Accurate positioning method of automatic transmission conductor breaking device

Abstract

The invention discloses a precise positioning method of an automatic transmission conductor breaking device, and relates to the technical field of line detection, and the method comprises the following steps: obtaining eddy current detection signal data of the surface of a transmission conductor; performing sliding window segmentation processing on the eddy current detection signal data, and extracting a transient impedance sequence; the transient impedance sequence is processed through a phase sensitive demodulation circuit, and impedance real part information and impedance imaginary part information are separated; impedance real part information features are extracted, impedance imaginary part information frequency domain energy is deconstructed, and an impedance track feature vector is generated; and carrying out space segmentation on the impedance track feature vector, and calculating defect position information and severity level based on the offset of the track center point. The problem that an existing wire detection technology is insufficient in automation level and difficult to meet the requirement for accurate defect positioning of an automatic wire breaking device is solved.

Description

Technical Field

[0001] This invention relates to the field of line detection technology, and more specifically, to a precise positioning method for an automatic disconnection device for power transmission lines. Background Technology

[0002] As a key component of power transmission, the structural integrity of transmission lines directly affects the safe and stable operation of the power grid. For a long time, conductor inspection has mainly relied on the following traditional methods: First, manual inspection, which uses telescope observation, infrared thermography and other means to assess the surface condition. This method is inefficient and heavily relies on the experience of the inspectors, making it difficult to detect internal defects and minor damage. Second, offline testing, which requires taking samples of the conductor and sending them to the laboratory for destructive testing such as metallographic analysis and tensile testing. This is not only complicated and costly, but can also cause permanent damage to the line.

[0003] Non-destructive testing methods based on eddy current testing technology are increasingly being applied to the field of conductor defect identification. These methods effectively identify defects such as surface cracks and internal corrosion by analyzing the distribution characteristics of eddy currents induced by alternating magnetic fields within the conductor. Typical applications include using impedance plane analysis to determine the defect type by observing changes in impedance trajectories, or using spectral analysis to extract the frequency domain response characteristics of defects at different depths. These methods have improved the automation level of inspection to a certain extent and provided new technical approaches for conductor condition assessment.

[0004] For example, the invention patent announcement CN112051326A discloses a crack identification method based on irregular area calculation, which includes the following steps: using an eddy current detector to detect the surface of a turbine runner blade to obtain the original impedance signal; automatically searching for the maximum value point in the imaginary part of the impedance signal; using the maximum value point as the origin, automatically searching for two near-zero points as the start and end points of the target signal to determine the range of the target signal; using the start and end points of the target signal to truncate the real and imaginary parts of the impedance and draw an impedance diagram; then, using the tension spline curve method in the impedance diagram to calculate the area of ​​the closed region in the target signal; dividing the calculated smaller area by the larger area, if the calculation result is greater than 0.1, it is identified as a crack, thus completing the crack identification based on irregular area calculation. This method can automatically identify crack defects.

[0005] For example, the invention patent announcement CN113777157A discloses a live-line detection system and method for cable lead seal defects based on impedance spectrum. This system includes a portable computer detection system and an eddy current detection sensor. The portable computer detection system includes a housing, a signal generation module, a signal receiving module, a signal processing module, a storage module, a power supply module, and a computer system housed within the housing, as well as a display screen and several buttons embedded in the housing. The eddy current detection sensor is a cylindrical probe, including a probe housing and an internal coil and magnetic core. The probe is connected to the portable computer detection system via a probe connector and wiring. The portable computer detection system processes and displays the impedance information of the probe coil collected by the eddy current detection sensor probe to obtain an impedance spectrum and analyze cable defects. This invention's detection system and method, based on the detection of cables under lead seal conditions, can promptly identify potential hazards in cable lines, ensuring power grid safety.

[0006] The above-disclosed technical solutions have at least the following technical problems: Existing technologies lack a dynamic adaptation mechanism between detection speed and signal quality, and cannot automatically adjust the scanning speed based on the real-time signal-to-noise ratio.

[0007] Existing solutions are mostly limited to the identification of defect types, lack the ability to accurately locate the absolute coordinates of defects, fail to establish an accurate mapping relationship between impedance characteristics and spatial location, and cannot provide accurate spatial positioning information for automatic disconnection devices.

[0008] To address the above problems, this invention proposes a solution. Summary of the Invention

[0009] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a precise positioning method for an automatic power transmission line disconnection device, so as to solve the problem that the existing conductor detection technology is not automated enough to meet the requirements of the automatic disconnection device for precise defect positioning.

[0010] To achieve the above objectives, the present invention provides the following technical solution: A precise positioning method for an automatic conductor disconnection device includes the following steps: During the detection process along the conductor, eddy current detection signals and displacement data are simultaneously acquired; the synchronous eddy current detection signals are segmented to extract transient impedance sequences; the moving speed is dynamically adapted and controlled based on the signal-to-noise ratio of the real and imaginary parts of the impedance in the transient impedance sequence; an electromagnetic coupling model is established to decouple the transient impedance sequence, separating the pure real impedance sequence and the imaginary impedance sequence; based on the decoupled real and imaginary impedance sequences, a feature vector is generated and fused with the displacement data to obtain preliminary defect positioning information and positional reliability; based on the positional reliability, the dynamic weight of the moving speed is corrected, and the absolute coordinates and severity level of the defect on the conductor are calculated and output.

[0011] In a preferred embodiment, the real and imaginary impedance information is obtained by decomposing the transient impedance sequence using a phase-sensitive demodulation circuit. Specifically, the transient impedance sequence is synchronously demodulated using the quadrature lock-in amplification principle to separate the component in phase with the reference signal as the real impedance information, which characterizes the resistance change characteristics of the conductor surface. The transient impedance sequence is then quadraturely demodulated to separate the component orthogonal to the reference signal as the imaginary impedance information, which characterizes the magnetic induction change characteristics inside the conductor. Finally, the real and imaginary impedance information are low-pass filtered to obtain a smooth real and imaginary impedance sequence.

[0012] In a preferred embodiment, the dynamic adaptation and control of the moving speed includes: calculating the real-time signal-to-noise ratio (SNR) of the impedance real part sequence and the impedance imaginary part sequence respectively; weighting and fusing the real-time SNR based on the dynamic weight of the moving speed to obtain a comprehensive SNR index; comparing the comprehensive SNR index with a preset SNR threshold range, and generating a speed control command based on the comparison result.

[0013] In a preferred embodiment, the step of establishing an electromagnetic coupling model to decouple the transient impedance sequence includes: establishing a coupled field distribution model containing electromagnetic interference from nearby conductors; reconstructing background interference components from the transient impedance sequence using the coupled field distribution model; eliminating the background interference components from the transient impedance sequence to obtain a decoupled transient impedance sequence; and performing quadratic orthogonal demodulation on the decoupled transient impedance sequence to separate the pure impedance real part sequence and the impedance imaginary part sequence.

[0014] In a preferred embodiment, generating the feature vector and fusing it with the displacement data includes: extracting time-domain features from the real part sequence of the pure impedance to obtain real part feature parameters; calculating the frequency-domain energy distribution of the imaginary part sequence of the pure impedance to obtain imaginary part spectral features; fusing the real part feature parameters and the imaginary part spectral features at the feature level to generate a comprehensive feature vector; and performing spatiotemporal registration of the comprehensive feature vector with the synchronously acquired displacement data to obtain preliminary location information and location confidence of the defect.

[0015] In a preferred embodiment, the step of correcting the dynamic weight of the moving speed based on the location confidence and calculating and outputting the absolute coordinates and severity level of the defect on the conductor includes: determining the defect type characteristics based on the preliminary location information of the defect; determining the weight correction coefficient based on the mapping relationship between the defect type characteristics and the location confidence; adjusting the weight ratio of each component in the dynamic weight of the moving speed according to the weight correction coefficient; updating the speed control strategy based on the corrected dynamic weight of the moving speed, while calculating the absolute coordinates of the defect in combination with the moving displacement data, and determining the severity level of the defect based on the feature vector amplitude.

[0016] In a preferred embodiment, the reconstructed background interference component specifically involves: based on the established coupled field distribution model, using the measured transient impedance sequence as a constraint, and performing inversion calculations on the model parameters under the assumption of no defects; matching the theoretical impedance output by the model with the measured transient impedance sequence in both amplitude and phase; and gradually adjusting the model parameters according to the minimum weighted error criterion to achieve the best fit between the theoretical response and the measured data, thereby forming the reconstructed background interference component.

[0017] In a preferred embodiment, the inversion calculation process further includes: compensating for magnetic flux disturbance based on the attitude and displacement information of the detection device to separate the static coupled field from the dynamic disturbance field; introducing a sparse constraint and phase residual correction mechanism to calculate the reconstructed background interference component using the corrected model parameters.

[0018] In a preferred embodiment, the model parameters include the inter-line mutual inductance coefficient, conductor conductivity and permeability parameters, and conductor induced current density distribution.

[0019] This invention achieves high-precision, adaptive, and fully automated detection and location of conductor defects by simultaneously acquiring eddy current signals and displacement data, dynamically controlling the detection speed based on the signal-to-noise ratio, decoupling and separating the pure impedance sequence based on an electromagnetic coupling model, and fusing multi-dimensional features and displacement information for defect location and confidence assessment. It also incorporates adaptive speed weighting based on location confidence and outputs the absolute coordinates and severity level of the defect, thus significantly improving the accuracy and operational efficiency of automatic conductor disconnection devices. Furthermore, closed-loop control and multi-feature fusion help suppress environmental interference, enhance system robustness, and effectively solve the problem that existing conductor detection technologies lack sufficient automation to meet the precise defect location requirements of automatic disconnection devices. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the precise positioning method of the automatic wire disconnection device for power transmission lines provided in an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1, Figure 1 The present invention provides a precise positioning method for the automatic power transmission line disconnection device, comprising the following steps: S1, During the movement detection along the conductor, eddy current detection signals and movement displacement data are collected simultaneously; S2, segment the synchronous eddy current detection signal and extract the transient impedance sequence; S3, dynamically adapts and controls the moving speed based on the signal-to-noise ratio of the real and imaginary parts of the impedance information in the transient impedance sequence; S4. Establish an electromagnetic coupling model to decouple the transient impedance sequence and separate the pure impedance real part sequence from the impedance imaginary part sequence. S5. Based on the decoupled impedance real part sequence and impedance imaginary part sequence, a feature vector is generated and fused with the displacement data to obtain the preliminary location information and location confidence of the defect. S6, based on the positional confidence level, corrects the dynamic weight of the moving speed, and calculates the absolute coordinates and severity level of the output defect on the conductor.

[0023] This embodiment suppresses interference through electromagnetic coupling decoupling, and combines signal-to-noise ratio adaptive speed regulation and multi-feature fusion positioning to achieve high-precision and automated detection and positioning of wire defects, effectively improving the accuracy, anti-interference ability and system adaptability of detection in complex environments.

[0024] S1, during the movement detection along the conductor, eddy current detection signals and movement displacement data are acquired simultaneously, specifically: The detection device, equipped with an eddy current detection probe and a displacement measurement module, is fixed on the power transmission line, and the automatic moving mechanism along the line is activated. While the detection device is moving, the built-in timing synchronization module performs time synchronization calibration on the eddy current detection signal sampling system and the displacement measurement system to ensure that the sampling times of the two types of data correspond one-to-one, and to achieve complete matching of signal and spatial position information.

[0025] In this embodiment, the eddy current detection signal acquisition step is specifically as follows: The detection device is equipped with a high-frequency excitation coil and an induction coil. A sinusoidal alternating current of a certain frequency is passed through the excitation coil, generating an alternating magnetic field around the coil. When the coil approaches the power transmission line, eddy currents are generated inside the line due to electromagnetic induction. The distribution of these eddy currents inside the line is closely related to whether there are defects such as fractures, corrosion, or inclusions on the surface or inside the line. When the probe moves along the line, the change in the distribution of eddy currents in the line causes changes in the amplitude and phase of the induced voltage of the detection coil, thus forming a time-series signal. The acquisition system samples this induced voltage at high frequency, generates a continuous sequence of raw eddy current detection signals, and stores it in the device's buffer.

[0026] In this embodiment, the displacement data acquisition step specifically includes: The displacement measurement module includes an coded displacement sensor and an inertial measurement unit. The coded sensor is mounted on the drive wheel axle of the moving mechanism and measures the linear displacement along the guide wire by pulse counting. The inertial measurement unit outputs acceleration and angular velocity information in real time to compensate and correct the displacement data, so as to eliminate measurement errors caused by wheel slip or guide wire tilt. The corrected displacement data is recorded synchronously with timestamps to achieve strict time correspondence with the eddy current detection signal.

[0027] It should be noted that the eddy current detection signal refers to the voltage response signal received by the probe's induction coil in an alternating magnetic field. The amplitude and phase of this signal change with the characteristics of the conductor defect and are key data for judging the structural integrity of the conductor.

[0028] S2, the synchronized eddy current detection signal is segmented, and the transient impedance sequence is extracted, specifically as follows: The original time-domain signal of the synchronously acquired eddy current detection is preprocessed to remove DC bias, power frequency interference, and high-frequency noise. The preprocessing includes three parts: baseline drift removal, bandpass filtering, and pulse artifact suppression. Baseline drift removal uses moving window least squares fitting and subtracts the fitting result to eliminate low-frequency trends. Bandpass filtering uses a well-designed digital bandpass filter, with the passband covering the excitation frequency and several harmonics nearby to retain frequency components related to the electromagnetic response of the tested conductor. Pulse artifact suppression combines median filtering with a short pulse removal algorithm to replace abnormal short-term abrupt change samples to prevent misjudgment during segmentation.

[0029] In this embodiment, the signal segmentation step based on energy and phase characteristics specifically includes: Short-time energy calculation and phase stability evaluation are performed on the preprocessed eddy current signal to determine the candidate positions of the segmentation window. The short-time energy uses the sliding window sum of squares statistic to reflect the location of the transient enhancement of the signal, and the phase stability uses a phase consistency metric based on reference excitation to eliminate false positives caused by noise. When both the energy and phase indicators exceed the preset threshold, they are identified as candidate mutation points. Neighboring candidate points are clustered and merged to eliminate redundant segmentation caused by high sampling rate. Finally, a set of non-overlapping time periods are obtained as valid signal segments.

[0030] In this embodiment, the transient impedance sequence extraction step specifically includes: For each valid signal segment obtained after segmentation, the in-phase component and quadrature component synchronized with the reference excitation are obtained within the segment window based on the phase-sensitive synchronous demodulation method. The demodulation result forms a complex voltage response sequence. The complex voltage response sequence is normalized according to the known excitation current amplitude and phase to obtain a complex impedance sequence, which is the transient impedance sequence. Subsequently, the complex impedance sequence is smoothed and denoised in a short time to retain the amplitude and phase changes related to the defect and suppress measurement noise. The transient impedance sequence also corresponds one-to-one with the synchronous displacement timestamp to form a transient impedance-displacement pair sequence with spatial annotation for subsequent decoupling and positioning.

[0031] It should be noted that the transient impedance sequence refers to the complex sequence of local electromagnetic response impedance at each moment obtained by segmenting the eddy current detection signal in time during the conductor movement detection process. This sequence reflects the transient changes in the electromagnetic characteristics of the conductor under test at different spatial positions. It is an electromagnetic characterization of local structural anomalies, material defects and geometric disturbances of the conductor. Unlike the traditional steady-state impedance, the transient impedance sequence does not describe the average response at a fixed frequency, but focuses on the time-space joint change process to capture the subtle differences in conductivity, permeability and geometric dimensions along the conductor direction.

[0032] S3 dynamically adapts and controls the moving speed based on the signal-to-noise ratio of the real and imaginary parts of the impedance information in the transient impedance sequence.

[0033] In this embodiment, the real and imaginary parts of the impedance are obtained by decomposing the transient impedance sequence using a phase-sensitive demodulation circuit, specifically: The transient impedance sequence is synchronously demodulated by using the quadrature lock-in amplification principle, and the component in phase with the reference signal is separated as the real part of the impedance information. The real part of the impedance information characterizes the resistance change characteristics of the conductor surface. The transient impedance sequence is subjected to orthogonal demodulation to separate the component orthogonal to the reference signal as the imaginary part of the impedance information, which characterizes the magnetic induction change characteristics inside the conductor. The real part and imaginary part of the impedance information are then subjected to low-pass filtering to obtain a smooth real part and imaginary part of the impedance sequence.

[0034] In this embodiment, the dynamic adaptation and control of movement speed includes: The real-time signal-to-noise ratio (SNR) of the impedance real part sequence and the impedance imaginary part sequence are calculated separately. Based on the dynamic weight of the movement speed, the real-time SNR is weighted and fused to obtain a comprehensive SNR index. The comprehensive SNR index is compared with a preset SNR threshold range, and a speed control command is generated based on the comparison result.

[0035] In this embodiment, the real-time signal-to-noise ratio calculation step is specifically as follows: Sliding window analysis is performed on the real and imaginary impedance sequences respectively, and the ratio of signal power to noise power is calculated in each time window. The signal power is obtained by mean square amplitude estimation, and the noise power is obtained by detrended residual statistical estimation. This forms the time-series distribution of the real and imaginary signal-to-noise ratios, reflecting the reliability of the impedance measurement signal at different spatial locations.

[0036] In this embodiment, the signal-to-noise ratio calculation formula is:

[0037] In the formula, For the overall signal-to-noise ratio index. The real part of the impedance signal-to-noise ratio is derived from the ratio of signal power to noise power due to the change in the surface impedance of the conductor. The signal-to-noise ratio (SNR) is the imaginary part of the impedance, which is derived from the ratio of signal to noise power based on the change in inductive reactance within the conductor. and These are the dynamic weighting coefficients for movement speed, used to automatically adjust the contribution of the real and imaginary parts to the overall signal-to-noise ratio based on the real-time detection environment. Both are adaptively calculated by the system control module based on the current movement speed and magnetic field stability.

[0038] It should be noted that the signal-to-noise ratio (SNR) is an important indicator for measuring the quality of eddy current detection signals. It is defined as the ratio of the useful power of the signal to the noise interference power, and directly determines the accuracy and stability of impedance measurement. The noise components include not only the electronic noise of the sensor, but also the dynamic noise introduced by the movement of the conductor, electromagnetic coupling inhomogeneity, and environmental magnetic field disturbances. When the overall signal-to-noise ratio (SNR) is higher than the upper threshold, the system determines that the detection signal quality is stable and outputs an acceleration control command to increase the movement detection speed. When it is within the preset range, the current movement speed is maintained to ensure the spatial resolution of data sampling. When it is lower than the lower threshold, a deceleration control command is output to extend the signal sampling time, thereby improving data integrity and subsequent positioning accuracy. The control module continuously updates the overall SNR based on the real-time feedback signal, forming a closed-loop speed regulation mechanism.

[0039] S4. Establish an electromagnetic coupling model to decouple the transient impedance sequence, separating the pure real part sequence and the imaginary part sequence of impedance. The specific steps are as follows: Establish a coupled field distribution model that includes electromagnetic interference from nearby conductors; The background interference component is reconstructed from the transient impedance sequence using the coupled field distribution model. The background interference component is removed from the transient impedance sequence to obtain the decoupled transient impedance sequence; The decoupled transient impedance sequence is subjected to quadratic demodulation to separate the pure impedance real part sequence and the impedance imaginary part sequence.

[0040] In this embodiment, establishing the coupled field distribution model including electromagnetic interference from nearby conductors specifically involves: Using the geometric parameters, conductivity, permeability, and spatial position of the surrounding conductors as input variables, the quasi-static Maxwell equations are used for solving. The eddy current density distribution and magnetic induction intensity distribution are obtained in the detection area through the finite element discretization method. The model also considers the mutual inductance effect and boundary effect between conductors, so that the calculated total magnetic field distribution can reflect the electromagnetic coupling characteristics between the conductors.

[0041] In this embodiment, the step of reconstructing the background interference components specifically includes: Based on the established coupled field distribution model, the measured transient impedance sequence is used as a constraint condition, and the model parameters are inverted and calculated under the defect-free assumption. During the inversion process, the spatial structure parameters of the conductor and adjacent conductors are used as input. The mutual inductance coefficient is used to characterize the magnetic coupling strength between conductors, and the conductor conductivity and permeability parameters are used to describe the electromagnetic properties of the conductor. Combined with the conductor induced current density distribution obtained from the model's internal solution, a preliminary theoretical impedance is formed. Within the set analysis time window, the system performs bidirectional amplitude and phase matching between the theoretical impedance output by the model and the measured transient impedance sequence. Based on the minimum weighted error criterion, the mutual inductance coefficient, conductivity, and induced current density distribution are gradually adjusted to achieve the best fit between the theoretical response and the measured data. During parameter updates, the system compensates for magnetic flux disturbances caused by motion based on the attitude and displacement information of the detection device, thereby separating the static coupled field from the dynamic disturbance field. Simultaneously, in model optimization, a parameter selection method is used to retain the main influencing parameters, and phase error correction is applied to improve computational accuracy, ensuring the physical rationality and computational stability of parameter convergence. After iterative convergence, the theoretical impedance calculated using the corrected mutual inductance matrix, conductivity, and induced current density becomes the reconstructed background interference component. This component reflects the magnetic coupling effect of neighboring conductors, conductor material properties, and environmental disturbance characteristics, providing a stable and reliable background reference for subsequent decoupling calculations.

[0042] In this embodiment, eliminating the background interference component specifically involves: Vector difference operation is performed on the transient impedance sequence to subtract the reconstructed background interference components from the original impedance sequence sample by sample to obtain the decoupled transient impedance sequence. This process is carried out in the complex impedance domain, which eliminates both amplitude shift and phase drift. The impedance sequence after decoupling retains only the response components related to the electromagnetic properties of the target conductor itself, avoiding interference from the coupling of adjacent conductors to subsequent feature recognition.

[0043] In this embodiment, the step of performing quadratic demodulation on the decoupled transient impedance sequence specifically involves: The decoupled transient impedance sequence is input into the secondary quadrature demodulation module and phase-sensitive detection is performed again to improve the suppression of residual noise. During the secondary demodulation process, a reference signal that is strictly synchronized with the initial excitation signal is used for in-phase and quadrature decomposition to separate the pure real part and imaginary part of the impedance sequence. The real part sequence is used to characterize the surface resistivity and geometric defect characteristics of the conductor, while the imaginary part sequence is used to characterize the changes in permeability and eddy current diffusion characteristics inside the conductor.

[0044] In this embodiment, the formula for the coupled field distribution model is:

[0045]

[0046] In the formula, It is the magnetic field intensity vector, which originates from the induced magnetic field between the detection coil and the conductor being measured; The electric field intensity vector is formed by the excitation voltage generated by the detection coil; The source current density represents the input excitation of the detection excitation coil; The electrical conductivity of a conductor is derived from the electrical properties of the wire material. The induced electric field term generated by the moving conductor is used to compensate for differences in spatial magnetic flux distribution; Permeability is a parameter derived from the internal magnetic response of a conductor. For time.

[0047] In this embodiment, the formula for reconstructing the background interference component is:

[0048] In the formula, For the reconstructed background interference components, The measured transient impedance sequence is derived from the complex impedance sampling at the output of the detection coil. To theoretically calculate impedance based on the coupled-field model, Mutual inductance is the coefficient between conductors, reflecting the magnetic coupling strength between the conductor and adjacent conductors. The induced current density in the conductor is derived from the electric field distribution within the model. For the conductor's conductivity, [ , [This represents the analysis time window.]

[0049] In this embodiment, the formula for performing vector step difference calculation on the transient impedance sequence is:

[0050] In the formula, For decoupling transient impedance sequence, and These are the real and imaginary parts of the original impedance, respectively. and To reconstruct the real and imaginary parts of the background component, It is the imaginary unit.

[0051] In this embodiment, the quadratic demodulation formula is:

[0052]

[0053] In the formula, This is a sequence of the pure impedance real parts. This is a sequence of the imaginary part of the pure impedance. The angular frequency of the excitation signal is taken from the system's excitation source. This is the phase synchronization correction angle, derived from the system synchronization control module. One complete demodulation cycle is used for integrating and smoothing residual noise. The system time constant is derived from the electromagnetic response inertia of the conductor.

[0054] It should be noted that the electromagnetic coupling model is a physical model that describes the electromagnetic interaction between the detection device and the surrounding conductors. Its core is to quantitatively characterize the interference effect of the external conductor on the detection signal through the field quantity correlation between mutual inductance and induced current. In this embodiment, the model adopts a quasi-static approximation form, assuming that the operating frequency is low and the field propagation delay is negligible, in order to ensure the stability and real-time performance of the calculation. The bidirectional matching refers to the simultaneous comparison and analysis of the theoretical impedance and the measured impedance sequence from two dimensions: signal amplitude and phase. The weighted sum of amplitude difference and phase offset is used as the matching degree evaluation index. The compensation specifically refers to mathematical modeling and signal correction of the additional magnetic flux disturbance caused by position changes and attitude adjustments during the movement of the detection device. By establishing a functional relationship between motion parameters and magnetic field distribution, the interference component caused by pure motion is separated from the measured signal, thereby retaining the effective electromagnetic response signal that is only related to the defect. Secondary orthogonal demodulation is a process of re-phase correction and separation of the demodulation result, which aims to further reduce the residual cross-coupling effect and the system phase mismatch error. By repeating the demodulation process, the orthogonal purity between the real and imaginary parts of the impedance can be significantly improved, thereby obtaining a more physically representative pure impedance sequence.

[0055] S5. Based on the decoupled impedance real part sequence and impedance imaginary part sequence, generate feature vectors and fuse them with the displacement data to obtain preliminary defect location information and location confidence. The specific steps are as follows: Time-domain feature extraction is performed on the pure impedance real part sequence to obtain the real part feature parameters; The frequency domain energy distribution of the imaginary part sequence of pure impedance is calculated to obtain the spectral characteristics of the imaginary part; The real part feature parameters and the imaginary part spectral features are fused at the feature level to generate a comprehensive feature vector. By performing spatiotemporal registration of the integrated feature vector with the synchronously acquired displacement data, preliminary location information and location reliability of the defect are obtained.

[0056] In this embodiment, the step of extracting time-domain features from the pure impedance real part sequence specifically includes: The decoupled impedance real part sequence is segmented into time windows. The mean and variance of the real part sample points within each time window are calculated to describe the fluctuation of the impedance real part in the time dimension. Then, autocorrelation function analysis is performed on each time window to extract the periodic correlation features within the sequence. The transient energy peak is extracted by the envelope detection algorithm. Finally, a multi-dimensional time-domain feature parameter set including mean drift rate, variance ratio, and energy peak intensity is formed, providing basic data of real part features for subsequent fusion.

[0057] In this embodiment, the frequency domain energy distribution calculation of the imaginary part sequence of the pure impedance is specifically performed as follows: The impedance imaginary part sequence is input into the fast Fourier transform module to obtain the spectral amplitude distribution; the energy density of different frequency bands is calculated by frequency band division and energy integration, and the main frequency energy peak, frequency band energy ratio and energy centroid drift are extracted; at the same time, the energy difference index of the imaginary part spectrum between high and low frequency bands is calculated to reflect the non-uniformity of dielectric response; the obtained parameter set constitutes the imaginary part spectral feature, which is used to characterize the dielectric response mode of the frequency domain signal.

[0058] In this embodiment, the step of performing feature-level fusion of the real part feature parameters and the imaginary part spectral features to generate a comprehensive feature vector is as follows: First, the two sets of features are linearly normalized to eliminate the influence of units and orders of magnitude. Then, a principal component mapping matrix is ​​introduced to perform principal component weighted fusion on the normalized real and imaginary features. The comprehensive feature vector is composed of the main feature components and can simultaneously reflect the changing trends of resistivity and reactance, providing a stable feature expression for defect identification.

[0059] In this embodiment, the step of performing spatiotemporal registration of the integrated feature vector with the synchronously acquired displacement data is as follows: Based on the time synchronization signal of the detection device, the comprehensive feature vector and the displacement sampling sequence are timestamped; the actual position of the feature sampling point in the three-dimensional coordinate system is calculated using a spatial interpolation algorithm; in the spatial domain, the preliminary distribution information of the defect position is obtained by analyzing the correspondence between the feature vector change gradient and the displacement trajectory; finally, the defect positioning confidence is calculated based on the feature change amplitude and the degree of consistency of displacement, and the preliminary result of the defect spatial position is output.

[0060] In this embodiment, the time-domain feature extraction formula is:

[0061] In the formula, This represents the time-domain fluctuation energy of the impedance real part sequence, used to characterize the amplitude stability of the signal. Let be the real part of the impedance at time i. This represents the average real impedance within that time window. This represents the number of sampling points in the current time window.

[0062] In this embodiment, the formula for the imaginary part spectrum energy distribution is:

[0063] In the formula, Let be the normalized energy density of the k-th frequency component, representing the proportion of energy concentration of the imaginary part of the signal in the frequency domain. The Fourier transform amplitude of the imaginary part sequence at the k-th frequency point is derived from the imaginary part spectrum. This represents the total number of sampling points for the spectrum.

[0064] In this embodiment, the feature fusion formula is:

[0065] In the formula, The combined feature vector serves as the high-dimensional feature output after fusion. This is the normalized eigenvector with real parts. This is the normalized imaginary part spectral eigenvector. and These are the weighting coefficients for the real and imaginary features, respectively, derived from the feature importance analysis matrix, and their values ​​are automatically determined based on the feature contribution.

[0066] In this embodiment, the formula for calculating the location reliability is:

[0067] In the formula, The defect location reliability score represents the level of confidence in the defect location results. This is the comprehensive feature vector of the current measurement point. This is the reference feature vector for the normal region. This is the feature matching error factor, used to measure the mode shift caused by feature differences. The spatial displacement difference originates from the motion detection path data and is used to reflect the spatiotemporal registration error.

[0068] It should be noted that feature-level fusion differs from decision-level fusion. It performs data layer synthesis in the feature space, enabling the fused features to simultaneously retain the statistical attributes and trends of the real and imaginary parts of impedance. Spatiotemporal registration refers to establishing a correspondence between impedance features and spatial locations through synchronized timestamps and displacement coordinates, ensuring that each feature point has a clear spatial reference. Normalized real and normalized imaginary feature parameters represent resistivity and reactance response indices under unified dimensions, respectively. The principal component mapping matrix is ​​a linear mapping relationship established through covariance decomposition, used to achieve dimensionality reduction and decorrelation of multidimensional features. The feature matching error factor λ is a comprehensive parameter used for confidence assessment, which depends on the Euclidean distance distribution between features; the smaller the value, the higher the feature stability. The final defect location confidence can be used as input for subsequent spatial clustering analysis and defect boundary identification.

[0069] S6. Based on the positional confidence level, adjust the dynamic weight of the moving speed, and calculate the absolute coordinates and severity level of the output defect on the conductor. The specific steps are as follows: Determine the defect type characteristics based on the preliminary defect location information; Based on the mapping relationship between defect type characteristics and location reliability, the weight correction coefficient is determined; The weight ratio of each component in the dynamic weight of movement speed is adjusted according to the weight correction coefficient; The speed control strategy is updated based on the modified dynamic weight of the moving speed. At the same time, the absolute coordinates of the defect are calculated by combining the moving displacement data, and the severity level of the defect is determined according to the amplitude of the feature vector.

[0070] In this embodiment, determining the defect type characteristics based on the preliminary defect location information specifically involves: Based on the initial location coordinates and comprehensive feature vectors of the defects, the amplitude and phase structure parameters of the local impedance response are extracted, including the real part rate of change, the imaginary part response delay, and the spectral energy ratio. These parameters are then input into a preset defect type identification model. The model uses the feature templates obtained through training to perform pattern matching on the defects, thereby identifying the defect type, such as corrosion defects, mechanical cracks, insulation degradation, or wire wear. Each type of defect corresponds to a set of feature templates and typical response feature intervals.

[0071] In this embodiment, determining the weight correction coefficient based on the mapping relationship between defect type features and location confidence specifically involves: In the established confidence type mapping matrix, different types of defects correspond to different confidence decay functions. The system calculates the weight correction coefficient based on the defect type and the current confidence value, which is used to correct the subsequent dynamic speed weight. When the confidence is low, the weight correction coefficient tends to amplify the position sampling accuracy weight to increase the spatial sampling density. When the confidence is high, the weight correction coefficient tends to enhance the speed smoothness weight to improve detection efficiency and path stability.

[0072] In this embodiment, adjusting the weight ratio of each component in the dynamic weight of movement speed according to the weight correction coefficient specifically means: The correction coefficient is introduced into the dynamic velocity weight calculation module to adjust the ratio of velocity component, direction component and acceleration component in real time. Through the weighted balancing strategy, the detection device automatically reduces the moving speed and increases the measurement density when it is close to the defect area, and maintains the normal scanning rate in the non-defect area. The adjusted dynamic weight vector is used to guide the movement control of the device in the next cycle.

[0073] In this embodiment, the speed control strategy is updated based on the modified dynamic weight of the moving speed. Simultaneously, the absolute coordinates of the defect are calculated using the moving displacement data, and the severity level of the defect is determined based on the eigenvector magnitude. Specifically: First, the corrected velocity control signal is integrated in the time domain, and the precise position of the detection device in the conductor coordinate system is calculated by combining the displacement data. Using the preliminary defect location coordinates as a reference, the spatial coordinates are corrected by the error compensation function to obtain the absolute coordinates of the defect. Then, the severity level of the defect is calculated based on the amplitude of the comprehensive feature vector and the frequency domain energy concentration, and it is divided into three categories: slight, moderate and severe.

[0074] In this embodiment, the formula for calculating the weight correction coefficient is:

[0075] In the formula, This is the weight adjustment coefficient at the current moment, representing the strength of the adjustment effect of the confidence level on the dynamic weights. These are the defect type feature coefficients, determined by the defect identification model based on the type feature template. This is the confidence decay rate coefficient, which reflects the sensitivity of the weight to changes in confidence.

[0076] In this embodiment, the dynamic weight adjustment formula for movement speed is:

[0077]

[0078] In the formula, For the corrected velocity component weights, The corrected acceleration component weights, and These are the original velocity and acceleration weights, respectively. and These are weighting balancing coefficients, derived from the system control parameter configuration, used to adjust the relative importance of velocity and acceleration in the dynamic response. This refers to the aforementioned weighting correction coefficient.

[0079] In this embodiment, the formula for calculating the absolute coordinates is:

[0080] In the formula, The absolute coordinates of the defect in the traverse coordinate system. The initial spatial location of the detection point. The instantaneous value of the corrected speed control signal over time is derived from the speed output after dynamic weighting correction. The integral process represents the cumulative displacement calculation of the device over time, used to obtain the precise spatial location of the defect.

[0081] In this embodiment, the formula for calculating the severity level of the defect is:

[0082] In the formula, This is the comprehensive value for the severity level of the defects. The magnitude of the real part of the eigenvector. This is a reference value for the real part characteristic amplitude under normal conditions. The energy concentration of the imaginary part of the spectrum. This is a reference value for the imaginary part spectral energy under normal conditions. and These are the weighting coefficients for the real and imaginary features, used to balance the impact of the two types of features on the severity of the defect.

[0083] It should be noted that the defect type characteristic refers to the defect category identified through the impedance sequence response pattern, which is jointly determined by the dynamic changes of the real and imaginary parts of the impedance; the weight correction coefficient is used to establish a quantitative correlation between the positioning confidence and the motion control strategy, enabling the detection device to dynamically adjust the scanning strategy within different confidence intervals; the dynamic weight of the moving speed refers to the multi-dimensional weight set used to balance speed, direction, and acceleration control during the conductor detection process, through which adaptive optimization of the detection path can be achieved; the integral term in the absolute coordinate calculation formula is essentially the spatiotemporal accumulation of the corrected speed signal to ensure that the positioning accuracy is not affected by speed fluctuations; the comprehensive severity level value, by combining the real part amplitude and the imaginary part energy information, can reflect the electromagnetic response intensity and physical degradation degree of the defect, and the higher the value, the more severe the defect.

[0084] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0085] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0086] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0087] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0088] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0089] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A precise positioning method for an automatic power transmission line disconnection device, characterized in that, Includes the following steps: Simultaneously acquire eddy current detection signals and displacement data during the movement along the conductor; segment the eddy current detection signals and extract the transient impedance sequence corresponding to the displacement point; The detection moving speed is dynamically adjusted based on the signal-to-noise ratio in the transient impedance sequence; The transient impedance sequence is decoupled to separate the pure real part sequence and the imaginary part sequence of impedance; By fusing the impedance real part sequence, impedance imaginary part sequence, and displacement data, a feature vector is generated, and the location confidence of the defect is obtained; Adjust the weights based on the location reliability correction speed, calculate and output the absolute coordinates and severity level of the defect.

2. The precise positioning method of the automatic transmission line disconnection device according to claim 1, characterized in that, The separation of the pure impedance real part sequence and the impedance imaginary part sequence includes: The acquired transient impedance sequence is subjected to orthogonal demodulation processing to separate the first orthogonal component as the real part of the impedance and the second orthogonal component as the imaginary part of the impedance. The real part and imaginary part of the impedance information are filtered to obtain smooth sequences of the real part and imaginary part of the impedance.

3. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 2, characterized in that, The dynamic adjustment of the detection movement speed includes: Calculate the real-time signal-to-noise ratio of the impedance real part sequence and the impedance imaginary part sequence respectively; Based on the dynamic weight of movement speed, the real-time signal-to-noise ratio is weighted and fused to obtain a comprehensive signal-to-noise ratio index; The overall signal-to-noise ratio (SNR) index is compared with the preset SNR threshold range, and a speed control command is generated based on the comparison result.

4. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 3, characterized in that, The decoupling of the transient impedance sequence includes: An electromagnetic coupling field distribution model is established based on the layout parameters of the adjacent conductors; The background electromagnetic interference component is reconstructed from the transient impedance sequence using an electromagnetic coupling field distribution model. By eliminating background electromagnetic interference components from the transient impedance sequence, a decoupled transient impedance sequence is obtained. The decoupled transient impedance sequence is orthogonally demodulated to separate the pure real part sequence and the imaginary part sequence of impedance.

5. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 4, characterized in that, The process of fusing the impedance real part sequence, the impedance imaginary part sequence, and the displacement data to generate a feature vector includes: Time-domain feature extraction is performed on the impedance real part sequence to obtain real part feature parameters; The frequency domain energy distribution of the impedance imaginary part sequence is calculated to obtain the spectral characteristics of the imaginary part; The real part feature parameters and the imaginary part spectral features are fused at the feature level to generate a comprehensive feature vector. By performing spatiotemporal registration of the integrated feature vector with the synchronously acquired displacement data, preliminary location information and location reliability of the defect are obtained.

6. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 5, characterized in that, The step of adjusting the speed weights based on location confidence, calculating and outputting the absolute coordinates and severity level of the defect includes: The defect type characteristics are determined based on the feature vector, and the weight correction coefficient is obtained according to the mapping relationship between the type characteristics and the location confidence. The dynamic weight of the movement speed is adjusted using the weight correction coefficient. The speed control strategy is updated based on the adjusted dynamic weight of the movement speed, and the absolute coordinates of the defect are calculated based on the movement displacement data. The severity level of the defect is determined by combining the feature vector.

7. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 4, characterized in that, The process of reconstructing the background electromagnetic interference component from the transient impedance sequence includes: Based on the pre-defined electromagnetic coupling model, under the assumption of no defects, the measured transient impedance sequence is used as a constraint condition for model parameter inversion calculation. The matching error between the theoretical impedance output by the calculation model and the measured transient impedance sequence is calculated. The model parameters are iteratively adjusted according to the weighted error minimization criterion to achieve the best fit between the theoretical impedance and the measured transient impedance sequence, so as to generate the reconstructed background interference component.

8. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 7, characterized in that, The inversion calculation process also includes: Based on the attitude and displacement information of the detection device, magnetic flux disturbance is compensated to achieve separation of static coupled field and dynamic disturbance field; Sensitivity analysis was performed on the electromagnetic coupling model parameters to screen key parameters, and the measured phase data was compensated and corrected based on the preset phase-frequency mapping relationship to optimize the model parameters. The reconstructed background interference components are calculated using the corrected parameters.

9. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 8, characterized in that, The key parameters are model parameters whose impact on impedance response exceeds a preset threshold.

10. The precise positioning method of the automatic wire disconnection device for power transmission lines according to claim 8, characterized in that, The segmentation is based on displacement data, and the synchronous eddy current detection signal is divided into segments with equal intervals or equal displacements to extract the transient impedance sequence.

Images