A method and system for ice detection on overhead power lines

Abstract

This invention relates to the field of icing and snow accumulation detection technology, and discloses a method and system for detecting icing on overhead transmission lines. The method includes: deploying integrated torsional vibration and acoustic emission sensing units at both ends of the transmission line span to collect conductor angular velocity and acoustic emission signals; performing cross-correlation processing on the angular velocity signals within a time window to obtain the propagation delay of multi-band torsional guided waves, and calculating the group velocity in conjunction with the span length; comparing the group velocity with a baseline curve for an ice-free period, and inverting the icing thickness using a composite pole dispersion model; using the acoustic emission signals to detect sudden events; if the number and energy of events exceed a threshold and the group velocity rebounds, it is determined to be an icing shedding state; otherwise, it enters the thickness and event fusion stage, outputting the icing thickness, growth rate, icing shedding alarm, and confidence level through a confidence factor. This invention can realize dynamic monitoring and early warning of the entire process of icing on transmission lines.

Description

Technical Field

[0001] This invention relates to the field of icing and snow accumulation detection technology, and more specifically, to a method and system for detecting icing on overhead transmission lines. Background Technology

[0002] Overhead transmission lines are prone to icing in winter or under extreme weather conditions. The adhesion of ice increases the weight of the conductors and the lateral wind load, causing conductor galloping, abnormal tension, or even line breakage, which seriously threatens the safety of the power grid. To address this, existing technologies have proposed a variety of icing detection methods.

[0003] For example, one approach utilizes fiber optic sensing or fiber optic cable image recognition to determine ice thickness through light signal attenuation, B-OTDR reflection features, or image edge extraction. However, this approach is sensitive to ambient light interference and lacks reliability under complex conditions such as nighttime or snowfall. Another approach uses sensors to measure wind speed, temperature, humidity, and tension to establish a correlation model between meteorological and mechanical factors to estimate ice thickness. However, this method is susceptible to errors in weather forecasts and fluctuations in conductor tension baselines, leading to potential biases. Some methods attempt to use radar wave or electromagnetic wave scattering characteristics to identify the ice and snow cover on the conductor surface. However, radar devices are expensive and suffer from signal distortion under multipath reflection or conductor swaying conditions. In summary, existing technologies still have the following shortcomings: detection methods relying on fiber optics or images have poor robustness under complex weather conditions and are prone to misjudgment; analysis methods based on meteorology and tension are sensitive to the environment and lack sufficient computational accuracy; radar wave-based solutions involve complex equipment, are susceptible to interference, and are difficult to apply on a large scale; and the aforementioned methods primarily focus on static thickness detection, lacking dynamic response and confidence assessment of the ice growth and shedding process. Summary of the Invention

[0004] In view of this, the present invention proposes a method and system for detecting icing on overhead transmission lines to solve the above problems.

[0005] On the one hand, the present invention proposes a method for detecting icing on overhead transmission lines, comprising: An integrated torsional vibration and acoustic emission sensing unit is installed at both ends of the transmission line span to collect the angular velocity signal and acoustic emission signal of the conductor in the transmission line. Based on the cross-correlation processing of the angular velocity signal of the conductor within a preset time window, the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor is obtained. Based on the transmission line span length and propagation delay, the group velocity in multiple frequency bands is obtained. The group velocity in each frequency band is compared with the baseline group velocity curve established during the ice-free period of the conductor. The ice thickness of the conductor is obtained by inversion using the composite pole dispersion model of the conductor and the ice layer. The sudden events that occur on the conductor are obtained based on the acoustic emission signals of the conductor. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, proceed to the next step. Based on the ice thickness of the conductor and sudden events, combined with the confidence factor, the output shows the current ice thickness, growth rate, de-icing alarm information, and confidence level.

[0006] Furthermore, the integrated torsional vibration and acoustic emission sensing unit includes an angular velocity sensor, a broadband acoustic emission sensor, a short-time ring actuator, and a clock module.

[0007] Furthermore, the specific content of the cross-correlation processing is as follows: within a preset time window, a Fourier transform is performed on the angular velocity signal of the conductor to decompose the angular velocity signal of the conductor into multiple frequency bands. In each frequency band, a generalized cross-correlation-phase transform is performed to obtain the normalized correlation function of the angular velocity signal of the conductor at both ends of the transmission line span. Based on the search for the cross-correlation peak position, the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor is obtained.

[0008] Furthermore, the specific content of the composite rod dispersion model of the conductor and the ice layer is as follows: based on the cross-sectional radius of the conductor, the pitch of the conductor, the number of strands of the conductor, and the material of the conductor as parameters, the equivalent polar inertia, the equivalent density of the conductor, and the equivalent shear modulus of the conductor are calculated. The polar inertia of the ice-covered annular cross section was calculated by using the ice layer thickness, ice material density, and ice shear modulus as parameters. Based on the conservation of mass and the torsional stiffness distribution relationship, the equivalent polar inertia of the conductor, the equivalent density of the conductor, the equivalent shear modulus of the conductor, the density of the icing material, the shear modulus of the icing material, and the polar inertia of the icing annular section are calculated to obtain the total polar inertia, the total equivalent density, and the total equivalent shear modulus.

[0009] Furthermore, the specific steps for obtaining the icing thickness of the conductor are as follows: the transmission line span length is divided by the propagation delay in each frequency band to obtain the group velocity in each frequency band; the group velocity in each frequency band is compared with the baseline group velocity curve established during the ice-free period of the conductor to obtain the relative offset of the group velocity in each frequency band; the relative offset of the group velocity in each frequency band is input into the composite rod dispersion model of the conductor and the icing layer for fitting to obtain multi-frequency band fitting residuals; when the multi-frequency band fitting residuals do not exceed a preset residual threshold, the corresponding ice layer thickness is output as the icing thickness of the conductor; when the multi-frequency band fitting residuals exceed the preset residual threshold, the frequency band fitting and redundancy point correction mechanism is triggered to remove and correct points in abnormal frequency bands to obtain the average icing thickness within a preset time window; the average icing thickness is output as the icing thickness of the conductor.

[0010] Furthermore, the specific content of obtaining the sudden event of the conductor based on the acoustic emission signal of the conductor is as follows: the broadband acoustic emission sensor continuously collects the high-frequency acoustic emission signal on the surface of the conductor and preprocesses the original signal. The preprocessing includes noise filtering, amplitude normalization and time window segmentation. Within a preset time window, a threshold triggering method is used to detect whether a sudden pulse with an amplitude exceeding the background noise appears in the acoustic emission signal. If a sudden pulse appears, it is recorded as a sudden event.

[0011] Furthermore, the specific content of determining the state of icing detachment as follows: If the number and energy of sudden events exceed a preset threshold and the group velocity rebounds, the starting point of the sudden event is accurately located using an adaptive information criterion method, and the characteristic parameters of the sudden event are calculated. The characteristic parameters include peak amplitude, duration, energy, and spectral centroid. The number and energy of the sudden events are compared with a preset threshold. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds simultaneously, the conductor is determined to be in an icing detachment state.

[0012] Furthermore, the specific content of outputting the current icing thickness, growth rate, de-icing alarm information, and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor, is as follows: The icing thickness of the conductor obtained by inversion within a preset time window is used as the ice layer thickness data point to form a thickness sequence that changes over time. The thickness sequence is differentially calculated to obtain the rate of change of the icing thickness of the conductor within adjacent time windows, which is used as the icing growth rate. At the same time, the conductor status is determined in conjunction with sudden events that occur on the conductor. If it is determined to be a state of icing detachment, a de-icing alarm information is generated. In the information output process, a confidence factor is introduced, and the detection results are divided into three levels: high confidence, medium confidence, and low confidence, based on the magnitude of the confidence factor.

[0013] Furthermore, the confidence factor is specifically derived as follows: a signal-to-noise ratio score is obtained based on the cross-correlation processing result of the angular velocity signal of the conductor; a consistency score is obtained based on the multi-band fitting residual; a correlation score is obtained based on the sudden events that occur on the conductor; and the confidence factor is obtained by weighted fusion of the signal-to-noise ratio score, the consistency score, and the correlation score.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention presents a method for detecting icing on overhead transmission lines. It directly reflects the increase in polar inertia and density caused by icing by the change in group velocity of the conductor's own torsional guided wave. Furthermore, it captures the ice cracking and detachment process using acoustic emission events, forming a novel detection path combining structural dynamics and acoustic coupling. This method does not rely on external environmental data and does not require high-cost fiber optic or radar equipment, thus fundamentally improving the stability and universality of the method. The method extracts group velocities across multiple frequency bands and performs fitting and inversion using a composite rod dispersion model. Multi-band fitting residual control ensures cross-frequency consistency of the results. Abnormal frequency band elimination and redundant point correction mechanisms prevent distortion at individual frequency points from affecting the overall results. Comparison with benchmark group velocity curves eliminates global material and environmental offsets, thus resulting in higher robustness and accuracy in thickness inversion, achieving centimeter-level thickness measurement. This invention provides an icing detection method for overhead transmission lines. It obtains the growth rate through thickness sequence difference to monitor the icing development trend; it distinguishes the energy and quantity of acoustic emission events, combined with the rebound characteristics of group velocity, to accurately identify de-icing events; and it uses multi-event localization to differentiate between uneven ice shedding and overall ice shedding, thus achieving full-process tracking and early warning of the three stages of icing growth, stabilization, and shedding. The method also constructs a confidence factor, which is fused from three independent scores: the signal-to-noise ratio score of the cross-correlation signal, reflecting signal quality; the consistency score of the multi-band fitting residual, reflecting the stability of thickness inversion; and the correlation score between acoustic emission events and thickness changes, reflecting the reliability of dynamic discrimination. The final output detection result carries a confidence level (high, medium, low), providing a clear and reliable reference for power grid dispatching.

[0015] On the other hand, the present invention proposes an icing detection system for overhead transmission lines, comprising: The integrated torsional vibration and acoustic emission sensing unit is configured to collect the angular velocity signal and acoustic emission signal of the conductor in the power transmission line. The processing unit is configured to perform cross-correlation processing on the angular velocity signal of the conductor within a preset time window to obtain the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor. The calculation unit is configured to obtain the group velocity in multiple frequency bands based on the transmission line span length and propagation delay, compare the group velocity in each frequency band with the baseline group velocity curve established during the ice-free period of the conductor, and invert the ice thickness of the conductor by combining the composite pole dispersion model of the conductor and the ice layer. The status assessment unit is configured to detect sudden events occurring on the conductor based on the acoustic emission signal of the conductor. If the number and energy of the sudden events exceed a preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, it proceeds to the next step. The output unit is configured to output the current icing thickness, growth rate, de-icing alarm information, and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor.

[0016] It should be noted that the icing detection system and method for overhead transmission lines of the present invention have the same beneficial effects, and will not be described in detail here. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a flowchart of a method for detecting icing on overhead transmission lines, provided as an embodiment of the present invention.

[0018] Figure 2 This is a functional block diagram of an icing detection system for overhead transmission lines provided in an embodiment of the present invention. Detailed Implementation

[0019] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] See Figure 1 As shown, this embodiment of the invention provides a method for detecting icing on overhead transmission lines, comprising: An integrated torsional vibration and acoustic emission sensing unit is installed at both ends of the transmission line span to collect the angular velocity signal and acoustic emission signal of the conductor in the transmission line. Based on the cross-correlation processing of the angular velocity signal of the conductor within a preset time window, the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor is obtained. Based on the transmission line span length and propagation delay, the group velocity in multiple frequency bands is obtained. The group velocity in each frequency band is compared with the baseline group velocity curve established during the ice-free period of the conductor. The ice thickness of the conductor is obtained by inversion using the composite pole dispersion model of the conductor and the ice layer. The sudden events that occur on the conductor are obtained based on the acoustic emission signals of the conductor. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, proceed to the next step. Based on the ice thickness of the conductor and sudden events, combined with the confidence factor, the output shows the current ice thickness, growth rate, de-icing alarm information, and confidence level.

[0021] In some embodiments of this application, the integrated torsional vibration and acoustic emission sensing unit includes an angular velocity sensor, a broadband acoustic emission sensor, a short-time ring actuator, and a clock module.

[0022] It should be noted that the angular velocity sensor is used to monitor the torsional vibration response of conductors in transmission lines under wind-induced disturbances or external excitations in real time. It can capture angular velocity changes at millisecond-level time resolution, providing basic signals for subsequent cross-correlation analysis and group velocity calculation. The broadband acoustic emission sensor is set on the surface of the conductor, with an operating frequency band covering 50kHz to 400kHz. It is used to collect high-frequency transient signals generated during the growth, crack propagation and shedding of the icing layer. After processing, it can effectively identify sudden events. The short-time ring vibrator is arranged inside the housing of the sensing unit. When the natural environmental disturbance is insufficient, it can apply low-amplitude, short-time pulse form of circumferential torsional vibration excitation to the conductor, thereby ensuring that the torsional guided wave still has a sufficient signal-to-noise ratio under severe weather conditions. The clock module adopts high-precision synchronization technology to provide a unified time reference for signal acquisition at both ends of the span, ensuring the synchronization of angular velocity signals and acoustic emission signals in data fusion and cross-correlation calculation. Through the synergy of the above components, the integrated torsional vibration and acoustic emission sensing unit of the present invention can collect both the dynamic torsional information of the conductor and the transient acoustic information of the icing layer. It also has self-excitation capability and time synchronization capability, forming an integrated monitoring platform that provides a stable and accurate data source for icing thickness inversion and de-icing status identification.

[0023] In some embodiments of this application, the cross-correlation processing specifically involves: performing a Fourier transform on the angular velocity signal of the conductor within a preset time window, decomposing the angular velocity signal of the conductor into multiple frequency bands, performing a generalized cross-correlation-phase transform in each frequency band to obtain the normalized correlation function of the angular velocity signals of the conductors at both ends of the transmission line span, and obtaining the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor based on the searched cross-correlation peak position.

[0024] Specifically, the angular velocity signals of the conductors collected by the integrated sensing unit for torsional vibration and acoustic emission at both ends of the transmission line span are sliced ​​using a preset time window, and expressed as follows: ; in, and The first p Angular velocity signals of conductors at both ends of the transmission line span within a time window. The length of the preset time window, The sampling frequency; The angular velocity signal of the conductor in each time window is subjected to trend and bandpass filtering and amplitude normalization, followed by discrete Fourier transform, expressed as:

[0025] ; in, and These are the angular velocity signals of the conductor after Fourier transform. For windowing functions, For a complex exponential term, ; The frequency band is divided into K non-overlapping frequency bands, and the sub-band cross spectrum and self-spectrum are calculated and expressed as: ;

[0026] ; in, For frequency band mask, For the first k Each frequency band, For frequency bandwidth, For the first m The corresponding frequency of each frequency point For the first p The first time window, the first k Cross spectrum within each frequency band and All are self-composed; Generalized cross-correlation-phase transform normalization of the subband cross spectrum, followed by inverse transform, yields the discrete time-delay correlation function, which can suppress amplitude fluctuations and highlight time delay information. This function is expressed as: ; in, For discrete time-delay correlation function, >0 is a constant. This is a discrete time delay index, corresponding to physical time delay. , The maximum number of search lag samples, It is the inverse transform function; The amplitudes of each function are normalized to facilitate comparison across different frequency bands. Then, peak search is used to obtain the initial propagation delay within the frequency band, expressed as: ; in, For the p-th time window, the initial value of the peak lag index for the k-th frequency band is... This represents the initial value of the propagation delay for the p-th time window and the k-th frequency band. The propagation delay is obtained by parabolic interpolation of the peak point and its neighbors, and is expressed as: ; ; in, This is the parabolic interpolation correction amount. Let p be the propagation delay of the k-th frequency band in the p-th time window.

[0027] It should be noted that by performing Fourier transform and frequency band division on the angular velocity signals at both ends of the span, the torsional vibration under broadband random excitation can be decomposed into multiple stable sub-band components. In each sub-band, a generalized cross-correlation-phase transform is used to eliminate the interference of amplitude differences and retain only the phase information, thereby making the propagation delay characteristics of the torsional guided wave more prominent and facilitating accurate detection. Furthermore, since the angular velocity signal under natural wind excitation often contains a large amount of random noise and uncorrelated components, direct analysis would mask the propagation characteristics. Therefore, cross-correlation calculation of the signals at both ends effectively filters out uncorrelated noise components, retaining only the part related to the torsional guided wave and improving the signal-to-noise ratio.

[0028] In some embodiments of this application, the composite rod dispersion model of the conductor and the icing layer specifically includes: calculating the equivalent polar inertia, equivalent density, and equivalent shear modulus of the conductor based on the conductor's cross-sectional radius, pitch, number of strands, and material as parameters, expressed as: ; in, For equivalent cross-sectional area, , The radius of the conductor; ; in, For equivalent extreme inertia, This is used to reflect the reduction of the polar inertia of the strand structure; ; in, For equivalent density, This is the equivalent shear rate. For the density of the material, The shear modulus of the material. , The equivalent reduction factor of the shear modulus to the twist angle; Using the ice layer thickness, ice material density, and ice shear modulus as parameters, the polar inertia of the ice-covered annular cross-section is calculated and expressed as: ; in, The inner radius of the ice layer. The outer radius of the ice layer. This refers to the thickness of the ice layer. ; in, The cross-sectional area of ​​the ice-covered ring is... The polar inertia is the cross-sectional area of ​​the ice-covered annulus. Based on the principles of mass conservation and torsional stiffness distribution, the equivalent polar inertia of the conductor, the equivalent density of the conductor, the equivalent shear modulus of the conductor, the density of the icing material, the shear modulus of the icing material, and the polar inertia of the icing annular section are calculated to obtain the total polar inertia, total equivalent density, and total equivalent shear modulus, which are expressed as: ; in, The total polar inertia; ; in, For the total equivalent density, This refers to the density of the ice layer. ; in, The total equivalent shear modulus, This represents the ice-covered shear modulus.

[0029] Furthermore, under the first-order approximation of a uniformly isotropic rod with a circular cross-section, the phase / group velocity of the torsional guided wave can be expressed as: ; Considering the dispersion effect of the stranded conductor structure and finite diameter at high frequencies, an empirical correction factor can be introduced to form a theoretical group velocity curve for fitting, thus obtaining the composite rod dispersion model, expressed as: ; in, This is an empirical correction factor, which can be obtained through on-site calibration.

[0030] It should be noted that by introducing the equivalent polar inertia, equivalent density, and equivalent shear modulus of the conductor, and combining them with the polar inertia, ice density, and ice shear modulus of the icing annular cross section, the changes in mass distribution and torsional stiffness of the conductor cross section after ice adhesion can be comprehensively reflected. Compared with existing methods that are only based on tension or external meteorological data, this invention, starting from the principle of structural dynamics, can directly quantify the influence of ice on the propagation speed of torsional waves, and has higher physical rationality. Through the principles of mass conservation and torsional stiffness distribution, the composite parameters of the conductor and ice are clearly calculated as the total polar inertia, total equivalent density, and total equivalent shear modulus. The dispersion model formed on this basis gives the functional relationship between ice thickness and group velocity frequency band characteristics, so that the thickness inversion depends not only on the single-point velocity value, but also on the fitting of the entire frequency band, thus improving the sensitivity of identification.

[0031] In some embodiments of this application, the specific steps for obtaining the icing thickness of the conductor are as follows: the transmission line span length is divided by the propagation delay in each frequency band to obtain the group velocity in each frequency band; the group velocity in each frequency band is compared with the baseline group velocity curve established during the ice-free period of the conductor band by band to obtain the relative offset of the group velocity in each frequency band; the relative offset of the group velocity in each frequency band is input into the composite rod dispersion model of the conductor and the icing layer for fitting to obtain a multi-frequency band fitting residual; when the multi-frequency band fitting residual does not exceed a preset residual threshold, the corresponding ice layer thickness is output as the icing thickness of the conductor; when the multi-frequency band fitting residual exceeds the preset residual threshold, a frequency band fitting and redundancy point correction mechanism is triggered to remove and correct points in abnormal frequency bands to obtain the average icing thickness within a preset time window; and the average icing thickness is output as the icing thickness of the conductor.

[0032] Specifically, the group velocity in each frequency band is obtained by dividing the transmission line span length by the propagation delay in each frequency band, expressed as: ; in, For the p-th time window, the group velocity in the k-th frequency band is... This refers to the span length of the transmission line; The baseline group velocity curve established during the ice-free period of the conductor is as follows: The center frequency of the kth frequency band The value at that location is Then the relative offset of the group velocity in the k-th frequency band is expressed as: ; in, Let be the relative offset of the group velocity within the k-th frequency band. The relative offset of the theoretical group velocity, calculated from the theoretical group velocity curve, is expressed as: ; in, This represents the relative offset of the theoretical group velocity. This is the theoretical group velocity curve at the k-th frequency band; By performing a weighted least-squares fit between the relative offset of the group velocity and the relative offset of the theoretical group velocity, the initial value of the multi-band fitting residual is obtained, expressed as: ; in, The initial values ​​for the multi-band fitting residuals, These are the weighting coefficients. The set of effective frequency bands within a preset time window; The icing thickness and multi-band fitting residual of the p-th time window are obtained through numerical optimization, and are expressed as follows: ; in, For the first p Ice thickness within a time window, For the first p Multi-band fitting residuals for each time window; The first p The multi-band fitting residuals for each time window are compared with a preset residual threshold. If the residuals do not exceed the preset residual threshold, then... The output is the icing thickness for the p-th time window; If the residual exceeds a preset threshold, an abnormal frequency band is identified, triggering a frequency band fitting and redundancy point correction mechanism. Specifically, the residuals of each frequency band are checked one by one, identifying frequency bands with significant deviations from the overall fitting trend in the multi-band fitting residuals and removing them as outliers. The remaining frequency bands are then divided into low-frequency, mid-frequency, and high-frequency bands, and fitting calculations are performed separately to obtain three independent icing thickness results. If the icing thickness results of the three frequency bands maintain high consistency, the weighted average of the three results yields a corrected icing thickness value. If the three independent icing thickness results differ significantly, a frequency-related local anomaly is identified within the time window, and only the frequency band with high consistency is selected as a valid reference value. This method effectively suppresses the interference of abnormal frequency bands on the overall thickness estimation, thereby improving the stability and robustness of the inversion results. After completing outlier handling and frequency band fitting, the system performs a weighted average of the icing thickness results obtained from each frequency band or frequency zone. The weighting factor can be set according to the frequency band signal-to-noise ratio, the magnitude of the multi-frequency band fitting residual, or the correlation coefficient, so that higher-quality data has a greater weight in the final calculation. The resulting average icing thickness is the representative icing thickness within the time window. Finally, the average icing thickness is output as the icing thickness of the conductor within that time period and is used as part of the continuous time series for subsequent growth rate calculation and de-icing status determination. In this way, not only can the accuracy of the thickness results in a single window be guaranteed, but also continuous and reliable data input can be provided for trend analysis and dynamic monitoring.

[0033] It should be noted that by calculating the group velocity through propagation delay and comparing it with the ice-free baseline curve to obtain the relative offset, and then fitting it with the composite rod dispersion model, the final output of the icing thickness is achieved. This realizes a complete mapping chain of "signal → delay → group velocity → thickness", ensuring the physical interpretability of the thickness inversion. On the other hand, by repeatedly executing this thickness inversion process within a continuous time window, time series icing thickness data can be generated, and the growth rate and trend of icing thickness can be further derived. This not only meets the needs of static icing thickness detection, but also provides a real-time data foundation for icing development prediction and de-icing alarm.

[0034] In some embodiments of this application, the specific content of obtaining the sudden event of the conductor based on the acoustic emission signal of the conductor is as follows: a broadband acoustic emission sensor continuously collects high-frequency acoustic emission signals from the surface of the conductor and preprocesses the original signal. The preprocessing includes noise filtering, amplitude normalization and time window segmentation. Within a preset time window, a threshold triggering method is used to detect whether a sudden pulse with an amplitude exceeding the background noise appears in the acoustic emission signal. If a sudden pulse appears, it is recorded as a sudden event.

[0035] In some embodiments of this application, the specific content of determining the icing detachment state if the number and energy of the sudden events exceed a preset threshold and the group velocity rebounds is as follows: the starting point of the sudden event is accurately located using an adaptive information criterion method, and the characteristic parameters of the sudden event are calculated. The characteristic parameters include peak amplitude, duration, energy, and spectral centroid. The number and energy of the sudden events are compared with a preset threshold. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds at the same time, the conductor is determined to be in an icing detachment state.

[0036] Specifically, in the acquired acoustic emission signals, the adaptive information criterion method is used to accurately locate the starting point of the burst pulse. It can automatically determine the true starting point of the burst event based on the abrupt change in the statistical characteristics of the signal, thereby avoiding misjudgment caused by environmental noise fluctuations in the traditional fixed threshold method. Subsequently, feature extraction is performed on each located burst event, and key parameters such as peak amplitude, duration, energy and spectral centroid are calculated to characterize the intensity and spectral characteristics of the event. Within a preset time window, the number and energy of sudden events are counted and compared with preset quantity and energy thresholds. When the number of sudden events and the cumulative energy both exceed the thresholds, it indicates that a large-scale crack propagation or ice layer detachment process has occurred on the conductor. At the same time, combined with the group velocity change calculated from the angular velocity signal, if a significant rebound phenomenon of the group velocity occurs within the same time window, that is, the group velocity value rises from the slowing trend under the state of increased ice thickness to a level close to the ice-free baseline curve, it is further confirmed that the sudden event is related to the ice detachment process. Under the condition that the above two conditions are met, the conductor's operating status is marked as ice shedding state, and a corresponding ice shedding alarm is generated; if the number or energy of sudden events exceeds the threshold but the group velocity does not rebound, the event is marked as micro-cracks or slight peeling inside the ice layer rather than overall shedding, and no ice shedding alarm is triggered, thereby ensuring the accuracy and reliability of the judgment results.

[0037] In some embodiments of this application, the specific content of outputting the current icing thickness, growth rate, de-icing alarm information and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor, is as follows: The icing thickness of the conductor obtained by inversion within a preset time window is used as the ice layer thickness data point to form a thickness sequence that changes over time. The thickness sequence is differentially calculated to obtain the rate of change of the icing thickness of the conductor in adjacent time windows, which is used as the icing growth rate. At the same time, the conductor status is determined in conjunction with the sudden events that occur on the conductor. If it is determined to be an icing detachment state, a de-icing alarm information is generated. In the information output process, a confidence factor is introduced, and the detection results are divided into three levels: high confidence, medium confidence and low confidence, according to the magnitude of the confidence factor.

[0038] In some embodiments of this application, the confidence factor is specifically obtained by: obtaining a signal-to-noise ratio score based on the cross-correlation processing result of the angular velocity signal of the conductor, obtaining a consistency score based on the multi-band fitting residual, obtaining a correlation score based on the sudden events that occur on the conductor, and weightedly fusing the signal-to-noise ratio score, consistency score and correlation score to obtain the confidence factor.

[0039] Specifically, based on the cross-correlation processing results of the angular velocity signal of the conductor within a preset time window, the peak value and sidelobe energy ratio of the normalized correlation function are extracted, and the signal-to-noise ratio score is calculated to measure the clarity and stability of the propagation delay estimation. The sidelobe energy is a small peak value of some secondary peaks on both sides of the peak value, which may be caused by factors such as noise. Based on the multi-band fitting residual results, the consistency between the group velocity fitting values ​​of each frequency band and the theoretical dispersion curve is statistically analyzed. If the residual difference between each frequency band is small and the overall residual is below the threshold, a higher consistency score is given; otherwise, the score is reduced to reflect the reliability of ice thickness inversion. Based on the sudden events detected by the acoustic emission signal of the conductor, the correlation between the number of sudden events, energy and group velocity rebound is calculated. When the event characteristics are highly matched with the ice shedding process, a high correlation score is assigned. If there is no obvious correspondence between the two, the score is reduced to reflect the degree of coupling between dynamic events and thickness changes. Finally, the three types of scores are weighted and fused according to preset weights to obtain the confidence factor. The confidence factor is used to classify the final output of icing thickness, growth rate, and de-icing alarm information. When the confidence factor is in the high-level range, the output result is judged to be highly reliable and can be directly used as a reference for operation and maintenance scheduling. When the confidence factor is in the low-to-medium level range, it indicates that the detection result may be affected by noise, residuals, or event inconsistencies, and needs to be verified in conjunction with other monitoring methods.

[0040] See Figure 2 As shown, this embodiment of the invention provides an icing detection system for overhead transmission lines, the system comprising: The integrated torsional vibration and acoustic emission sensing unit is configured to collect the angular velocity signal and acoustic emission signal of the conductor in the power transmission line. The processing unit is configured to perform cross-correlation processing on the angular velocity signal of the conductor within a preset time window to obtain the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor. The calculation unit is configured to obtain the group velocity in multiple frequency bands based on the transmission line span length and propagation delay, compare the group velocity in each frequency band with the baseline group velocity curve established during the ice-free period of the conductor, and invert the ice thickness of the conductor by combining the composite pole dispersion model of the conductor and the ice layer. The status assessment unit is configured to detect sudden events occurring on the conductor based on the acoustic emission signal of the conductor. If the number and energy of the sudden events exceed a preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, it proceeds to the next step. The output unit is configured to output the current icing thickness, growth rate, de-icing alarm information, and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor.

[0041] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for detecting icing on overhead transmission lines, characterized in that, include: An integrated torsional vibration and acoustic emission sensing unit is installed at both ends of the transmission line span to collect the angular velocity signal and acoustic emission signal of the conductor in the transmission line. Based on the cross-correlation processing of the angular velocity signal of the conductor within a preset time window, the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor is obtained. Based on the transmission line span length and propagation delay, the group velocity in multiple frequency bands is obtained. The group velocity in each frequency band is compared with the baseline group velocity curve established during the ice-free period of the conductor. The ice thickness of the conductor is obtained by inversion using the composite pole dispersion model of the conductor and the ice layer. The sudden events that occur on the conductor are obtained based on the acoustic emission signals of the conductor. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, proceed to the next step. Based on the ice thickness of the conductor and sudden events, combined with the confidence factor, the output shows the current ice thickness, growth rate, de-icing alarm information, and confidence level.

2. The method for detecting icing on overhead transmission lines according to claim 1, characterized in that, The integrated torsional vibration and acoustic emission sensing unit includes an angular velocity sensor, a broadband acoustic emission sensor, a short-time ring actuator, and a clock module.

3. The method for detecting icing on overhead transmission lines according to claim 2, characterized in that, The specific content of the cross-correlation processing is as follows: within a preset time window, the angular velocity signal of the conductor is subjected to Fourier transform, the angular velocity signal of the conductor is decomposed into multiple frequency bands, and a generalized cross-correlation-phase transform is performed in each frequency band to obtain the normalized correlation function of the angular velocity signal of the conductor at both ends of the transmission line span. Based on the search cross-correlation peak position, the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor is obtained.

4. The method for detecting icing on overhead transmission lines according to claim 3, characterized in that, The specific content of the composite rod dispersion model of the conductor and the ice layer is as follows: based on the cross-sectional radius of the conductor, the pitch of the conductor, the number of strands of the conductor, and the material of the conductor as parameters, the equivalent polar inertia, the equivalent density of the conductor, and the equivalent shear modulus of the conductor are calculated. The polar inertia of the ice-covered annular cross section was calculated by using the ice layer thickness, ice material density, and ice shear modulus as parameters. Based on the conservation of mass and the torsional stiffness distribution relationship, the equivalent polar inertia of the conductor, the equivalent density of the conductor, the equivalent shear modulus of the conductor, the density of the icing material, the shear modulus of the icing material, and the polar inertia of the icing annular section are calculated to obtain the total polar inertia, the total equivalent density, and the total equivalent shear modulus.

5. The method for detecting icing on overhead transmission lines according to claim 4, characterized in that, The specific steps for obtaining the icing thickness of the conductor are as follows: Divide the transmission line span length by the propagation delay in each frequency band to obtain the group velocity in each frequency band. Compare the group velocity in each frequency band with the baseline group velocity curve established during the ice-free period of the conductor, and obtain the relative offset of the group velocity in each frequency band. Input the relative offset of the group velocity in each frequency band into the composite rod dispersion model of the conductor and the icing layer for fitting to obtain multi-frequency band fitting residuals. When the multi-frequency band fitting residuals do not exceed the preset residual threshold, the corresponding ice layer thickness is output as the icing thickness of the conductor. When the multi-frequency band fitting residuals exceed the preset residual threshold, the frequency band fitting and redundancy point correction mechanism is triggered to remove and correct points in abnormal frequency bands to obtain the average icing thickness within a preset time window. The average icing thickness is output as the icing thickness of the conductor.

6. The method for detecting icing on overhead transmission lines according to claim 5, characterized in that, The specific content of obtaining the sudden event of the conductor based on the acoustic emission signal of the conductor is as follows: the broadband acoustic emission sensor continuously collects the high-frequency acoustic emission signal on the surface of the conductor and preprocesses the original signal. The preprocessing includes noise filtering, amplitude normalization and time window segmentation. Within a preset time window, a threshold triggering method is used to detect whether there is a sudden pulse in the acoustic emission signal with an amplitude exceeding the background noise. If a sudden pulse appears, it is recorded as a sudden event.

7. The method for detecting icing on overhead transmission lines according to claim 6, characterized in that, The specific content of determining that if the number and energy of sudden events exceed a preset threshold and the group velocity rebounds, the conductor is determined to be in an icing detachment state is as follows: The adaptive information criterion method is used to accurately locate the starting point of the sudden event and calculate the characteristic parameters of the sudden event, including peak amplitude, duration, energy and spectral centroid. The number and energy of the sudden events are compared with the preset threshold. If the number and energy of the sudden events exceed the preset threshold and the group velocity rebounds at the same time, the conductor is determined to be in an icing detachment state.

8. The method for detecting icing on overhead transmission lines according to claim 7, characterized in that, The specific content of outputting the current icing thickness, growth rate, de-icing alarm information, and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor, is as follows: The icing thickness of the conductor obtained by inversion within a preset time window is used as the ice layer thickness data point to form a thickness sequence that changes over time. The thickness sequence is differentially calculated to obtain the rate of change of the icing thickness of the conductor within adjacent time windows, which is used as the icing growth rate. At the same time, the conductor status is determined in conjunction with sudden events that occur on the conductor. If it is determined to be a state of icing detachment, a de-icing alarm information is generated. In the information output process, a confidence factor is introduced, and the detection results are divided into three levels: high confidence, medium confidence, and low confidence, based on the magnitude of the confidence factor.

9. A method for detecting icing on overhead transmission lines according to claim 8, characterized in that, The confidence factor is specifically derived as follows: a signal-to-noise ratio score is obtained based on the cross-correlation processing results of the angular velocity signal of the conductor; a consistency score is obtained based on the multi-band fitting residual; a correlation score is obtained based on the sudden events that occur on the conductor; and the confidence factor is obtained by weighted fusion of the signal-to-noise ratio score, the consistency score, and the correlation score.

10. An icing detection system for overhead transmission lines, implemented as described in any one of claims 1-9, characterized in that, The system includes: The integrated torsional vibration and acoustic emission sensing unit is configured to collect the angular velocity signal and acoustic emission signal of the conductor in the power transmission line. The processing unit is configured to perform cross-correlation processing on the angular velocity signal of the conductor within a preset time window to obtain the propagation delay of the torsional guided wave of the conductor in multiple frequency bands of the angular velocity signal of the conductor. The calculation unit is configured to obtain the group velocity in multiple frequency bands based on the transmission line span length and propagation delay, compare the group velocity in each frequency band with the baseline group velocity curve established during the ice-free period of the conductor, and invert the ice thickness of the conductor by combining the composite pole dispersion model of the conductor and the ice layer. The status assessment unit is configured to detect sudden events occurring on the conductor based on the acoustic emission signal of the conductor. If the number and energy of the sudden events exceed a preset threshold and the group velocity rebounds, it is determined to be an ice-covered detachment state; otherwise, it proceeds to the next step. The output unit is configured to output the current icing thickness, growth rate, de-icing alarm information, and confidence level based on the icing thickness of the conductor and sudden events, combined with a confidence factor.

Images