Power transmission line safety assessment method based on grounding electrode current

Abstract

The invention relates to the technical field of power transmission line safety monitoring, discloses a power transmission line safety assessment method based on grounding electrode current, and aims to solve the problem of low assessment reliability caused by the fact that the prior art is susceptible to far-field interference and environmental factors, and the method comprises the following steps: synchronously collecting instantaneous current and potential signals of a target and a reference monitoring position; calculating respective original impedance spectrums; performing spatial difference on the target original impedance spectrum and the average reference impedance spectrum of the reference position to obtain a differential impedance spectrum to suppress common mode influence; carrying out time difference on the differential impedance spectrum and a dynamic health baseline spectrum based on real-time environmental parameter prediction to obtain a diagnosis residual spectrum so as to remove environmental changes; and finally, performing safety evaluation and fault diagnosis based on the diagnosis residual spectrum. According to the method, local fault features can be accurately extracted through space-time dual difference, the accuracy and anti-interference capability of evaluation are improved, and quantitative evaluation and graded early warning of health states and identification of specific fault modes can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission line safety monitoring, in particular to a power transmission line safety evaluation method based on grounding electrode current. BACKGROUND

[0002] The grounding system of the power transmission line is the basis for the safe and stable operation of the power system. Its main function is to provide an effective discharge channel for current when lightning or short-circuit fault occurs, in order to protect equipment and personnel safety. Therefore, it is crucial to accurately and timely evaluate the health status of the grounding system.

[0003] At present, the state evaluation of the grounding system of the power transmission line usually relies on the measurement of its grounding impedance or related electrical parameters. However, in the actual operating environment, such measurement methods face multiple challenges, affecting the reliability of their evaluation results. On the one hand, the measurement signal is easily disturbed by the operating condition fluctuations of the power transmission line itself or the external far-field electromagnetic environment. These disturbances will be coupled into the measurement circuit, causing data distortion. On the other hand, the electrical characteristics of the grounding system itself are not constant, and they will normally fluctuate with seasonal or weather changes in environmental factors such as soil temperature and humidity. The existing evaluation methods are difficult to effectively distinguish between impedance changes caused by real faults (such as corrosion, fracture) and non-fault changes caused by the above disturbances and environmental factors, which often leads to false positives or false negatives of the evaluation system.

[0004] In addition, traditional evaluation methods often rely on comparing the measured value with a fixed threshold. This approach is not sensitive enough to identify the early stages of a slow developing degradation process, and can only detect abnormalities when the fault has become serious, lacking the ability to provide early warnings. Even if an anomaly is detected, existing technologies generally cannot provide information about the specific type of fault, and cannot distinguish between different situations such as grounding body corrosion or loose connecting bolts, resulting in a large amount of manual inspection required by maintenance personnel in subsequent fault troubleshooting, and low maintenance efficiency. SUMMARY

[0005] In view of the deficiencies of the prior art in the power transmission line grounding state evaluation method, such as single dimension of monitoring data, susceptibility to environmental factors and far-field electromagnetic interference, insufficient accuracy of evaluation results, and difficulty in distinguishing between fault and non-fault state changes, the present application provides a power transmission line safety evaluation method based on grounding electrode current, which solves the problem of low evaluation reliability and false positives or false negatives of traditional methods in complex operating environments.

[0006] To achieve the above purpose, the present application is implemented by the following technical solutions:

[0007] The present application provides a power transmission line safety evaluation method based on grounding electrode current, which adopts the following technical solutions:

[0008] A method for safety assessment of transmission lines based on grounding electrode current, comprising:

[0009] S1. Simultaneously acquire the instantaneous current signal of the grounding down conductor of the target monitoring location and at least one spatially adjacent reference monitoring location, and the instantaneous potential signal of the tower body to the ground to obtain the synchronously acquired signal;

[0010] S2. Based on the synchronously acquired signals, calculate the original impedance spectrum of the target monitoring location and the reference impedance spectrum of each reference monitoring location.

[0011] S3. Based on the reference impedance spectrum of each reference monitoring location, calculate the average reference impedance spectrum, and perform a differential operation between the original impedance spectrum of the target monitoring location and the average reference impedance spectrum to obtain the differential impedance spectrum.

[0012] S4. Based on the pre-constructed dynamic health baseline model associated with environmental parameters, predict the baseline differential impedance spectrum corresponding to the health status under the current environment, and subtract the differential impedance spectrum obtained in step S3 from the baseline differential impedance spectrum to obtain the diagnostic residual spectrum.

[0013] S5. Based on the diagnostic residual spectrum obtained in step S4, perform safety assessment and fault diagnosis.

[0014] By employing the above technical solution, this invention utilizes the spatiotemporal dual-difference mechanism to achieve accurate extraction of local fault characteristics in grounding systems. Its technical principle lies in:

[0015] Spatial differential mechanism: By synchronously acquiring signals from the target monitoring location and spatially adjacent reference monitoring locations (S1), and calculating the difference between the original impedance spectrum of the target location and the average reference impedance spectrum of the reference location (S3), common-mode effects can be effectively suppressed. Factors such as far-field electromagnetic interference and fluctuations in line operating conditions, which are common on transmission lines, have similar effects on multiple towers in geographical proximity, constituting common-mode components. The differential operation process cancels out these common-mode components, thereby highlighting local feature information that is only related to the state of the target monitoring location itself.

[0016] Time-difference mechanism: By introducing a dynamic health baseline model (S4) correlated with environmental parameters, the misjudgment problem caused by slow changes in grounding impedance due to non-fault factors such as soil temperature and humidity, and seasonal freeze-thaw cycles is solved. This model can predict the differential impedance spectrum shape that a healthy device should have under current environmental conditions. Subtracting the actual measured differential impedance spectrum from this dynamic baseline again, the resulting diagnostic residual spectrum is stripped of normal fluctuations caused by environmental factors. Therefore, the final diagnostic residual spectrum (S5) used for assessment is a high-purity fault characteristic quantity after dual noise filtering in both spatial and temporal dimensions. Its amplitude only deviates significantly from zero when the grounding system experiences structural degradation or sudden faults, thereby greatly improving the accuracy, sensitivity, and reliability of the assessment.

[0017] Preferably, the synchronous acquisition in step S1 is achieved by deploying synchronous timing units at both the target monitoring location and the reference monitoring location. The synchronous timing unit provides a unified high-precision time reference for the acquired instantaneous current signal of the grounding down conductor and the instantaneous potential signal of the tower body to the ground.

[0018] By employing the above technical solutions, deploying synchronization units (such as GPS or BDS modules) provides a time reference with microsecond-level or higher precision for monitoring terminals distributed in different geographical locations. This strict time alignment is a physical prerequisite for achieving effective spatial differential operations, ensuring that each spectral component compared in subsequent calculations originates from the response to system disturbances at the same moment, avoiding data misalignment caused by time asynchrony, and thus guaranteeing the effectiveness of common-mode suppression.

[0019] Preferably, after step S1 and before step S2, the method further includes: analyzing the instantaneous potential signal of the tower body to ground acquired in step S1, evaluating the excitation energy of the instantaneous potential signal of the tower body to ground at multiple frequency points, and obtaining an effective frequency set; step S2 is defined as: based on the synchronously acquired signal, calculating the original impedance spectrum of the target monitoring position and the reference impedance spectrum of each reference monitoring position at the frequency points within the effective frequency set.

[0020] By adopting the above technical solution, an evaluation step of the excitation signal energy is added before calculating the impedance spectrum. Since this method utilizes naturally occurring electrical disturbances in the system as the excitation source, their energy distribution is uneven across different frequency bands. This step, by setting an energy threshold, filters out frequency points with sufficient signal energy and high signal-to-noise ratio to form an effective frequency set. Subsequent impedance calculations are performed only at these effective frequency points, avoiding the risk of introducing huge calculation errors due to division operations at low signal-to-noise ratio frequency points, thus improving the quality and reliability of the original impedance spectrum data.

[0021] Preferably, in step S3, the average reference impedance spectrum is calculated by using the average reference spectrum formula to perform a complex arithmetic average or weighted average on the original impedance spectra of all reference monitoring locations within the reference tower set corresponding to the target monitoring location.

[0022] By employing the above technical solution, the impedance spectra at multiple reference locations are averaged to generate a more stable and representative common-mode interference reference. Compared to using a single reference point, this method can smooth out non-common-mode noise or minor perturbations present at individual reference points, avoid interference from abnormal states of individual reference points on the differential results, and enhance the robustness of spatial differential operations.

[0023] Preferably, in step S3, the differential impedance spectrum is calculated as follows: using the normalized differential impedance spectrum formula, the original impedance spectrum of the target monitoring location is subtracted from the average reference impedance spectrum by a complex number, and then divided by the average reference impedance spectrum.

[0024] By employing the above technical solution and performing normalization, the resulting differential impedance spectrum is a dimensionless relative rate of change. Its magnitude directly reflects the degree to which the target monitoring location deviates from the average state of the region, without being affected by the magnitude of the impedance spectrum baseline value itself. This makes it more convenient and effective to subsequently set uniform evaluation thresholds and conduct state comparisons.

[0025] Preferably, in step S4, the dynamic health baseline model is a function of frequency and environmental parameter vectors as inputs and expected health differential impedance spectrum under the environmental parameters as output. The environmental parameter vectors include at least one of soil temperature, soil moisture, air temperature, air humidity, and rainfall.

[0026] By adopting the above technical solution, the inputs and outputs of the dynamic health baseline model were clarified. Using physical environmental quantities that directly affect grounding resistivity (such as soil temperature and humidity) as model inputs gives the model interpretability based on physical mechanisms. By learning from historical health data, the model establishes a nonlinear mapping relationship between the electrical characteristics of the grounding system and environmental factors, enabling it to accurately predict and compensate for normal impedance fluctuations caused by environmental changes.

[0027] Preferably, step S5 specifically involves: performing aggregation calculations on the diagnostic residual spectrum obtained in step S4 to obtain a health index; and conducting safety assessment and fault diagnosis based on the health index.

[0028] By employing the above technical solution, the diagnostic residual spectrum distributed across multiple frequency points is aggregated into a single scalar index, namely the health index. This dimensionality reduction process transforms complex spectral information into an intuitive and quantitative assessment result, greatly facilitating automated status assessment, trend analysis, and the generation and management of early warning information.

[0029] Preferably, the safety assessment based on the health index specifically involves comparing the health index with a warning threshold and an alarm threshold, and determining the status as healthy, warning, or alarm based on the comparison result; the warning threshold and alarm threshold are determined based on statistical analysis of historical health data, or based on simulation or experimental data of typical faults.

[0030] By adopting the above technical solutions, a clear and hierarchical status assessment system was established. By setting two levels of thresholds—early warning and alarm—the entire process of fault monitoring, from early degradation to severe failure, was achieved. Maintenance personnel can adopt corresponding maintenance strategies based on different status levels; for example, they can focus on and schedule maintenance for "early warning" statuses, and respond immediately to alarm statuses, thereby improving the predictability and efficiency of maintenance work.

[0031] Preferably, the fault diagnosis is performed when the determination status is an alarm. The fault diagnosis includes: performing similarity matching between the diagnostic residual spectrum obtained in step S4 and multiple feature spectrum templates in a pre-built fault feature library to obtain a similarity score; and identifying the fault mode corresponding to the feature spectrum template with the highest similarity score as the current fault type based on the similarity score.

[0032] By adopting the above technical solution, a fault diagnosis process based on pattern matching is further introduced after an abnormal state is detected. This enables the method not only to determine whether a fault exists, but also to identify what kind of fault it is. By comparing the measured diagnostic residual spectrum with templates in the fault feature library, different fault modes such as grounding electrode corrosion, loose connecting bolts, and broken down conductors can be distinguished, providing clear guidance for subsequent accurate maintenance and shortening the fault diagnosis time.

[0033] Preferably, the fault feature library is constructed in at least one of the following ways: by accumulating diagnostic residual amplitude spectra recorded when faults occurred and were confirmed through on-site investigation in history, a feature spectrum template is formed; by establishing a three-dimensional simulation model of the tower-ground grid, specific physical defects are set in the three-dimensional simulation model of the tower-ground grid to simulate fault modes, and a feature spectrum template of the fault is generated through simulation calculation.

[0034] By adopting the above technical solutions, two complementary approaches to constructing a fault feature library are provided. Templates built using historical measured data have high realism, while templates generated using 3D simulation models can cover rare fault types that have never occurred historically but are theoretically possible. This combined approach ensures the completeness and practicality of the fault feature library, improving the coverage and accuracy of fault diagnosis.

[0035] This invention provides a method for safety assessment of transmission lines based on grounding electrode current. It has the following beneficial effects:

[0036] 1. This invention employs a mechanism combining spatial and temporal differential methods to improve the accuracy and anti-interference capability of the assessment results. Utilizing spatially adjacent monitoring locations as references for differential calculations effectively suppresses common-mode effects such as far-field electromagnetic interference and fluctuations in line operating conditions. Furthermore, by introducing a dynamic health baseline model correlated with environmental parameters for secondary differential calculations, normal fluctuations caused by non-fault factors such as soil temperature and humidity can be eliminated. This method ensures that the diagnostic characteristics ultimately used for assessment are only sensitive to local structural defects in the grounding system, reducing the misjudgment rate caused by changes in the environment and operating conditions.

[0037] 2. This invention aggregates complex spectrum diagnostic information into a quantitative health index and establishes two-level assessment thresholds for early warning and alarm, realizing quantitative assessment and graded early warning of the health status of the grounding system. This method solves the limitations of traditional methods that rely on qualitative judgment or single thresholds, and can identify early deterioration trends in equipment status. It provides a basis for decision-making for operation and maintenance departments to take preventive maintenance measures, thereby helping to avoid the sudden occurrence and expansion of faults and improving the level of precision in equipment asset management.

[0038] 3. After determining that the equipment alarm is detected, the present invention further realizes the diagnosis of specific fault types by pattern matching between the diagnostic residual spectrum and the fault feature library. This function enables the evaluation system to not only determine whether there is an anomaly in the grounding system, but also to identify specific fault modes such as grounding body corrosion and loose connectors. The diagnostic results provide clear guidance for on-site maintenance work, help to shorten the fault investigation time, reduce operation and maintenance costs, and improve the efficiency and accuracy of emergency repair work. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the architecture of a power transmission line safety assessment system according to an embodiment of the present invention;

[0040] Figure 2 This is a flowchart illustrating the overall workflow of a power transmission line safety assessment method according to an embodiment of the present invention.

[0041] Among them, 100 is the online monitoring terminal; 110 is the data acquisition module; 120 is the edge computing module; 130 is the communication module; 200 is the cloud analysis platform; 210 is the data receiving module; 220 is the central analysis module; 230 is the decision and output module; and 240 is the data storage module. Detailed Implementation

[0042] The technical solutions in 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] See attached document Figure 1 This invention provides a method for safety assessment of transmission lines based on grounding electrode current, which can be executed using an assessment system. This system may include online monitoring terminals deployed at multiple locations along the transmission line and a cloud-based analysis platform.

[0044] The online monitoring terminal 100 is deployed along the transmission line by towers, and its function is to collect and process on-site data. In one embodiment, the online monitoring terminal 100 includes: a data acquisition module 110 for acquiring the original electrical signals of the grounding system; an edge computing module 120 for performing preliminary processing and feature extraction on the original signals; and a communication module 130 for data interaction with a cloud analysis platform.

[0045] Specifically, the data acquisition module 110 may further include: a broadband current sensor 111 for measuring the instantaneous current flowing through the grounding down conductor; a tower potential reference sensor 112 for measuring the instantaneous potential of the tower relative to the ground; and a synchronization timing unit 113 for providing a high-precision time reference for the acquired current and potential data.

[0046] The cloud-based analysis platform 200 is used to receive and analyze data uploaded by one or more online monitoring terminals 100, execute complex diagnostic algorithms, and generate evaluation results. In one embodiment, the cloud-based analysis platform 200 includes: a data receiving module 210 for receiving data from each monitoring terminal; a central analysis module 220 for executing the core analysis and diagnostic logic of the method of the present invention; a decision and output module 230 for generating final evaluation conclusions and alarm information; and a data storage module 240 for storing historical data, models, and fault feature libraries.

[0047] After the data acquisition module 110 of the online monitoring terminal 100 collects data, the edge computing module 120 processes the data and sends the processed results to the cloud analysis platform 200 via the communication module 130. The data receiving module 210 of the cloud analysis platform 200 receives the data and passes it to the central analysis module 220 and the decision and output module 230 for in-depth analysis and decision-making, ultimately outputting the evaluation results. The data storage module 240 provides data support for the entire analysis process.

[0048] See attached document Figure 2 This invention provides a method for safety assessment of transmission lines based on grounding electrode current, which may include the following steps:

[0049] S201, synchronously acquire the instantaneous current signal of the grounding down conductor and the instantaneous potential signal of the tower body to the ground at the target monitoring location and at least one spatially adjacent reference monitoring location;

[0050] S202, Analyze the instantaneous ground potential signal of the tower body at the target monitoring location, evaluate its excitation energy at multiple frequency points, and determine the effective frequency set for subsequent calculations;

[0051] S203, based on the synchronously acquired signal, calculates the original impedance spectrum of the target monitoring position and the reference impedance spectrum of each reference monitoring position at frequency points within the effective frequency set.

[0052] S204. Based on the reference impedance spectrum, calculate an average reference impedance spectrum, and perform a difference operation between the original impedance spectrum and the average reference impedance spectrum to obtain the differential impedance spectrum.

[0053] S205, based on a pre-built dynamic health baseline model associated with environmental parameters, predicts the baseline differential impedance spectrum corresponding to the health status under the current environment, and subtracts the actual calculated differential impedance spectrum from the baseline differential impedance spectrum to obtain the diagnostic residual spectrum.

[0054] S206, based on the diagnostic residual spectrum, performs safety assessment and fault diagnosis, including calculating a quantitative health index and performing pattern matching with a fault feature library when the health index is abnormal to identify the fault type.

[0055] To make the technical solutions of the embodiments of the present invention clearer, the following will describe in detail each step of the aforementioned method flow.

[0056] In one embodiment, the execution of step S201 is described in detail. The goal of this step is to acquire high-fidelity electrical signals that are strictly aligned on the time base for subsequent system state identification.

[0057] This step can be further broken down into S201.1, signal acquisition, and S201.2, time synchronization.

[0058] In step S201.1, signal acquisition is completed by the data acquisition module 110 within the online monitoring terminal 100. To acquire the instantaneous current signal of the grounding lead, a broadband current sensor 111 can be used. The broadband current sensor 111 has a sufficiently wide measurement bandwidth, for example, covering the DC to megahertz (DC-MHz) range, to ensure that it can not only accurately measure the power frequency component but also capture high-frequency signal components generated by system disturbances, harmonics, or transient events. These high-frequency signal components can serve as the excitation source required for subsequent analysis in this invention. Specific implementations may include, but are not limited to, using a Rogowski coil or a high-precision closed-loop Hall effect current sensor.

[0059] Meanwhile, to obtain the instantaneous potential signal of the tower relative to the ground, a tower potential reference sensor 112 can be used. This sensor is used to establish a stable voltage measurement reference point. Specifically, it can be a non-contact electric field sensor that obtains the potential signal by measuring the electric field strength around the tower, which is proportional to the tower's potential; or a sensor based on the principle of capacitive voltage division, which constructs a high-resistance voltage divider circuit with a fixed voltage division ratio between the tower and the ground to obtain a voltage signal proportional to the tower's potential.

[0060] In step S201.2, time synchronization is achieved by the synchronization and timing unit 113 within the data acquisition module 110. This unit provides a unified, high-precision time reference for online monitoring terminals 100 deployed in different geographical locations. Specifically, this can be achieved by integrating a Global Positioning System (GPS) module or a BeiDou Navigation Satellite System (BDS) module.

[0061] The synchronization unit 113 can output a pulse-per-second (PPS) signal and a message containing standard time information. The controller of the online monitoring terminal 100, such as a field-programmable gate array (FPGA) or microcontroller (MCU), uses the rising or falling edge of the PPS signal as a hardware trigger signal to simultaneously initiate sampling of the analog-to-digital converters (ADCs) for both the current and potential channels. This method ensures that the sampling times of the current and potential signals are strictly aligned within a single monitoring terminal.

[0062] Simultaneously, the controller parses the time message and adds a precise timestamp to each acquired data block or data frame. Since the synchronization units 113 of all monitoring terminals along the entire line receive time signals from the same satellite system, the timestamps assigned to their respective data by different monitoring terminals have a unified reference, achieving synchronization accuracy at the microsecond or sub-microsecond level. This time synchronization across multiple spatially distributed monitoring points is a necessary prerequisite for performing multi-point spatiotemporal differential calculations in the subsequent step S204 to effectively separate common-mode interference and local fault characteristics. The specific circuit implementation of the time synchronization technology can be designed by those skilled in the art based on publicly available standards; it is a well-known technology in the field and will not be elaborated upon here.

[0063] In one embodiment, the execution method of step S202 is described in detail. This step is performed by the edge computing module 120 within the online monitoring terminal 100, and its purpose is to perform a quality assessment on the natural electrical signal used as the excitation source to ensure that the signal used for subsequent calculations has a sufficient signal-to-noise ratio, thereby ensuring the accuracy of the impedance spectrum calculation results.

[0064] This step can be broken down into S202.1, which is to calculate the excitation energy spectrum, and S202.2, which is to determine the effective frequency set.

[0065] In step S202.1, the edge computing module 120 processes the instantaneous potential signal of the tower body relative to the ground acquired in step S201. Time-frequency analysis is performed. One specific implementation method is to use Short-Time Fourier Transform (STFT), dividing the signal into multiple time windows and performing Fourier Transform on the data within each time window to obtain the frequency domain representation of the signal. For the implementation of STFT, those skilled in the art can use window functions such as Hanning windows or Hamming windows to suppress spectral leakage, which are well-known techniques in the field and will not be elaborated here.

[0066] After obtaining the frequency domain representation of the signal, its energy spectrum is calculated using the excitation energy spectrum formula, which is:

[0067] ;

[0068] in, Indicates that in the number At the tower location, within the time window Within, the frequency is The excitation energy spectrum of the potential signal; It is a time-domain signal In the time window The result of the short-time Fourier transform within is a complex number; It is the modulo operator, therefore express The amplitude, i.e., the amplitude spectrum of the signal; This indicates that the amplitude is squared, and the result is the energy spectrum of the signal.

[0069] In step S202.2, the edge computing module 120 calculates the excitation energy spectrum obtained in the previous step. With a preset effective excitation threshold spectrum Compare the data to filter out the effective frequency points.

[0070] Effective excitation threshold spectrum It is defined as a frequency-dependent function whose value corresponds to a specific frequency point. The threshold spectrum is the signal energy required to achieve the preset minimum signal-to-noise ratio (SNR). Since the background noise power spectral density of transmission lines varies across different frequency bands, this threshold spectrum is not a single fixed value. It can be obtained through statistical analysis of noise levels from historical monitoring data or through on-site calibration.

[0071] By analyzing each frequency point A point-by-point comparison is performed to determine the effective frequency set. This process is described using the definition of the effective frequency set, which is:

[0072] ;

[0073] in, Indicates that in the number At the tower location, within the time window Within, the set of frequency points that are determined to be valid; The standard definition of a set is the frequency of a set satisfying a specific condition. A set; Yes, the filtering condition means that this frequency point is selected only when the excitation energy spectrum of the actual signal is not less than the preset threshold spectrum. Only then were they included in the effective frequency set.

[0074] The final set of effective frequencies This will be used to guide the impedance spectrum calculation in the subsequent step S203, meaning that the impedance value will be calculated only at frequency points belonging to this set.

[0075] To proactively acquire broadband excitation signals that are difficult to obtain under normal operating conditions, the method of this invention also includes an opportunistic probing mechanism. This mechanism utilizes predictable power grid events that generate strong transient electrical signals to acquire high-quality data. This mechanism involves collaborative operation between the cloud analytics platform 200 and the online monitoring terminal 100.

[0076] The implementation of this mechanism may include S202.3, in which the cloud platform predicts the detection timing and issues instructions, and S202.4, in which the monitoring terminal executes the detection task.

[0077] In step S202.3, the central analysis module 220 of the cloud analysis platform 200 is responsible for identifying time windows that can be used for opportunistic detection. One specific implementation involves the cloud analysis platform 200 connecting to the power grid dispatch automation system to acquire and parse predetermined operation plans, thereby identifying events such as the closing or opening of transmission lines, the switching of large inductive or capacitive loads, and the operation of parallel capacitor banks or reactor banks. These events generate transient components spanning a wide frequency band in the power grid, which can serve as effective broadband excitation sources. Another implementation involves the central analysis module 220 performing data mining on historical monitoring data, using pattern recognition algorithms to identify waveform characteristics corresponding to unplanned but periodic or regular transient disturbances, thereby predicting their future occurrence time.

[0078] After identifying the detection opportunity, the decision and output module 230 of the cloud analysis platform 200 generates and issues opportunistic detection commands to one or more target online monitoring terminals 100. The commands may include the following information: a unique identifier for the target monitoring terminal, the start and end times of the detection task, and a specific set of data acquisition parameters. The data acquisition parameters may include a transient sampling rate higher than the conventional acquisition rate, and triggering conditions for transient capture.

[0079] In step S202.4, the communication module 130 of the target online monitoring terminal 100 receives and parses the opportunistic detection command. The edge computing module 120 adjusts the operating mode of the data acquisition module 110 according to the command content within a specified start and end time period.

[0080] Specifically, the edge computing module 120 temporarily increases the sampling rate of the analog-to-digital converter (ADC) in the data acquisition module 110 from a normal value (e.g., 100 kS / s) to a transient sampling rate (e.g., 1 MS / s or higher) set in the instruction.

[0081] Simultaneously, the edge computing module 120 activates transient capture mode. One implementation method involves using a circular buffer to continuously cache high-sampling-rate data. The module monitors the amplitude or rate of change of the acquired signal in real time. Once the monitored value exceeds the trigger condition set in the instruction, the module will save and process the data stored in the circular buffer for a period of time before and after the trigger moment (i.e., pre-trigger and post-trigger) as a complete transient event waveform.

[0082] By implementing an opportunistic detection mechanism, this invention can actively acquire excitation and response data with high signal-to-noise ratio and wide bandwidth that are difficult to obtain under steady-state conditions, thereby providing data support for constructing a more complete and accurate broadband impedance spectrum.

[0083] In one embodiment, the execution method of step S203 is described in detail. This step is performed by the edge computing module 120 within the online monitoring terminal 100. Its purpose is to calculate the original impedance spectrum characterizing the frequency response characteristics of the grounding system using the electrical signal from the quality assessment in step S202.

[0084] This step can be broken down into S203.1, which is frequency domain transformation, and S203.2, which is complex impedance calculation.

[0085] In step S203.1, the edge computing module 120 matches the time window in step S202. Corresponding, synchronously acquired tower body to ground instantaneous potential signal and instantaneous current signal of grounding lead The data segments are then subjected to frequency domain transformation. One specific implementation method is to use the Fast Fourier Transform (FFT) algorithm. This transformation converts the discrete time-domain signal sequence into a discrete frequency-domain complex sequence, yielding the frequency domain representation. and The specific implementation of the Fast Fourier Transform is a well-known technique in this field and will not be elaborated upon here.

[0086] In step S203.2, the edge computing module 120 calculates the complex impedance of the grounding system based on the frequency domain signal obtained in the previous step. This calculation is not performed at all frequency points, but only for the effective frequency set determined in step S202. Frequency points within This selective calculation avoids performing division operations at frequencies where the excitation signal is insufficient, thereby improving the reliability of the calculation results.

[0087] The impedance spectrum is calculated using the complex impedance spectrum formula, which is:

[0088] ;

[0089] in, This represents the complex impedance value at frequency f at tower number k. This value is a complex number that contains the impedance amplitude and phase information of the grounding system at that frequency. It is the frequency domain representation of the tower's potential signal to ground; It is the frequency domain representation of the grounding lead current signal; This represents a complex number division operation, that is, the frequency domain representation of a potential signal divided by the frequency domain representation of a current signal; This is the execution condition of the formula, explicitly stating that this calculation only applies to frequencies belonging to the effective frequency set. frequency points efficient.

[0090] By repeating this calculation across all frequency points within the effective frequency set, a dataset consisting of complex impedance values ​​at multiple discrete frequency points is obtained. This dataset represents the original impedance spectrum of the target monitoring location. The same steps are used to calculate the reference impedance spectrum for each tower selected as a reference monitoring location. These calculated impedance spectrum data will be uploaded to the cloud analysis platform 200 for subsequent differential analysis.

[0091] In one embodiment, the execution method of step S204 is described in detail. This step is executed by the central analysis module 220 within the cloud analysis platform 200. Its purpose is to process the original impedance spectrum of a single monitoring point by introducing spatial dimension reference information, so as to suppress or eliminate the common-mode effects introduced by factors such as far-field interference and changes in the overall operating conditions of the line, thereby highlighting the local characteristics caused by changes in the state of the grounding system itself.

[0092] This step begins in S204.1 by defining and selecting a reference tower cluster for each target monitoring location to be evaluated (hereinafter referred to as the target tower).

[0093] In step S204.1, the central analysis module 220 determines the corresponding set of reference towers for the target tower numbered k. .

[0094] One implementation is that the selection is statically configured based on the physical topology of the line. For example, for the towers on a single-circuit line. Its reference tower cluster It can be fixedly defined as two towers that are physically directly adjacent to each other, that is... .

[0095] Another implementation is that the selection is dynamic. Before each calculation, the central analysis module 220 filters from a broader pool of candidate towers according to preset rules. These preset rules may include: electrical connection and environmental similarity rules, meaning candidate towers should be located on the same power line as the target tower, or within the same geographically susceptible area; and data validity rules, meaning only towers selected within the current analysis time window are considered. Within the range, the uploaded impedance spectrum data must be complete and have passed quality verification for the towers; or a geographical proximity rule must be used, i.e., all towers with similar data to the target tower must be selected. Towers whose straight-line distance is less than a preset threshold.

[0096] After determining the reference tower cluster Then, step S204.2 generates the average reference spectrum.

[0097] In step S204.2, the central analysis module 220 analyzes the reference tower cluster. All members within the same time window The original impedance spectrum is calculated to generate an average reference impedance that can represent the common-mode effect in this region. .

[0098] In one specific implementation, this operation is the arithmetic mean of complex vectors. The spectrum is calculated using the average reference spectrum formula, which is:

[0099] ;

[0100] in, Represented as target tower Calculated at frequency The average reference complex impedance on; It is the target tower's number; These are frequency points, and the calculation is performed point-by-point on all valid frequency points involved in the analysis. In step S204.1, the target tower is... A defined set of reference towers; It is a set The number of reference towers, i.e. ; It is an iterable collection The index represents the number of a reference tower; It is numbered The original complex impedance of the reference tower is calculated within the same time window; Represents a set The original impedance spectra of all reference towers are summed using complex numbers; The overall representation is the complex arithmetic mean of the impedance spectra of all reference towers within the cluster.

[0101] In another preferred embodiment, to further improve the representativeness of the reference spectrum, a weighted average method can be used to generate it. The formula for the weighted average reference spectrum is:

[0102] ;

[0103] in, To be assigned to reference towers The weighting coefficients. The methods for determining the weighting coefficients may include: weights based on geographical distance, i.e., weights relative to the reference tower. To the target tower The distance is inversely proportional; or it is based on signal quality weights, i.e., the weights are related to the reference tower. The signal-to-noise ratio or the proportion of effective frequency points of the signal within the current time window is directly proportional.

[0104] This averaging operation can smooth out or suppress non-common-mode noise or disturbances that exist only at individual reference points while preserving the spectral components common to each reference point, thus obtaining a more robust common-mode interference reference.

[0105] In step S204.3, the central analysis module 220 performs differential calculations to quantify the difference in the impedance spectrum of the target tower relative to the reference standard, thereby suppressing common-mode effects. This calculation requires an effective frequency range agreed upon by all participants. Therefore, a shared set of effective frequencies needs to be determined first. This collection contains the target towers. and its reference cluster The intersection of the effective frequency sets of all towers within the area. The common effective frequency set formula is used for calculation, which is:

[0106] ;

[0107] in, It is the target tower The set of common effective frequencies determined by the difference calculation; The target tower Its own effective frequency set; The target tower Reference tower cluster; It is a cluster A reference tower The effective frequency set; Indicates the cluster The intersection operation is performed on the effective frequency sets of all reference towers within the range, and the result is the set of effective frequency points common to all reference towers. This is the intersection operator for sets.

[0108] After determining the common effective frequency set Subsequently, in one implementation, the differential impedance spectrum is calculated using complex subtraction. The formula for the differential impedance spectrum is:

[0109] ;

[0110] in, Indicates the target tower In frequency The differential impedance spectrum value is a complex number; The target tower The original complex impedance; It is the target tower The calculated average reference complex impedance; representing complex number subtraction; This is the execution condition of the formula, explicitly stating that this calculation applies only to frequency points belonging to the common effective frequency set. efficient.

[0111] In another preferred embodiment, to obtain a measure of the relative change in the reference value, normalized difference or complex division can be used. For example, the normalized differential impedance spectroscopy formula can be used for calculation, as follows:

[0112] ;

[0113] The difference spectrum obtained by this normalization process is dimensionless, which better highlights the relative rate of change and has advantages for setting a unified fault discrimination threshold.

[0114] The differential impedance spectrum obtained through this differential operation This reduces the impact of common-mode factors such as fluctuations in the overall operating conditions of the line and far-field electromagnetic interference, making the target towers more stable. The differences in the grounding system's condition relative to the average condition of its neighboring area are highlighted. Therefore, this differential impedance spectrum is a more sensitive diagnostic measure for local fault characteristics and can be used for subsequent health status assessment.

[0115] In one embodiment, the execution of step S205 is described in detail. This step is performed by the central analysis module 220 within the cloud analysis platform 200. Its purpose is to compare the differential impedance spectrum calculated in real time with a benchmark that reflects the health status of the equipment, thereby generating a diagnostic residual spectrum sensitive to fault characteristics.

[0116] The key point here is that the benchmark is not a fixed, static spectrum, but a dynamic health baseline model. This step can be broken down into S205.1, namely the construction and application of the dynamic health baseline model.

[0117] In step S205.1, the central analysis module 220 first performs analysis on each target tower. A proprietary dynamic health baseline model is constructed. This model aims to learn and predict differential impedance spectroscopy under healthy conditions. An inherent mapping relationship between the baseline and a set of measurable environmental or operational parameters. Introducing environmental or operational parameters allows the baseline to adapt to slow changes in the electrical characteristics of the grounding system caused by non-fault factors such as soil temperature, humidity, and seasonal freeze-thaw cycles, avoiding misjudgments due to these normal variations.

[0118] The model building process, i.e. the training phase, can be implemented as follows:

[0119] The central analysis module 220 collects and stores data on the target towers. Historical differential impedance spectroscopy data for a specific period The equipment is considered to be in a healthy state during specific periods, such as the initial commissioning phase or after regular inspections / maintenance confirming the grounding system is intact. Simultaneously, environmental parameters related to the tower are collected, forming a multi-dimensional parameter vector. This vector may include, but is not limited to, soil temperature, soil moisture, air temperature, air humidity, and rainfall. These parameters can be obtained through dedicated sensors deployed locally on the tower or by accessing third-party meteorological data services.

[0120] The dynamic health baseline model is constructed as a function. Its input is frequency and environmental parameter vector The output is the expected health differential impedance spectrum under these environmental conditions. This is described using the dynamic health baseline model formula, which is:

[0121] ;

[0122] in, It is the target tower predicted by the model under environmental condition E. In frequency The complex impedance value at the healthy baseline; This represents the functional relationship characterized by the model; For frequency; A multidimensional parameter vector describing the current environmental state, for example =[temperature, humidity, ...].

[0123] One specific implementation method is that the model Multiple regression analysis can be used to construct and identify... The linear or polynomial relationship between the real and imaginary parts of the equation and various environmental parameters.

[0124] In another preferred embodiment, to capture more complex nonlinear relationships, the model F can be constructed using machine learning algorithms, such as feedforward neural networks, support vector regression, or Gaussian process regression. By training these algorithms using historical health data, a model capable of predicting a health baseline based on environmental parameters can be obtained. The training and implementation of such machine learning algorithms can be designed by those skilled in the art based on publicly available theories, and are well-known techniques in the field, so they will not be elaborated upon here.

[0125] During the model application phase, the central analysis module 220 acquires the current real-time environmental parameter vector. And input it into the already trained model. In this way, the target tower can be calculated at the current moment. The expected health differential impedance spectrum, i.e., the dynamic health baseline. .

[0126] To ensure the long-term effectiveness of the model, the central analysis module 220 can also periodically use new data that has been confirmed to be in a healthy state to improve the model. Perform online updates or retraining to enable it to adapt to the long-term, slowly changing characteristics of the device and environment.

[0127] In step S205.2, after acquiring the dynamic health baseline at the current moment, the central analysis module 220 compares it with the measured differential impedance spectrum to calculate the diagnostic residual spectrum. The purpose of this calculation is to quantify the degree to which the current equipment state deviates from its expected health state.

[0128] The calculation determines the common effective frequency set in step S204.3. The process is performed point by point. In one specific implementation, the diagnostic residual spectrum formula is used for calculation. The diagnostic residual spectrum formula is as follows:

[0129] ;

[0130] in, The target tower In frequency The diagnostic residual spectrum value is a complex number. At the current moment Calculated target tower Measured differential impedance spectrum; In step S205.1, based on the environmental parameters at the current moment... The expected health differential impedance spectrum predicted by the dynamic health baseline model; This is the execution condition of the formula, explicitly stating that this calculation applies only to frequency points belonging to the common effective frequency set. efficient.

[0131] The obtained diagnostic residual spectrum It is a complex quantity. Under healthy conditions, the measured value... Compared with the baseline value predicted by the model The difference is small, therefore The amplitude is close to zero. When the grounding system deteriorates or fails, the measured value will deviate from the healthy baseline, leading to... The amplitude increases from a level close to zero.

[0132] To facilitate the setting of standardized assessment thresholds and the performance of status assessments, it is typically necessary to convert the complex diagnostic residual spectrum into a scalar spectrum. One specific implementation method is to calculate its amplitude spectrum. This is done using the diagnostic residual amplitude spectrum formula, which is:

[0133] ;

[0134] in, The target tower In frequency The diagnostic residual amplitude spectrum is a non-negative real number. The modulo operator for complex numbers indicates the calculation of complex numbers. The amplitude.

[0135] In another preferred embodiment, to obtain a relative deviation that is insensitive to the amplitude of the baseline itself, a normalized diagnostic residual can be calculated. The formula for the normalized diagnostic residual amplitude spectrum is:

[0136] ;

[0137] This type of normalized residual is dimensionless, which better reflects the relative severity of the bias.

[0138] Regardless of the method used to calculate the diagnostic residual amplitude spectrum These are all diagnostic characteristic quantities that are sensitive to local faults in the grounding system and have been compensated for changes in operating conditions, far-field interference, and environmental factors. This spectrum will be used for health status assessment and fault diagnosis in step S206.

[0139] In yet another preferred embodiment, to quantify the effective value or average power of the residual spectrum, its root mean square (RMS) value can be calculated as a health index. The formula for the RMS health index is:

[0140] ;

[0141] in, For the common effective frequency set The total number of mid-frequency points.

[0142] In step S206.2, the decision and output module 230 will calculate the health index. The health status of the target tower grounding system is determined by comparing it with a set of preset evaluation thresholds.

[0143] In one embodiment, a warning threshold can be set. and an alarm threshold (in The evaluation rules are as follows:

[0144] like If so, the system status is determined to be "healthy";

[0145] like If the system status is "warning", it indicates that there is initial deterioration and corresponding maintenance suggestions can be generated.

[0146] like If the system status is "alarm", it indicates a serious fault and an alarm message can be sent to the maintenance personnel immediately.

[0147] Evaluation threshold ( The determination of a warning threshold may include: statistical analysis based on historical health data, for example, setting a warning threshold. Set to healthy state The mean of the historical data distribution plus three standard deviations ( ), set alarm threshold Set as the mean plus six standard deviations ( Alternatively, based on simulation or experimental data of typical faults, that is, by injecting known fault modes into the model, calculating the resulting health index, and setting a threshold accordingly.

[0148] In step S206.2, when the health index If the warning or alarm threshold is exceeded, the decision and output module 230 can further execute a fault diagnosis process to identify the specific fault mode. This process is based on a pre-built fault feature library.

[0149] This step can be broken down into S206.3, namely, the construction of the fault feature library. The decision and output module 230 or an offline training module is responsible for this task. The fault feature library is a database that stores various known fault modes and their corresponding diagnostic feature spectra.

[0150] Specifically, the library can be constructed as a collection containing multiple tuples. Its structure is , among which, among which This represents the total number of known fault modes as preset. For any tuple in this set... , where index From 1 to Integers:

[0151] It is the first Labels for different failure modes, such as "severe corrosion of the grounding electrode", "loose grounding lead bolts", or "broken grounding lead".

[0152] It is related to failure mode The associated feature spectrum template, its data structure and diagnostic residual amplitude spectrum Consistent. In one implementation, This could be a typical spectrum from a single measurement under this failure mode; in another preferred embodiment, to improve the robustness of the template, It can be a statistical spectrum, such as the mean spectrum obtained by averaging or median filtering multiple sample spectra of the failure mode.

[0153] Characteristic Spectrum Template The acquisition methods may include one or more of the following: One implementation method is through the accumulation of historical data. That is, when a fault confirmed through on-site investigation occurs in the past, At that time, the diagnostic residual amplitude spectrum recorded before its occurrence was used as a template. Stored in the database.

[0154] Another approach is through electromagnetic field simulation. This involves creating an accurate three-dimensional simulation model of the tower-ground grid and simulating a failure mode by introducing specific physical defects within the model. (For example, modifying the material conductivity of the grounding electrode to simulate corrosion, or setting a small air gap at the connection point to simulate bolt loosening), then obtaining the impedance spectrum under this fault through simulation calculation, and processing it through the same S204 and S205 steps as the actual measurement, finally generating the characteristic spectrum template of this fault. .

[0155] After the fault feature database is built, step S206.4 performs real-time fault mode matching and identification. When the target tower... Health Index When the alarm threshold is exceeded, the decision and output module 230 outputs its current diagnostic residual amplitude spectrum. With fault feature library Each feature spectrum template in Perform similarity matching.

[0156] In one specific implementation, the matching is achieved by calculating a similarity score. For example, the similarity is calculated using a cosine similarity formula, which is:

[0157] ;

[0158] in, Indicates the current measured spectrum With templates in the library The similarity score has a range of [-1, 1]; and The measured spectrum and the template spectrum at frequencies are respectively The amplitude on; This represents the dot product of two spectral vectors; Represents the measured spectral vector The Euclidean norm. Besides cosine similarity, other measures that can quantify the similarity or distance between two spectra, such as the Pearson correlation coefficient or the reciprocal of the Euclidean distance, can also be applied to this invention.

[0159] The decision and output module 230 iterates through all templates in the library. A series of similarity scores are calculated. The final identified fault modes... The label is determined by the template with the highest similarity score. To avoid incorrectly classifying unknown faults into existing patterns, a matching confidence threshold can also be set. The final recognition rule is:

[0160] ;

[0161] That is, only when the largest similarity score is greater than or equal to the confidence threshold. Only then will the fault be determined as the corresponding mode. Otherwise, classify it as an "unknown fault" and indicate that further manual investigation is required.

[0162] In another preferred embodiment, fault identification can be achieved using a pre-trained classification model. For example, a support vector machine (SVM), decision tree, or neural network classifier can be trained using data from a fault feature library as the training set. During real-time diagnosis, the current diagnostic residual amplitude spectrum is used. As input, the classifier directly outputs the label of the most likely fault mode. The training and application of such classifiers can be carried out by those skilled in the art, and are well-known technologies in the field, so they will not be described in detail here.

[0163] Ultimately, the decision and output module 230 will identify the specific fault modes. The alarm information is output together to provide maintenance personnel with more accurate fault location and repair guidance.

Claims

1. A method for transmission line safety assessment based on ground electrode current, characterized in that, The method comprises the following steps: S1, synchronously collecting the instantaneous current signals of the grounding down conductor and the instantaneous potential signals of the tower body at the target monitoring position and at least one reference monitoring position adjacent to the target monitoring position in space, to obtain the synchronously collected signals; S2, based on the synchronously collected signals, calculating the original impedance spectrum of the target monitoring position and the reference impedance spectrum of each reference monitoring position; S3, based on the reference impedance spectrum of each reference monitoring position, calculating the average reference impedance spectrum, and performing difference operation on the original impedance spectrum of the target monitoring position and the average reference impedance spectrum to obtain the difference impedance spectrum; S4, according to a pre-constructed dynamic health baseline model associated with environmental parameters, predicting the baseline difference impedance spectrum corresponding to the health state under the current environment, and subtracting the difference impedance spectrum obtained in the S3 step from the baseline difference impedance spectrum to obtain a diagnostic residual spectrum; S5, based on the diagnostic residual spectrum obtained in the S4 step, performing safety evaluation and fault diagnosis.

2. The ground electrode current-based transmission line safety assessment method of claim 1, wherein, The synchronous collection in the S1 step is realized by deploying a synchronous time unit at the target monitoring position and the reference monitoring position, which provides a unified high-precision time reference for the collected instantaneous current signals of the grounding down conductor and the instantaneous potential signals of the tower body.

3. The ground electrode current based transmission line safety assessment method of claim 1, wherein, After the S1 step and before the S2 step, the method further comprises the following steps: analyzing the instantaneous potential signals of the tower body collected in the S1 step, evaluating the excitation energy of the instantaneous potential signals of the tower body at multiple frequency points, and obtaining an effective frequency set; The S2 step is limited to: based on the synchronously collected signals, calculating the original impedance spectrum of the target monitoring position and the reference impedance spectrum of each reference monitoring position at the frequency points in the effective frequency set.

4. The ground electrode current based transmission line safety assessment method of claim 3, wherein, In the S3 step, the average reference impedance spectrum is calculated by using an average reference spectrum formula to perform complex arithmetic average or weighted average on the original impedance spectra of all reference monitoring positions in the reference tower set corresponding to the target monitoring position.

5. The ground electrode current based transmission line safety assessment method of claim 1, wherein, In the S3 step, the difference impedance spectrum is calculated by using a normalized difference impedance spectrum formula to perform complex subtraction operation on the original impedance spectrum of the target monitoring position and the average reference impedance spectrum, and then dividing the average reference impedance spectrum.

6. The ground electrode current based transmission line safety assessment method of claim 1, wherein, In the S4 step, the dynamic health baseline model is a function with frequency and environmental parameter vector as input and the expected health difference impedance spectrum under the environmental parameters as output, and the environmental parameter vector includes at least one of soil temperature, soil humidity, air temperature, air humidity and rainfall.

7. The ground electrode current based transmission line safety assessment method of claim 1, wherein, The S5 step specifically comprises: performing aggregation operation on the diagnostic residual spectrum obtained in the S4 step to obtain a health index; based on the health index, performing safety evaluation and fault diagnosis.

8. The ground electrode current based transmission line safety assessment method of claim 7, wherein, The safety evaluation based on the health index specifically comprises: comparing the health index with a warning threshold and an alarm threshold; when the health index is less than the warning threshold, the state is determined to be healthy; when the health index is greater than or equal to the warning threshold and less than the alarm threshold, the state is determined to be in warning. determining that the state is alarm when the health index is greater than or equal to the alarm threshold; wherein the pre-warning threshold and the alarm threshold are determined based on statistical analysis of historical health data, or based on simulation or experimental data of typical faults.

9. The ground electrode current based transmission line safety assessment method of claim 8, wherein, The fault diagnosis is performed when the state is determined to be alarm, and the fault diagnosis comprises: performing similarity matching between the diagnostic residual amplitude spectrum obtained in the S4 step and a plurality of feature spectrum templates in a pre-constructed fault feature library, to obtain a similarity score; According to the similarity score, the fault mode corresponding to the feature spectrum template with the maximum similarity score is identified as the current fault type.

10. The ground electrode current based transmission line safety assessment method of claim 9, wherein, The construction method of the fault feature library comprises at least one of the following: forming feature spectrum templates by accumulating diagnostic residual amplitude spectra recorded when faults occurred in history and confirmed by on-site investigation; establishing a tower-ground net three-dimensional simulation model, setting specific physical defects in the tower-ground net three-dimensional simulation model to simulate fault modes, and generating feature spectrum templates of faults through simulation calculation.

Images