Cable full life cycle condition monitoring method based on electromagnetic harmonic characteristics analysis

Abstract

The present invention belongs to the technical field of power equipment monitoring and relates to a cable full life cycle status monitoring method based on electromagnetic harmonic characteristic analysis, including: generating a discrete signal sequence; generating a denoised signal sequence through threshold filtering and reconstruction; dividing the denoised signal sequence into a segmented signal set according to a preset segment length, each segment of the sub-signal corresponding to a local area of ​​the cable's physical length; extracting multi-dimensional feature parameters and generating a global feature vector by segment combination; judging the degree of cable aging through multi-level thresholds, and triggering an alarm and positioning process based on the judgment results; generating a multi-node signal data set; calculating the distance range of the fault point relative to each monitoring node to generate a candidate distance set; performing weighted fusion on the candidate distance set, and outputting the final positioning result of the fault point. The present invention solves the problem that the positioning algorithm of the traditional method does not integrate multi-dimensional features and spatiotemporal correlation.

Description

Technical Field

[0001] The present invention belongs to the technical field of power equipment monitoring and relates to a cable full life cycle status monitoring method based on electromagnetic harmonic characteristic analysis. Background Art

[0002] Cables, the core carrier of power transmission, are prone to failures such as partial discharge and breakdown over long periods of operation due to insulation aging, mechanical stress, or environmental corrosion, threatening power grid security. Traditional monitoring methods rely on regular manual inspections and partial discharge testing, but they struggle to capture early signs of aging in real time and cannot accurately locate hidden fault points, resulting in delayed fault warnings and inefficient repairs. Especially in complex electromagnetic environments or long cable runs, noise interference and signal attenuation further reduce monitoring reliability.

[0003] Conventional solutions often use a single sensor to monitor the signal at the head end of the cable, combined with threshold alarms or impedance methods to locate faults. For example, the fault distance is estimated by measuring changes in cable impedance, or time-domain reflectometry is used to capture signal reflections for rough positioning. While these methods can achieve basic monitoring, they are limited by the narrow signal bandwidth and weak anti-interference capabilities, making it difficult to distinguish between noise and true fault characteristics, and prone to false alarms and missed faults. Furthermore, traditional positioning technology relies on data from a single node and cannot address issues such as multipath propagation and nonlinear signal attenuation. Positioning errors often reach tens of meters, making it difficult to meet high-precision requirements.

[0004] Based on the above problems, the traditional positioning algorithm fails to integrate multi-dimensional features and spatiotemporal correlations, resulting in poor stability of the results. Summary of the Invention

[0005] In order to solve the above problems, the present invention provides a cable full life cycle status monitoring method based on electromagnetic harmonic characteristic analysis.

[0006] The cable life cycle condition monitoring method based on electromagnetic harmonic characteristics analysis includes the following steps:

[0007] S1. Acquire the continuous electromagnetic harmonic signal naturally generated during the operation of the cable, and discretize it at a preset sampling frequency to generate a discrete signal sequence;

[0008] S2, decompose the discrete signal sequence, separate the noise component and the effective harmonic component, and reconstruct the denoised signal sequence through threshold filtering;

[0009] S3, dividing the denoised signal sequence into a set of segmented signals according to a preset segment length, where each segment of the sub-signal corresponds to a local area of ​​the physical length of the cable;

[0010] S4, performing a joint analysis of the time domain and frequency domain on each sub-signal in the segmented signal set, extracting multi-dimensional feature parameters, and generating a global feature vector by segment combination;

[0011] S5. Based on the global feature vector, determine the degree of cable aging through multi-level thresholds, and trigger an alarm and positioning process according to the determination result;

[0012] S6. When it is determined that the cable is in a serious aging state, multiple monitoring nodes along the cable are activated to synchronously collect electromagnetic harmonic signals to generate a multi-node signal data set;

[0013] S7, analyzing the characteristic parameter attenuation law and phase propagation delay difference of the multi-node signal data set, calculating the distance range of the fault point relative to each monitoring node, and generating a candidate distance set;

[0014] S8. Perform weighted fusion on the candidate distance set to screen the optimal solution that meets the signal attenuation consistency, phase propagation delay matching, and energy distribution correlation, and output the final fault location result.

[0015] A further embodiment of the present invention generates a discrete signal sequence, comprising the following steps:

[0016] The electromagnetic harmonic sensor is installed at the head end of the cable or a preset monitoring point. The electromagnetic harmonic sensor has a built-in broadband induction coil, which is used to couple the alternating electromagnetic field between the cable conductor and the shielding layer to generate a continuous electromagnetic harmonic signal.

[0017] The continuous electromagnetic harmonic signal is periodically sampled by an analog-to-digital conversion circuit to generate the discrete signal sequence. The sampling frequency is determined according to the rated voltage level of the cable and the highest harmonic frequency.

[0018] A further solution of the present invention generates a denoised signal sequence by threshold filtering and reconstruction, comprising the following steps:

[0019] Discrete signal sequence input signal decomposition algorithm, decomposing the signal into multiple sub-signal components according to different frequency bands;

[0020] A dynamic threshold is set to determine the boundary between noise and effective harmonics. Sub-signal components above the threshold are considered noise and set to zero, while sub-signal components below the threshold are retained as effective harmonics. The dynamic threshold is adaptively calculated based on the noise standard deviation of the sub-signal component and the signal length.

[0021] All processed sub-signal components are combined inversely according to the original frequency band to generate a denoised signal sequence.

[0022] A further solution of the present invention is to divide the denoised signal sequence into a set of segmented signals according to a preset segment length, comprising the following steps:

[0023] The physical length is determined based on the ratio of the total cable length to the preset segment length;

[0024] Calculate the number of sampling points contained in each sub-signal in combination with the preset sampling frequency;

[0025] Based on the number of sampling points, continuous data segments are intercepted from the denoised signal sequence in time order to generate a segmented signal set corresponding to a local area of ​​the cable physical length.

[0026] A further solution of the present invention is to extract continuous data segments from the denoised signal sequence in chronological order, comprising the following steps:

[0027] If the remaining signal length is less than one segment, it is extended to a complete segment through zero padding. Zero padding refers to the processing method of filling zero-value data at the end of the signal to meet the segment length requirement. The zero padding operation only fills the end of the signal with zero values. During feature extraction, the valid data segment is automatically identified to eliminate the interference of the zero-padding part on the analysis.

[0028] A further solution of the present invention generates a global feature vector, comprising the following steps:

[0029] Each sub-signal of the segmented signal set is converted from a time domain signal to a frequency domain signal through fast Fourier transform, and the ratio of the square of each harmonic amplitude to the sum of the square of the total harmonic amplitude is calculated to calculate the harmonic energy distribution characteristics;

[0030] The fluctuation degree of the signal amplitude is counted in the time domain, and the signal fluctuation characteristics are generated by calculating the ratio of the maximum value of the signal to the root mean square value;

[0031] Analyze the phase difference between adjacent sub-signals at the same frequency to generate phase consistency features;

[0032] The three types of features of each sub-signal are concatenated in segmented order to generate a global feature vector.

[0033] A further solution of the present invention uses a multi-level threshold to determine the degree of cable aging and triggers an alarm and positioning process based on the determination result, including the following steps:

[0034] The value of each characteristic parameter in the global characteristic vector is sequentially compared with the preset safety threshold and fault threshold;

[0035] If the characteristic parameters do not exceed the preset safety threshold, the cable is determined to be in a healthy state;

[0036] If the characteristic parameter exceeds the safety threshold but is lower than the fault threshold, the cable is judged to be in a mild aging state and a warning signal is generated;

[0037] If any characteristic parameter exceeds the fault threshold, the cable is judged to be in a serious aging state, a fault alarm signal is generated, and the positioning process is activated.

[0038] A further solution of the present invention generates a multi-node signal data set, comprising the following steps:

[0039] When a severe aging alarm signal is triggered, multiple monitoring nodes deployed at the beginning, middle, and end of the cable send activation instructions;

[0040] After receiving the instructions, each monitoring node is synchronously triggered by a unified clock source to collect electromagnetic harmonic signals, and a multi-node signal data set is generated by timestamp alignment; the monitoring nodes are deployed at preset positions at the beginning, middle and end of the cable.

[0041] A further solution of the present invention generates a candidate distance set, comprising the following steps:

[0042] For the electromagnetic harmonic signals collected by each monitoring node, the harmonic energy distribution characteristics, signal fluctuation characteristics, and phase consistency characteristics are extracted;

[0043] Based on the characteristic parameter attenuation model and the phase propagation model, the fault distance interval corresponding to each monitoring node is calculated, and the overlapping intervals are screened through cross-validation to generate the candidate distance set; the attenuation model is an exponential decay function of the harmonic amplitude with distance, and the phase propagation model is a linear relationship between phase difference and distance.

[0044] A further solution of the present invention outputs the final location result of the fault point, including the following steps:

[0045] For each candidate distance in the candidate distance set, calculate its signal attenuation consistency index, phase propagation delay matching index, and energy distribution correlation index with each monitoring node;

[0046] The three types of indicator values ​​are weighted and summed according to the set weight coefficients of each indicator to generate a comprehensive score;

[0047] The candidate distance with the highest comprehensive score is selected as the optimal solution and converted into the actual fault location based on the physical coordinates of the cable head end.

[0048] In summary, the present invention has the following beneficial technical effects:

[0049] 1. Electromagnetic harmonic sensors capture electromagnetic signals from cables in real time. Combined with signal decomposition and multi-dimensional feature extraction techniques, this effectively separates noise interference and extracts key harmonic features reflecting insulation degradation, resolving the misjudgment problem often associated with environmental noise in traditional methods. A multi-level threshold-based judgment mechanism distinguishes between healthy status, mild aging, and severe aging, significantly improving the accuracy of aging assessments. This provides a reliable basis for preventive maintenance and avoids power outages caused by sudden failures.

[0050] 2. Utilizing multi-node collaborative acquisition and spatiotemporal synchronization technology, the system analyzes signal attenuation patterns and phase propagation delay differences to construct a multidimensional constraint model to screen candidate distance sets, overcoming the vulnerability of single-node positioning to the nonlinear effects of signal attenuation. The weighted fusion algorithm further integrates signal consistency, delay matching, and energy correlation indicators, making it particularly suitable for complex fault scenarios in long-distance cables and significantly reducing manual inspection costs.

[0051] 3. The combination of dynamic threshold filtering, segmented signal refinement analysis, and adaptive characteristic parameter calibration technology enables the system to adapt to cable monitoring requirements across different voltage levels, material properties, and environmental interference. The introduction of global eigenvectors and a multi-level decision model automates the entire process from raw signal to fault decision-making, reducing reliance on manual intervention. It also supports historical data backtracking and iterative model optimization, providing a scalable technical foundation for intelligent power grid operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. The drawings are used to provide a further understanding of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0053] Figure 1 A schematic diagram of the flow chart in the embodiment of the present application is disclosed.

[0054] Figure 2 The present invention discloses a schematic structural diagram in an embodiment of the present application. DETAILED DESCRIPTION

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] The following is combined with Figure 1-Figure 2 The preferred embodiments of the present invention are described in detail.

[0057] Refer to the attached Figure 1 The present invention proposes a cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis, which includes the following steps:

[0058] S1. Acquire the continuous electromagnetic harmonic signal naturally generated during the operation of the cable, and discretize it at a preset sampling frequency to generate a discrete signal sequence;

[0059] S2, decompose the discrete signal sequence, separate the noise component and the effective harmonic component, and reconstruct the denoised signal sequence through threshold filtering;

[0060] S3, dividing the denoised signal sequence into a set of segmented signals according to a preset segment length, where each segment of the sub-signal corresponds to a local area of ​​the physical length of the cable;

[0061] S4, performing a joint analysis of the time domain and frequency domain on each sub-signal in the segmented signal set, extracting multi-dimensional feature parameters, and generating a global feature vector by segment combination;

[0062] S5. Based on the global feature vector, determine the degree of cable aging through multi-level thresholds, and trigger an alarm and positioning process according to the determination result;

[0063] S6. When it is determined that the cable is in a serious aging state, multiple monitoring nodes along the cable are activated to synchronously collect electromagnetic harmonic signals to generate a multi-node signal data set;

[0064] S7, analyzing the characteristic parameter attenuation law and phase propagation delay difference of the multi-node signal data set, calculating the distance range of the fault point relative to each monitoring node, and generating a candidate distance set;

[0065] S8. Perform weighted fusion on the candidate distance set to screen the optimal solution that meets the signal attenuation consistency, phase propagation delay matching, and energy distribution correlation, and output the final fault location result.

[0066] In one embodiment of the present invention, step S1 includes the following steps:

[0067] The electromagnetic harmonic sensor is used to capture the continuous electromagnetic harmonic signals naturally generated during cable operation in real time. The electromagnetic harmonic sensor discretizes the continuous electromagnetic harmonic signals at a preset sampling frequency to generate a discrete signal sequence.

[0068] Specifically, the electromagnetic harmonic sensor is fixedly installed at the head end of the cable or a preset monitoring point. The electromagnetic harmonic sensor has a built-in wide-band induction coil, which is used to couple the alternating electromagnetic field between the cable conductor and the shielding layer to generate a continuous electromagnetic harmonic signal corresponding to the operating status of the cable; then, the continuous electromagnetic harmonic signal is periodically sampled using an analog-to-digital conversion circuit. The sampling frequency is determined according to the rated voltage level of the cable and the highest harmonic frequency, ensuring that the discrete signal sequence after sampling completely retains the time-frequency characteristics of the original signal.

[0069] A continuous electromagnetic harmonic signal refers to the periodic electromagnetic field fluctuations generated in the surrounding space by current changes when the cable is energized. Its frequency includes the fundamental wave and integer multiple harmonic components. The sampling frequency refers to the set value for the number of signal points collected per second by the analog-to-digital conversion circuit. This value must satisfy the Nyquist sampling theorem and be at least twice the highest frequency component of the cable's electromagnetic harmonic signal. A discrete signal sequence is a collection of digital signals arranged in chronological order after sampling, with each number corresponding to the amplitude of the electromagnetic harmonic signal at the sampling moment.

[0070] For example, assume a CYZ-EMD-1000 electromagnetic harmonics sensor is installed at the headend of a 10kV cable, with a sampling frequency of 200MHz. The continuous electromagnetic harmonic signal captured by the sensor's induction coil is converted into a discrete signal sequence via a 24-bit high-precision analog-to-digital converter. Each data point in the sequence has an amplitude range of -5V to +5V, corresponding to the instantaneous electromagnetic field strength during cable operation. Technicians verify the integrity of the discrete signal sequence using an oscilloscope to ensure that the sampled discrete signal sequence fully preserves the time-frequency characteristics of the original signal.

[0071] In one embodiment of the present invention, step S2 includes the following steps:

[0072] Signal decomposition is performed on the discrete signal sequence to separate the noise component and the effective harmonic component; the noise component is eliminated by threshold filtering, and a denoised signal sequence is reconstructed, and the denoised signal sequence retains the electromagnetic harmonic characteristics related to the cable operation status.

[0073] Specifically, the discrete signal sequence input signal decomposition algorithm decomposes the signal into multiple sub-signal components according to different frequency bands, wherein the high-frequency sub-signal components contain random noise introduced by the external environment of the cable, and the low-frequency sub-signal components contain regular harmonics generated by the cable itself; for each sub-signal component, a dynamic threshold is set to determine the boundary between noise and effective harmonics, sub-signal components above the threshold are determined to be noise and set to zero, and sub-signal components below the threshold are retained as effective harmonics; all processed sub-signal components are reversely merged according to the original frequency band to generate a denoised signal sequence.

[0074] Among them, signal decomposition algorithms include but are not limited to wavelet transform and empirical mode decomposition, which are widely used in the field of signal processing; wavelet transform is a time-frequency analysis method that decomposes the signal into wavelet basis functions of different scales and frequencies, decomposes the signal and removes high-frequency noise components to reconstruct the denoised signal sequence; empirical mode decomposition is an adaptive signal decomposition method that decomposes a complex signal into several intrinsic mode functions, each of which represents an inherent vibration mode in the signal.

[0075] Reverse merging refers to the operation of recombining the filtered sub-signal components into a complete signal sequence according to the reverse process of decomposition, and reconstructing the denoised signal that retains the effective harmonic characteristics of the cable.

[0076] The dynamic threshold is a filter critical value that is adaptively adjusted based on noise energy. Its value is related to the statistical characteristics of the sub-signal components and is adaptively calculated based on the noise energy of each sub-signal component, satisfying the following formula:

[0077]

[0078] in, Indicates a dynamic threshold. represents the noise standard deviation, which is a statistic that measures the degree of data dispersion and reflects the degree of fluctuation of the noise signal. N represents the signal length, which is the number of data points contained in the signal.

[0079] In one embodiment of the present invention, step S3 includes the following steps:

[0080] According to the preset segment length, the denoised signal sequence is divided into several continuous sub-signals. Each sub-signal corresponds to a local area of ​​the cable's physical length, generating a segmented signal set to achieve refined analysis of the local area of ​​the cable status.

[0081] Specifically, the physical length corresponding to each sub-signal is determined based on the ratio of the total cable length to the preset segment length. The number of sampling points contained in each sub-signal is calculated in combination with the sampling frequency set in step S1. Based on the number of sampling points, continuous data segments are extracted from the denoised signal sequence in chronological order to generate a set of sub-signals corresponding to local regions of the cable's physical length. If the remaining signal length is less than a segment, it is expanded to a complete segment through zero padding. Zero padding refers to a processing method that pads the end of a signal with zero-valued data to meet the segment length requirement. Zero padding only pads the end of the signal with zero values. During feature extraction, valid data segments are automatically identified, eliminating interference from the zero-padding portion on the analysis.

[0082] Among them, the preset segment length refers to the total number of segments pre-set according to the cable length and monitoring accuracy requirements, and its value is proportional to the total cable length.

[0083] The number of sampling points refers to the amount of discrete signal data contained in each sub-signal segment. It is determined by the ratio of the cable physical segment length to the signal propagation speed and satisfies the following formula:

[0084]

[0085] Wherein, N is the number of sampling points of each sub-signal; The physical length of each cable segment is determined by the ratio of the total cable length to the preset segment length; is the sampling frequency, obtained in combination with step S1; It is the signal propagation speed, determined by combining the cable material properties.

[0086] For example, for a 100m long 10kV cable, the preset segment length is M = 10m, then the physical length of each segment is 10; the signal propagation speed is , the sampling frequency value is , substitute into the formula to calculate the number of sampling points of each sub-signal:

[0087]

[0088] It can be seen that a sub-signal is intercepted every 10 points in the denoised signal sequence, and a total of 10 sub-signals are generated.

[0089] In one embodiment of the present invention, step S4 includes the following steps:

[0090] Each sub-signal in the segmented signal set is jointly analyzed in the time domain and frequency domain to extract multi-dimensional feature parameters reflecting the cable operation status. The multi-dimensional feature parameters are combined by segment to generate a global feature vector.

[0091] Specifically, the following operations are performed on each sub-signal: The time-domain signal is converted to the frequency-domain signal via a fast Fourier transform. The ratio of the square of each harmonic amplitude to the sum of the squares of the total harmonic amplitude is calculated to generate a harmonic energy distribution signature. The degree of signal amplitude fluctuation in the time domain is statistically analyzed, and the signal fluctuation signature is generated by calculating the ratio of the maximum signal to the root mean square value. The phase differences of adjacent sub-signals at the same frequency are analyzed for consistency to generate a phase consistency signature.

[0092] Finally, the three types of features of each sub-signal are concatenated in segmented order to generate a global feature vector.

[0093] The harmonic energy distribution characteristic, which refers to the proportion of each harmonic energy to the total signal energy, is used to quantify the degree of cable insulation degradation. An imbalance in this proportion indicates partial discharge or carbonization of the insulation material. The signal fluctuation characteristic refers to the degree to which the time-domain signal amplitude deviates from the average level. An increase in this characteristic indicates conductor deformation due to mechanical stress or environmental corrosion.

[0094] Phase consistency compares the phase differences of adjacent sub-signals at the same frequency. If the difference exceeds a preset threshold, the signal is considered inconsistent. The threshold is calculated based on the maximum phase difference of the same type of cable in good condition.

[0095] In one embodiment of the present invention, step S5 includes the following steps:

[0096] Based on the generated global feature vector, the aging degree of the cable is judged through multi-level thresholds and the corresponding alarm and positioning process is triggered according to the judgment results.

[0097] Specifically, the value of each characteristic parameter in the global characteristic vector based on step S4 is sequentially compared with the preset safety threshold and fault threshold:

[0098] If the characteristic parameter does not exceed the preset safety threshold, the cable is judged to be in a healthy state; if the characteristic parameter exceeds the safety threshold but is lower than the fault threshold, the cable is judged to be in a slightly aged state and a warning signal is generated; if any characteristic parameter exceeds the fault threshold, the cable is judged to be in a severely aged state, a fault alarm signal is generated and the positioning process is activated.

[0099] The preset safety threshold corresponds to the normal aging boundary of the cable and is the maximum allowable value of the cable's characteristic parameters during normal operation. This threshold is derived from long-term monitoring data of the same type of cable under standard operating conditions. The fault threshold corresponds to the critical point of cable failure and is the minimum abnormal value of the characteristic parameters when the cable suffers irreversible damage. This threshold is calibrated using accelerated aging experiments and fault simulation data.

[0100] The early warning signal is a textual or visual reminder indicating that the cable is experiencing early signs of aging. The fault warning signal indicates a serious cable defect requiring immediate attention. The location process triggers the subsequent multi-node coordinated fault location process.

[0101] For example, the global feature vector generated in step S4 includes the harmonic energy distribution feature , signal fluctuation characteristics , phase consistency characteristics Spend;

[0102] Preset safety threshold: ≤15%, ≤3.5, ≤20 degrees;

[0103] Failure threshold: ≥25%, ≥5.0, ≥30 degrees;

[0104] because Exceeds the safety threshold but does not reach the failure threshold, The degree exceeds the safety threshold but does not reach the failure threshold, If the cable exceeds the safety threshold but does not reach the fault threshold, it is determined that the cable is in a mild aging state and an early warning signal is generated. 、 、 If any characteristic parameter exceeds the preset fault threshold, the cable is directly judged to be in a serious aging state, a fault alarm signal is generated, and the positioning process is activated.

[0105] In one embodiment of the present invention, step S6 includes the following steps:

[0106] When the cable is determined to be in a serious aging state, multiple monitoring nodes deployed along the cable are activated to synchronously collect electromagnetic harmonic signals from each monitoring node to generate a multi-node signal data set.

[0107] Specifically, when a severe aging alarm is triggered in step S5, an activation command is sent to multiple monitoring nodes deployed at the cable's headend, midsection, and end. Upon receiving the command, each monitoring node synchronizes its electromagnetic harmonics sensor based on a unified clock source and collects electromagnetic harmonic signals at its location in real time at the sampling frequency determined in step S1. The sampled data from all nodes is aligned by timestamp to form a multi-node signal dataset.

[0108] Monitoring nodes are hardware devices used for local signal acquisition. They integrate sensors, communication modules, and processing units and are fixedly installed at pre-set locations along the cable. A unified clock source refers to a timing system that synchronizes the sampling times of each monitoring node using GPS or the Network Time Protocol, ensuring time alignment of multi-node signals. A multi-node signal dataset represents a structured data set containing node numbers, physical locations, timestamps, and discrete signal sequences.

[0109] N monitoring nodes are deployed along the cable, and the acquisition time of each monitoring node is T seconds. The total data volume of the multi-node signal data set is the product of N times T times the sampling frequency.

[0110] For example, assume that 5 monitoring nodes are evenly spaced on a 100-meter-long cable (positions 0m, 25m, 50m, 75m, and 100m). When a severe aging alarm signal is triggered, each monitoring node starts collecting data at the exact second through the GPS synchronization module. Hz frequency continuously collects electromagnetic harmonic signals for 10 seconds, and the total data volume is .

[0111] After the multi-node signal data sets are aligned through timestamps, the differences in the signals of each monitoring node at the same time point can be compared, similar to multiple cameras shooting the same event from different angles and analyzing the object's motion trajectory through image synchronization.

[0112] In one embodiment of the present invention, step S7 includes the following steps:

[0113] Based on the multi-node signal data set generated in step S6, by analyzing the characteristic parameter attenuation law and phase propagation delay difference of the signal of each monitoring node, the distance range of the fault point relative to each monitoring node is calculated, and a candidate distance set for subsequent precise positioning is generated.

[0114] Specifically, for the electromagnetic harmonic signals collected by each monitoring node, the harmonic energy distribution characteristics, signal fluctuation characteristics, and phase consistency characteristics defined in step S4 are extracted. A characteristic parameter attenuation model is constructed based on the cable's distributed parameters to describe the attenuation of the characteristic parameters with cable transmission distance. A phase propagation model is also established to reflect the linear relationship between signal phase difference and cable transmission distance. The characteristic parameters of each monitoring node are input into this model, and the possible distance intervals of the fault point are calculated. A set of candidate distances that meet the multi-node consistency constraints is selected through cross-validation.

[0115] Among them, cross-validation refers to overlapping and comparing the distance intervals calculated by different nodes, and only retaining the distance values ​​that meet both the amplitude attenuation and phase delay constraints.

[0116] The candidate distance set refers to a set of discrete distance values ​​of all possible fault locations, where each distance value corresponds to a physical coordinate on the cable.

[0117] The characteristic parameter attenuation model is a mathematical expression of the variation of characteristic parameters with distance when electromagnetic harmonic signals propagate in a cable. Its characteristic parameters are calibrated by cable material properties and historical data and satisfy the following formula:

[0118]

[0119] in, represents the harmonic amplitude of node j; Indicates the harmonic amplitude at the cable head end; is the attenuation coefficient, which is calibrated by cable material parameters and experiments; is the distance from the fault point to node j. The farther the fault point is from the node, the more significant the amplitude attenuation.

[0120] The phase propagation model is a linear relationship between the signal phase difference and the transmission distance, which is determined by the electromagnetic wave propagation speed of the cable and satisfies the following formula:

[0121]

[0122] in, represents the phase difference of node j; f represents the signal frequency; represents the distance from the fault point to node j; Indicates the signal propagation speed.

[0123] For example, in the multi-node signal data set collected by the five monitoring nodes in step S6, the 3rd harmonic amplitude of node 1 (0 meters) is 5V, the 3rd harmonic amplitude of node 2 (25 meters) is 3.8V, and the 3rd harmonic amplitude of node 3 (50 meters) is 2.9V;

[0124] According to the characteristic parameter attenuation model fitting , when the distance between the fault point and node 3 is x=50, the theoretical amplitude should be 2.5V, and the actual measured value is =2.9V, indicating that the fault point is downstream of node 3;

[0125] Combined with the phase propagation model The phase difference calculation distance of node 4 (75 meters) is 72 meters, and that of node 5 (100 meters) is 98 meters. After cross-validation, the candidate distance set {70-75 meters, 95-100 meters} is screened out.

[0126] In one embodiment of the present invention, step S8 includes the following steps:

[0127] Based on the weighted fusion of the candidate distance set generated in step S7, the optimal solution that meets multiple constraints is screened out. The multiple constraints include signal attenuation consistency, phase propagation delay matching, and energy distribution correlation, and the final positioning result of the fault point relative to the cable head end is output.

[0128] Specifically, for each candidate distance in the candidate distance set, the degree of conformity with the signal attenuation consistency index, phase propagation delay matching index, and energy distribution correlation index of each monitoring node is calculated respectively; the weighted sum of the three types of indicator values ​​is weighted according to the set weight coefficients of each indicator to generate a comprehensive score; the candidate distance with the highest comprehensive score is selected as the optimal solution, and converted into the actual fault location according to the physical coordinates of the cable head end.

[0129] The signal attenuation consistency index measures the degree of match between the theoretical signal attenuation value and the actual measured value at the candidate distance. A smaller value indicates a more consistent attenuation pattern. The phase propagation delay matching index measures the deviation between the theoretical phase difference and the actual phase difference at the candidate distance. A smaller value indicates a more accurate delay model. The energy distribution correlation index measures the correlation between the energy signature at the candidate distance and the cable aging model. A larger value indicates a more significant fault signature. The weight coefficient is a weighting factor assigned based on the reliability of the indicator and determined through expert experience.

[0130] For example, the candidate distance set is {70 meters, 75 meters, 95 meters, 98 meters};

[0131] 1. Calculation of the 70-meter candidate distance:

[0132] Signal attenuation consistency index = 0.12 (theoretical attenuation 3.2V, measured 3.1V);

[0133] Phase delay matching index = 0.08 (theoretical phase difference 15 degrees, measured 14 degrees);

[0134] Energy correlation index = 0.85 (consistent with the characteristics of the aging model);

[0135] Determined through expert experience, the weight coefficient is: =0.3, =0.5, =0.2;

[0136] Comprehensive score = 0.12×0.3 + 0.08×0.5 + 0.85×0.2 = 0.23.

[0137] 2. Calculation of the 98-meter candidate distance:

[0138] Signal attenuation consistency index = 0.25; phase delay matching index = 0.05; energy correlation index = 0.90; determined through expert experience, weight coefficients: =0.3, =0.5, =0.2;

[0139] Comprehensive score = 0.25×0.3 + 0.05×0.5 + 0.90×0.2 = 0.265.

[0140] The comprehensive scores of the 70-meter candidate distance are compared with the comprehensive scores of the 98-meter candidate distance, and 98 meters is determined to be the optimal solution. The positioning result of the fault point being 98 meters away from the head end is output.

[0141] See attached Figure 2 The present invention also proposes a cable full life cycle status monitoring system based on electromagnetic harmonic characteristics analysis, which includes the following modules:

[0142] The signal acquisition module is used to obtain the continuous electromagnetic harmonic signal naturally generated during the operation of the cable, and discretize it at a preset sampling frequency to generate a discrete signal sequence;

[0143] The signal processing module decomposes the discrete signal sequence, separates the noise component and the effective harmonic component, and reconstructs the denoised signal sequence through threshold filtering;

[0144] The segment analysis module divides the denoised signal sequence into a set of segmented signals according to the preset segment length, where each segment of the signal corresponds to a local area of ​​the cable's physical length;

[0145] The feature extraction module performs a joint analysis of the time and frequency domains on each sub-signal in the segmented signal set, extracts multi-dimensional feature parameters, and generates a global feature vector based on the segment combination;

[0146] An aging determination module, which determines the degree of cable aging through multi-level thresholds based on the global feature vector and triggers an alarm and positioning process according to the determination result;

[0147] The multi-node collaborative module activates multiple monitoring nodes along the cable to synchronously collect electromagnetic harmonic signals and generate a multi-node signal data set when it determines that the cable is in a serious aging state;

[0148] A distance calculation module analyzes the attenuation law of characteristic parameters and the phase propagation delay difference of the multi-node signal data set, calculates the distance range of the fault point relative to each monitoring node, and generates a candidate distance set;

[0149] The positioning optimization module performs weighted fusion on the candidate distance set, selects the optimal solution that meets the signal attenuation consistency, phase propagation delay matching and energy distribution correlation, and outputs the final positioning result of the fault point.

[0150] It should be noted that the formulas described above, through the principle of dimensional consistency and mathematical standardization (e.g., normalization, dimensionless parameter conversion, or unified unit system), can translate physical quantities of different attributes into unitless standard values ​​or homogeneous, superimposable parameters. This eliminates the interference of different dimensions on operational logic, ensuring that the formulas retain the distribution characteristics of the original data while maintaining mathematical rationality and adaptability to objective laws. These are merely exemplary embodiments of the present invention and are not intended to limit the scope of the invention.

[0151] The modules can be implemented in whole or in part through software, hardware, or a combination thereof, supporting hardware embedded in or independent of a processor in a computer device, and also supporting software stored in a memory in a computer device, so that the processor can call and execute operations corresponding to the modules.

[0152] It should be noted that the human body information (including but not limited to human device information and personal information, etc.) and data (including but not limited to data used for analysis, stored data and displayed data, etc.) involved in the present invention are all information and data authorized by the human body or fully authorized by all parties. The collection, use and processing of relevant data require relevant legal standards.

[0153] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A cable life cycle condition monitoring method based on electromagnetic harmonic characteristics analysis, characterized in that: The following steps are involved: S1. Acquire the continuous electromagnetic harmonic signal naturally generated during the operation of the cable, and discretize it at a preset sampling frequency to generate a discrete signal sequence; S2, decompose the discrete signal sequence, separate the noise component and the effective harmonic component, and reconstruct the denoised signal sequence through threshold filtering; Discrete signal sequence input signal decomposition algorithm, decomposing the signal into multiple sub-signal components according to different frequency bands; A dynamic threshold is set to determine the boundary between noise and effective harmonics. Sub-signal components above the threshold are considered noise and set to zero, while sub-signal components below the threshold are retained as effective harmonics. The dynamic threshold is adaptively calculated based on the noise standard deviation of the sub-signal component and the signal length. All processed sub-signal components are combined inversely according to the original frequency band to generate a denoised signal sequence; S3, dividing the denoised signal sequence into a set of segmented signals according to a preset segment length, where each segment of the sub-signal corresponds to a local area of ​​the physical length of the cable; S4, performing a joint analysis of the time domain and frequency domain on each sub-signal in the segmented signal set, extracting multi-dimensional feature parameters, and generating a global feature vector by segment combination; Each sub-signal of the segmented signal set is converted from a time domain signal to a frequency domain signal through fast Fourier transform, and the ratio of the square of each harmonic amplitude to the sum of the square of the total harmonic amplitude is calculated to calculate the harmonic energy distribution characteristics; The fluctuation degree of the signal amplitude is counted in the time domain, and the signal fluctuation characteristics are generated by calculating the ratio of the maximum value of the signal to the root mean square value; Analyze the phase difference between adjacent sub-signals at the same frequency to generate phase consistency features; The three types of features of each segment signal are concatenated in segment order to generate a global feature vector; S5. Based on the global feature vector, determine the degree of cable aging through multi-level thresholds, and trigger an alarm and positioning process according to the determination result; S6. When it is determined that the cable is in a serious aging state, multiple monitoring nodes along the cable are activated to synchronously collect electromagnetic harmonic signals to generate a multi-node signal data set; S7, analyzing the characteristic parameter attenuation law and phase propagation delay difference of the multi-node signal data set, calculating the distance range of the fault point relative to each monitoring node, and generating a candidate distance set; S8. Perform weighted fusion on the candidate distance set to screen the optimal solution that meets the signal attenuation consistency, phase propagation delay matching, and energy distribution correlation, and output the final fault location result.

2. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: Generating a discrete signal sequence includes the following steps: The electromagnetic harmonic sensor is installed at the head end of the cable or a preset monitoring point. The electromagnetic harmonic sensor has a built-in broadband induction coil, which is used to couple the alternating electromagnetic field between the cable conductor and the shielding layer to generate a continuous electromagnetic harmonic signal. The continuous electromagnetic harmonic signal is periodically sampled by an analog-to-digital conversion circuit to generate the discrete signal sequence. The sampling frequency is determined according to the rated voltage level of the cable and the highest harmonic frequency.

3. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: Dividing the denoised signal sequence into a segmented signal set according to a preset segment length includes the following steps: The physical length is determined based on the ratio of the total cable length to the preset segment length; Calculate the number of sampling points contained in each sub-signal in combination with the preset sampling frequency; Based on the number of sampling points, continuous data segments are intercepted from the denoised signal sequence in time order to generate a segmented signal set corresponding to a local area of ​​the cable physical length.

4. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: Extracting continuous data segments from the denoised signal sequence in chronological order includes the following steps: If the remaining signal length is less than one segment, it is extended to a complete segment through zero padding. Zero padding refers to the processing method of filling zero-value data at the end of the signal to meet the segment length requirement. The zero padding operation only fills the end of the signal with zero values. During feature extraction, the valid data segment is automatically identified to eliminate the interference of the zero-padding part on the analysis.

5. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: The cable aging degree is judged by multi-level thresholds, and the alarm and positioning process is triggered according to the judgment results. The following steps are involved: The value of each characteristic parameter in the global characteristic vector is sequentially compared with the preset safety threshold and fault threshold; If the characteristic parameters do not exceed the preset safety threshold, the cable is determined to be in a healthy state; If the characteristic parameter exceeds the safety threshold but is lower than the fault threshold, the cable is judged to be in a mild aging state and a warning signal is generated; If any characteristic parameter exceeds the fault threshold, the cable is judged to be in a serious aging state, a fault alarm signal is generated, and the positioning process is activated.

6. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: Generating a multi-node signal dataset includes the following steps: When a severe aging alarm signal is triggered, multiple monitoring nodes deployed at the beginning, middle, and end of the cable send activation instructions; After receiving the instructions, each monitoring node is synchronously triggered by a unified clock source to collect electromagnetic harmonic signals, and a multi-node signal data set is generated by timestamp alignment; the monitoring nodes are deployed at preset positions at the beginning, middle and end of the cable.

7. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 1 is characterized in that: Generating a candidate distance set includes the following steps: For the electromagnetic harmonic signals collected by each monitoring node, the harmonic energy distribution characteristics, signal fluctuation characteristics, and phase consistency characteristics are extracted; Based on the characteristic parameter attenuation model and the phase propagation model, the fault distance interval corresponding to each monitoring node is calculated, and the overlapping intervals are screened through cross-validation to generate the candidate distance set; the attenuation model is an exponential decay function of the harmonic amplitude with distance, and the phase propagation model is a linear relationship between phase difference and distance.

8. The cable life cycle status monitoring method based on electromagnetic harmonic characteristics analysis according to claim 7 is characterized in that: Outputting the final fault location result includes the following steps: For each candidate distance in the candidate distance set, calculate its signal attenuation consistency index, phase propagation delay matching index, and energy distribution correlation index with each monitoring node; The three types of indicator values ​​are weighted and summed according to the set weight coefficients of each indicator to generate a comprehensive score; The candidate distance with the highest comprehensive score is selected as the optimal solution and converted into the actual fault location according to the physical coordinates of the cable head end.

Images