embodiment, the conductor is either the metal sheath of an optical fiber composite overhead ground wire (OPGW) or a conventional overhead ground wire; this choice avoids impacting the main transmission circuit.

[0035] The aforementioned line status detection system is not a single device, but rather a system organically composed of multiple functional modules. In a typical implementation, the system may include:

[0036] 1. Main control and data processing unit: This is usually an industrial PC or embedded system, responsible for running monitoring software, controlling the collaborative work of various hardware modules, executing signal processing algorithms (such as peak detection, feature extraction, Mahalanobis distance calculation, etc.), managing the baseline database, and generating alarms.

[0037] 2. Signal Generation and Injection Module: The core is a high-speed arbitrary waveform generator (AWG) or direct digital frequency synthesizer (DDS), used to generate Gaussian pulses in TDR mode or linear frequency modulated signals in FDR mode. The signal is amplified by a power amplifier (PA) to the appropriate injection power.

[0038] 3. Signal Reception and Acquisition Module: The core components are a high-bandwidth, low-noise amplifier (LNA) and a high-speed analog-to-digital converter (ADC). The LNA amplifies the weak signals reflected back from the line, while the ADC performs high-fidelity digital acquisition at a sampling rate much higher than the Nyquist frequency (e.g., 2.5 GSPS or higher for a 500MHz bandwidth signal).

[0039] 4. High-voltage isolation and coupling unit: This is a crucial module for safely connecting low-voltage detection electronic equipment to high-potential transmission line conductors, and is an important prerequisite for the realization of this invention. This unit is typically installed in a substation, using a non-contact coupler (such as a broadband directional coupler or a dedicated capacitive / inductive coupling probe) to inject and extract signals. This coupler must have a sufficient withstand voltage rating (e.g., 500kV) and be able to achieve effective isolation from DC to the upper limit of the detection frequency band (e.g., 500MHz), while ensuring good impedance matching (typically 50 ohms) to reduce reflections at the connection point.

[0040] 5. Clock and Synchronization Module: Provides a highly stable system clock and ensures that the signal generation, injection, and acquisition processes are synchronized with picosecond-level precision. This is fundamental for achieving coherent accumulation processing to improve the signal-to-noise ratio. Typically, the trigger signal for signal injection also serves as the start trigger signal for data acquisition.

[0041] Before routine monitoring, a baseline reflection feature database is established; this process, known as baseline calibration, forms the basis for all subsequent diagnostics. This calibration is performed after line construction is complete or after a comprehensive overhaul, ensuring all electrical fitting clamps are reliably tightened. The detection system injects a series of high-frequency probe signals into selected transmission line conductors and receives the returned complete sequence of reflection signals. The system first applies an adaptive threshold peak detection algorithm (e.g., setting the detection threshold to 5 to 8 times the noise floor standard deviation) to the raw reflection signal sequence to initially identify all reflection peaks generated by impedance discontinuities (primarily towers and clamps). To obtain high-precision location, the system applies Gaussian or parabolic curve fitting algorithms to each initially identified reflection peak and its neighboring sampling point data to calculate the peak time with sub-sampling point accuracy. Based on the propagation speed of electromagnetic waves and this precise round-trip time, the precise physical location of each reflector along the line can be calculated. Finally, the system associates these calculated location sequences with the as-built drawing data of the line (including tower number, clamp type, and attachment point information), thereby clarifying the physical identity of each reflection peak (e.g., "Tower No. XX - Phase A - Suspension Clamp"). After identity binding is completed, the system extracts the reflection signal features at each identified clamp location to form a multi-dimensional baseline feature vector, and stores it along with the physical identity in the database to complete baseline calibration.

[0042] In routine monitoring, this method includes the following steps:

[0043] Step 1: Signal Injection. The detection system injects a high-frequency probe signal from the detection end into a preset conductor of the transmission line, which serves as the sensing medium. In this embodiment, the probe signal can employ two preferred technical modes:

[0044] Time Domain Reflectometry (TDR) mode: Employs Gaussian or raised cosine pulses with pulse widths ranging from 2 ns to 10 ns. For example, a Gaussian pulse with a pulse width of 4 ns provides sufficient distance resolution to distinguish different fittings on the same tower.

[0045] Frequency Domain Reflection (FDR) mode: Employs a linear frequency modulated signal with a scan bandwidth of 50MHz to 500MHz. For example, a chirp signal scanned from 50MHz to 250MHz, with a bandwidth of 200MHz, can also provide meter-level distance resolution.

[0046] Step 2: Signal Reception. At the detection end, the detection system uses a high-sensitivity receiver to capture the weak reflected signals generated at various electrical fitting clamp positions due to impedance discontinuities as the probe signal propagates along the line. These signals form a sequence of reflected signals in chronological order.

[0047] Step 3: Signal Processing and Feature Extraction. To effectively extract weak reflected signals from background noise, a key processing step is to perform point-by-point alignment and coherent accumulation of the N consecutively acquired reflected signal sequences (e.g., N=1024). This significantly improves the signal-to-noise ratio (theoretically by about 15dB). Subsequently, the accumulated signal undergoes digital bandpass filtering to further remove out-of-band noise. After processing, the system determines the position of the corresponding power fitting clamp based on the round-trip time of each reflected signal and extracts real-time reflection features characterizing the current connection state of that clamp.

[0048] Step 4: Feature Comparison and Diagnosis. Before feature comparison, a crucial time axis calibration step is introduced to eliminate the overall time axis drift caused by line length expansion and contraction due to temperature changes. The system selects several stable and strong reflection peaks generated by the tower body from the baseline database as "anchor points." Preferably, reflection peaks generated by the metal tower body (especially the tower head) in the line are selected as anchor points. Because the tower structure is large and stable, its reflection characteristics are almost unaffected by local minor defects such as loose clamps, making it an ideal reference point. By calculating the cross-correlation function between the real-time reflection signal sequence and the waveforms of these anchor points, the position with the largest correlation peak is found, thereby accurately calculating the overall translation of the current time axis relative to the baseline. This translation is used to compensate and correct the time axis of the entire real-time reflection signal sequence, ensuring the positional accuracy of subsequent comparisons.

[0049] After calibration, the system compares the real-time reflection characteristics of each power fitting clamp with the baseline reflection characteristics at the same location, which are pre-stored in the database. This feature extraction and comparison process is multi-dimensional. Specifically, for each located clamp reflection waveform, the system extracts multiple parameters, including peak amplitude, pulse energy (obtained by integrating the pulse waveform), pulse full width at half maximum (FWHM), rise / fall time, symmetry coefficient, and ringing factor (the ratio of subsequent small peaks to the main peak energy), which together constitute a real-time feature vector V. real During the comparison, the system retrieves the baseline feature vector V at the corresponding location from the database. base And the covariance matrix S. The Mahalanobis distance between the real-time eigenvectors and the baseline eigenvectors is calculated:

[0050]

[0051] To quantify the difference between the two, where V real It is a real-time feature vector, V base S is the baseline eigenvector, and S is the eigenvector covariance matrix obtained statistically using the clamps in a tightened state. -1 This is its inverse matrix, where T represents the matrix transpose operation. Transpose allows for dimension matching, enabling matrix multiplication and other operations. The square root of the transpose yields the Mahalanobis distance, used to measure the difference between two vectors considering feature correlation and normal fluctuation range. This covariance matrix S is extracted during the baseline calibration phase by performing multiple measurements (e.g., 100 sets of data collected at different ambient temperatures) on each line clamp in a healthy state, extracting its multidimensional feature vectors. From these 100 sets of feature vector samples, the covariance matrix S, representing the normal fluctuation range of the features in the healthy state of the line clamp, can be statistically calculated. It serves as a benchmark for assessing the degree of deviation in subsequent monitoring. The introduction of Mahalanobis distance considers the normal fluctuation range of each feature parameter and their inter-parameter correlation, making the measurement of difference more scientific and reliable. (Refer to...) Figure 3 and Figure 4 The comparison shows that the differences were quantified by calculating the Mahalanobis distance of the fitting clamps in two different states. Figure 3 and Figure 4 The comparison vividly illustrates the difference in the characteristics of the reflected signal from the clamp under normal and loose conditions.

[0052] Step 5: Alarm Generation. When the calculated Mahalanobis distance D exceeds the preset looseness judgment threshold (for example, this threshold can be set to 3 or 5 based on historical data and operation and maintenance experience), the system determines that the corresponding power fitting clamp is loose and generates an alarm message, indicating the specific location of the loose clamp (such as the suspension clamp of phase XX on tower XX of line XX).

[0053] Example 2

[0054] To further enhance signal identifiability and detection sensitivity, this invention also provides an enhanced technical solution. In specific implementation, during line construction or maintenance, a passive impedance characteristic code element is pre-fixed at a non-critical load-bearing position of each power fitting clamp.

[0055] The specific implementation of this component may include:

[0056] 1. Physical form: It can be manufactured using flexible printed circuit board (FPC) technology, allowing it to conform to the curvature of the wire clip surface. The substrate can be made of materials with excellent weather resistance and high and low temperature resistance, such as polyimide (PI).

[0057] 2. Fixing method: Use epoxy resin structural adhesive that is resistant to high and low temperatures, UV rays, and has high shear strength to firmly bond it to the non-load-bearing, clean surface of the wire clamp. The surface must be sanded and cleaned before installation to ensure good mechanical and electrical contact.

[0058] 3. Design Considerations: The resonant frequency should be designed to avoid radio interference bands commonly found in power line corridors to improve the signal-to-noise ratio. Its structure (such as a patch-type spiral resonator or an array of multiple quarter-wavelength stubs of varying lengths) needs to be optimized using electromagnetic simulation software (such as HFSS or CST) to achieve the highest reflection intensity (i.e., lowest insertion loss) and the most suitable Q value at the target frequency.

[0059] Under this scheme, when the detection system monitors, its analysis target is no longer the weak and irregularly shaped natural reflection of the wire clamp itself, but the high-intensity and uniquely shaped resonant reflection signal generated by this feature element. When the wire clamp becomes loose, the contact resistance between its body and the feature element changes, which severely damages the quality factor Q of the resonant structure. This change in Q directly leads to significant distortion of the characteristics of the resonant reflection signal (for example, a sharp decrease in the amplitude and broadening of the reflected pulse in TDR mode; and a shallowing and broadening of the resonant peak in FDR mode). Since the amplitude of this characteristic distortion caused by the deterioration of Q is much greater than the change in natural reflection, the system can determine the loose state of the wire clamp with extremely high sensitivity, effectively reducing the probability of missed and false alarms.

[0060] Example 3

[0061] This embodiment further expands the dimensions of monitoring methods by introducing a diagnostic scheme based on dynamic excitation and differential comparison to discover early or potential defects that are difficult to detect under static conditions.

[0062] One approach utilizes environmental disturbances for dynamic analysis. The detection system links with anemometers along the line or a distributed vibration sensing system (DVS) integrated into the OPGW. When wind speed exceeds a preset threshold (e.g., 15 m / s) or abnormal line vibration occurs, the system is triggered, continuously executing steps 1 to 3 at a high-density frequency (e.g., 10 Hz) to obtain a set of real-time reflection characteristic sequences that change rapidly over time at each clamp location. The diagnostic logic then shifts to dynamic analysis: the system calculates the standard deviation or peak-to-peak value of a key parameter (e.g., reflection amplitude) within this characteristic sequence over time. A secure clamp should maintain stable reflection characteristics under this disturbance, with the calculated result close to zero; however, a mechanically loose clamp will exhibit "jittering" reflection characteristics with vibration, leading to a significant increase in the calculated result. When this result exceeds a set dynamic jitter threshold, the clamp can be determined to be loose. Combined with... Figure 5 and Figure 6 As the comparison shows, the signal reflection characteristics 'jitter' with vibration, leading to a significant increase in the calculation results. Figure 5 and Figure 6 The characteristic stability of tightened and loose wire clamps under dynamic disturbances was compared intuitively.

[0063] Another more advanced approach is to diagnose the problem by utilizing changes in the line's own operating state, particularly suitable for diagnosing overheating defects caused by loosening. This additional step is closely coordinated with the line's power outage and restoration operations. First, in response to the line's power-on command, a measurement is performed while the line is in a cold state at ambient temperature to obtain the cold-state reflection characteristics of all clamps. Then, after the line has operated under heavy load for a period of time and reached a thermodynamically stable state, another measurement is performed to obtain the hot-state reflection characteristics of all clamps. For a healthy clamp, the evolution of its reflection characteristics from cold to hot is small and predictable. However, for a clamp that has become loose due to poor contact, it becomes a hot spot after being energized, with a local temperature much higher than the conductor itself, causing its reflection characteristics to undergo a drastic and abnormal evolution from cold to hot. By calculating the magnitude of this characteristic evolution and comparing it with an overheating defect threshold set based on statistical data of healthy clamps, defect points with overheating risk can be accurately identified.

[0064] To achieve the highest possible signal-to-noise ratio, this thermal diagnostic method can be further upgraded to a three-phase differential comparison mode. The detection system must be capable of simultaneously detecting the three-phase conductors (A, B, and C). During the aforementioned cold and hot state measurements, the system simultaneously acquires the cold and hot state reflection characteristics of each corresponding clamp on the three phases and calculates their respective characteristic evolution variables. In a system with a basically balanced three-phase load, common-mode factors such as ambient temperature and solar radiation have a consistent impact on the three phases. Therefore, by calculating the difference between the characteristic evolution variable of each phase and its three-phase average, a differential-mode characteristic evolution variable can be obtained. This differential modulus can effectively cancel out all common-mode interference. Only when a clamp in a certain phase has a single-point overheating defect will the value of its differential-mode characteristic evolution variable significantly deviate from zero. When this differential modulus value exceeds a preset differential-mode imbalance threshold, the system can determine with extreme sensitivity and unambiguity that the clamp in that specific phase has an overheating defect. This lateral comparison differential diagnostic logic elevates the reliability and sensitivity of the detection to a higher level.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring the loosening of electrical fitting clamps, characterized in that, Includes the following steps: Step 1: Inject a high-frequency probe signal from a probe end into a preset conductor of the transmission line, which serves as the sensing medium. Step 2: At the detection end, receive the sequence of reflected signals generated by the impedance discontinuity at multiple electrical fitting clamp positions along the line when the probe signal propagates along the preset conductor. Step 3: Process the reflected signal sequence to determine the position of the corresponding power fitting clamp based on the round-trip time of each reflected signal, and extract the real-time reflection features characterizing the connection status of the power fitting clamp. Step 4: Compare the real-time reflection characteristics of each power fitting clamp with the baseline reflection characteristics of a power fitting clamp in the same location under a tightened state, which are stored in advance. Step 5: When the difference generated by the comparison exceeds the preset looseness judgment threshold, it is determined that the corresponding power fitting clamp is loose.

2. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, The high-frequency probe signal is a Gaussian pulse or raised cosine pulse with a pulse width of 2ns to 10ns for time-domain reflectometry; or, the high-frequency probe signal is a linear frequency modulated signal with a scanning bandwidth of 50MHz to 500MHz for frequency-domain reflectometry.

3. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, In step 3, processing the reflection signal sequence includes: before extracting the real-time reflection features, performing point-by-point alignment and coherent accumulation on multiple consecutive reflection signal sequences synchronously acquired from the detection end, and performing digital bandpass filtering on the accumulated signal.

4. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, Before the comparison in step 4, a time axis calibration step is also included. This step includes: selecting a stable reflection peak with amplitude and waveform generated by the tower body from the baseline reflection characteristics as an anchor point; determining the peak position by calculating the cross-correlation function between the real-time reflection signal sequence and the anchor point to solve the overall translation of the time axis; and using the translation to compensate and correct the time axis of the entire real-time reflection signal sequence.

5. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, In step 3, the process of extracting the real-time reflection features includes: for each located power fitting clamp, extracting one or more feature parameters of its reflected signal waveform to form a real-time feature vector; the feature parameters are selected from: peak amplitude, pulse energy obtained by integrating the pulse waveform, pulse full width at half maximum (FWHM), pulse rise time, pulse fall time, pulse symmetry coefficient, and ringing factor defined by the energy ratio of subsequent small peaks to the main peak; In step 4, the comparison step specifically involves: retrieving the baseline feature vector corresponding to the position of the power fitting clamp from a pre-stored database; and using the inverse matrix of the covariance matrix of the feature vector obtained statistically when the power fitting clamp is in a tightened state, calculating the Mahalanobis distance between the real-time feature vector and the baseline feature vector as the difference.

6. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, A passive impedance characteristic element with a preset high-frequency resonance characteristic is fixed on each power fitting clamp in advance. The characteristic element is a spiral resonant ring or a stub array. The baseline reflection features and real-time reflection features extracted in step 3 are characteristics of the resonant reflection signal with a specific shape generated by the feature code element. In step 5, the loosening is determined based on the characteristic distortion caused by the change in the quality factor Q value of the resonant reflected signal.

7. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, The execution of step 1 is triggered by environmental disturbance data. The steps for determining the trigger include: Linkage steps: Obtain real-time disturbance data characterizing the environmental state of the line from the external environmental monitoring system, including wind speed or line vibration data; Triggering step: When the disturbance data exceeds a preset disturbance threshold, the detection end is triggered to continuously execute steps 1 to 3 multiple times at a preset high density frequency to obtain a set of time-varying real-time reflection characteristics for each power fitting clamp. Dynamic analysis steps: In step 5, the standard deviation or peak-to-peak value of a certain characteristic parameter in the set of real-time reflection characteristics is calculated as a function of time. When the calculation result exceeds a dynamic jitter threshold, it is determined that the corresponding power fitting clamp is loose.

8. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, It also includes an additional step for diagnosing overheating defects, the additional step comprising: Step a, cold state feature acquisition: In response to the line closing and power supply command, when the line is in a cold state at ambient temperature, steps 1 to 3 are executed to acquire the cold state reflection features of each power fitting clamp. Step b, thermal characteristic acquisition: After the line is loaded and reaches a thermodynamically stable state, steps 1 to 3 are executed again to acquire the thermal reflection characteristics of each power fitting clamp. Step c, Overheating Defect Judgment: Calculate the feature evolution variable from the cold reflection feature to the hot reflection feature. When the value of the feature evolution variable exceeds a preset overheating defect threshold, it is determined that the corresponding power fitting clamp has an overheating defect.

9. The method for monitoring the loosening of power fitting clamps according to claim 8, characterized in that, In the additional steps, steps a and b are applied simultaneously to the three-phase conductors A, B, and C of the transmission line to obtain the cold-state reflection characteristics and hot-state reflection characteristics of each power fitting clamp on the three phases. Furthermore, the overheating defect judgment process in step c further includes: c1. Differential Mode Calculation: For each power fitting clamp in the same location, based on its characteristic evolution variables on phases A, B, and C, calculate the differential mode characteristic evolution variables for each phase. c2. Differential mode diagnosis: When the value of the differential mode characteristic evolution of any phase exceeds a preset differential mode imbalance threshold, it is determined that there is an overheating defect in the power fitting clamp of that phase.

10. The method for monitoring the loosening of power fitting clamps according to claim 1, characterized in that, In step 4, the pre-stored baseline reflection features are obtained through the following steps: S41. Peak detection: For the reflected signal sequence obtained when all power fitting clamps are in a known tight state, an adaptive threshold algorithm is applied to identify all reflection peaks; S42. Precise positioning: Apply Gaussian or parabolic fitting to each identified reflection peak to calculate the peak time with subsampling point accuracy, and solve for the precise position of each reflection peak accordingly. S43. Identity Binding and Feature Extraction: The positions of all calculated reflection peaks are associated with the as-built drawing data of the line to clarify the physical identity of each reflection peak, and the feature parameters of the reflection peak are extracted and stored in the database as the baseline reflection features.

Images