Hardware-in-the-loop test method and device for power distribution network traveling wave fault detection system

Abstract

The application provides a hardware-in-the-loop test method and device for a power distribution network traveling wave fault detection system, relates to the technical field of power distribution network fault detection and simulation test, and comprises the following steps: obtaining pre-constructed hierarchical reference data, the hierarchical reference data comprising first-level reference data adapted to initial positioning requirements and second-level reference data adapted to accurate positioning requirements; based on the access characteristics of a terminal to be tested, controlling the terminal to be tested to upload differentiated test data to first-level and second-level master stations in stages; and under a target test mode, based on the hierarchical reference data, in combination with the uploaded data of the terminal to be tested and the research and judgment results of the two-level master stations, performing full-link performance quantitative evaluation on the power distribution network traveling wave fault detection system. The application realizes accurate matching of test reference and two-zone master station hierarchical research and judgment mechanisms and full-link performance quantitative verification.

Description

Technical Field

[0001] This invention relates to the field of power distribution network fault detection and simulation testing technology, specifically to a hardware-in-the-loop testing method and apparatus for a power distribution network traveling wave fault detection system. Background Technology

[0002] Currently, accurate fault location in distribution networks heavily relies on the wavefront identification performance of traveling wave terminals and the collaborative analysis capabilities of distribution automation master stations. Existing hardware-in-the-loop testing (HIL) techniques for traveling wave terminals typically employ general-purpose testing logic, using the overall fault analysis time or wave peak time of the master station as a unified verification benchmark. This approach lacks customization for the specific operating architecture and hierarchical analysis mechanisms of distribution networks. This general testing method suffers from issues such as benchmark timing lag, inability to distinguish between genuine first wavefronts and noisy pseudo-wavefronts, and limited test object coverage, making it difficult to adapt to the differentiated characteristics of retrofittable FTU terminals, existing FTU terminals, and fault indication terminals in terms of dual-zone master station access partitions, data upload logic, and analysis permissions. Furthermore, existing testing dimensions are one-sided, focusing only on wavefront timing accuracy and neglecting key indicators such as wavefront quality screening, hierarchical data upload timing compliance, and consistency of analysis between dual-zone master stations. It also lacks adaptability to complex and abnormal operating conditions such as BeiDou time synchronization interruptions, resulting in significant deviations between test results and actual operating conditions. This fails to meet the requirements for refined testing and standardized network access verification of distribution network traveling wave terminals and master station systems. Summary of the Invention

[0003] To address the issues of existing technologies, such as coarse general traveling wave test benchmarks, weak scenario adaptability, and inability to adapt to the dual-zone master station architecture and differentiated access of various types of terminals in distribution networks, this application achieves high-precision quantitative evaluation and error tracing of the entire link performance of traveling wave fault detection in distribution networks through a hardware-in-the-loop testing method and device for a traveling wave fault detection system in distribution networks.

[0004] To achieve the above objectives, this application adopts the following technical solution: A hardware-in-the-loop testing method for a power distribution network traveling wave fault detection system includes: acquiring pre-constructed hierarchical reference data, wherein the hierarchical reference data includes first-level reference data adapted to initial positioning requirements and second-level reference data adapted to fine positioning requirements; based on the access characteristics of the terminal under test (DUT), controlling the DUT to upload differentiated test data hierarchically to the first-level master station and the second-level master station; the test data is generated hierarchically by the DUT based on the collected traveling wave signal; in a target test mode, based on the hierarchical reference data, combined with the uploaded data of the DUT and the judgment results of the two-level master stations, performing a full-link performance quantitative evaluation of the power distribution network traveling wave fault detection system; wherein the target test mode includes an absolute time test mode based on a common time reference or a relative time test mode based on the relative timing of multiple terminals.

[0005] Preferably, the hierarchical reference data is generated by locking the physical initial moment of the traveling wave based on a multi-constraint joint identification mechanism. The multi-constraint joint identification mechanism includes: temporal gradient mutation constraint, used to screen mutation candidate points based on steady-state noise reference; waveform kurtosis disturbance resistance constraint, used to filter out pseudo mutation points based on local waveform kurtosis; baseline step offset constraint, used to verify persistent step mutation based on the baseline difference before and after mutation; based on the mutation points verified by the multi-constraint joint identification mechanism, the physical initial moment of the traveling wave is locked, and the first-level reference data and the second-level reference data are generated accordingly.

[0006] Preferably, the execution process of the multi-constraint joint identification mechanism includes: preprocessing the acquired original traveling wave time series, extracting the steady-state waveform data before the fault to calculate the mean steady-state baseline, and establishing a fault-free reference level; calculating the first-order instantaneous change gradient of the waveform point by point based on the system sampling interval, and quantifying the severity of waveform change at each time series node; statistically analyzing the maximum gradient value within the steady-state interval before the fault as the system noise fluctuation reference, constructing an adaptive change threshold, and determining that point as a candidate change point when the gradient of a sampling point is greater than or equal to the adaptive change threshold; and processing the change... The ratio of the fourth central moment to the variance of the instantaneous signal amplitude after removing the mean is calculated from the waveform data within the short-time window near the candidate point to obtain the local waveform kurtosis. Only candidate points whose local waveform kurtosis is greater than a preset kurtosis threshold are retained. The mean of the short-time steady-state waveform after the mutation occurs is statistically analyzed, and the baseline difference before and after the mutation is calculated. When the baseline difference before and after the mutation is greater than or equal to a preset offset threshold, the mutation is determined to be a fault-induced continuous step mutation. All the mutation points that pass the verification are traversed, and the earliest mutation point in the time dimension is selected as the physical initial time of the traveling wave.

[0007] Preferably, the step of controlling the terminal under test to upload differentiated test data to the first-level master station and the second-level master station based on the access characteristics of the terminal under test includes: dividing the test data into lightweight feature data and complete waveform data according to the data upload permissions of the terminal under test; uploading the lightweight feature data to the first-level master station according to a first timing requirement to trigger the first-level master station to perform initial fault location; and uploading the complete waveform data to the second-level master station according to a second timing requirement to trigger the second-level master station to perform fine fault location and secondary verification; wherein the response time limit of the first timing requirement is shorter than the response time limit of the second timing requirement.

[0008] Preferably, before uploading differentiated test data from the terminal under test to the first-level master station and the second-level master station, the method further includes: scoring the quality of the waveforms collected by the terminal under test based on the signal-to-noise ratio, amplitude ratio, and rise edge steepness; filtering invalid waveforms based on the quality score and verifying the compliance of the caching and uploading of valid waveforms; and using the data corresponding to the valid waveforms that pass the verification as the data source for the lightweight feature data and the complete waveform data.

[0009] Preferably, in the target test mode, based on the hierarchical benchmark data, combined with the uploaded data of the terminal under test and the judgment results of the two-level master station, a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system is performed. This includes: in the absolute time test mode, based on the same absolute time benchmark of the signal generator and the terminal under test, linking the hierarchical judgment mechanism of the first-level master station and the second-level master station, synchronously verifying the terminal wavefront identification accuracy, hierarchical upload compliance, fault algorithm matching degree, and dual-zone judgment consistency; or, in the relative time test mode, based on the relative timing of the multi-terminal cluster, verifying the consistency of parameters of the same model of terminal and the compatibility of terminals of different brands or types, to evaluate the system compatibility in multi-terminal mixed access scenarios.

[0010] Preferably, in the relative time test mode, based on the relative timing of a multi-terminal cluster, the consistency verification of parameters for terminals of the same model and the compatibility verification of terminals of different brands or types are performed, including: synchronously connecting multiple brands and types of terminals under test to the hardware-in-the-loop test loop, uniformly receiving standardized fault waveform signals; each terminal under test independently completing wavefront identification, quality screening, and graded data upload, generating its own relative judgment timing and test data; comparing the wavefront timing, quality score, and upload timing deviation of terminals of the same model to assess the stability of equipment hardware and algorithms; quantifying the timing deviation and data interaction compatibility of terminals of different brands or types in mixed access scenarios; and verifying the stability of the initial positioning and fine positioning results of the dual-level master station based on the differentiated data uploaded by multiple terminals.

[0011] Preferably, the full-link performance quantitative evaluation of the distribution network traveling wave fault detection system further includes: distinguishing between terminal wavefront identification error, quality screening misjudgment error, hierarchical upload timing offset error, fault algorithm matching deviation, master station data interaction delay error, and terminal clock drift error; generating a comprehensive performance quantitative score for terminal network access based on preset thresholds in the hierarchical benchmark data and the distinguished types of errors; determining the compliance of terminal network access based on the comprehensive performance quantitative score, and generating differentiated access adaptation suggestions based on the terminal type.

[0012] Furthermore, this application also provides a hardware-in-the-loop testing device for a distribution network traveling wave fault detection system, applicable to the hardware-in-the-loop testing method for the distribution network traveling wave fault detection system as described above, comprising: a data acquisition module for acquiring pre-constructed hierarchical reference data, wherein the hierarchical reference data includes first-level reference data adapted to initial positioning requirements and second-level reference data adapted to fine positioning requirements; a data uploading module for uploading test data differentially to the first-level master station and the second-level master station based on the access characteristics of the terminal under test, wherein the test data is generated hierarchically by the terminal under test based on the collected traveling wave signal; and a performance evaluation module for performing a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system based on the uploaded data of the terminal under test and the judgment results of the two-level master stations under a target test mode; wherein the target test mode includes an absolute time test mode based on a common time reference or a relative time test mode based on the relative timing of multiple terminals.

[0013] Furthermore, this application also provides a hardware-in-the-loop testing system for traveling wave fault detection in a distribution network, comprising: a hardware-in-the-loop signal generator, a cluster of terminals under test, a first-level master station, and a second-level master station; the hardware-in-the-loop signal generator is configured to output a standardized fault waveform signal; the cluster of terminals under test is configured to acquire traveling wave signals and generate test data in stages; the first-level master station and the second-level master station are configured to receive the test data and perform hierarchical fault analysis; the system is configured to execute the hardware-in-the-loop testing method for the traveling wave fault detection system in a distribution network as described above.

[0014] Beneficial effects: This application addresses the problem of existing general testing methods having a single benchmark and being unable to adapt to the dual-zone master station architecture in power distribution networks. It acquires hierarchical benchmark data adapted to the initial and fine-tuning requirements of the fault location, and controls the hierarchical uploading of differentiated test data to the dual-level master station based on the access characteristics of the terminal under test. Simultaneously, it performs full-link performance quantification evaluation under absolute or relative time testing modes. Through the synergy of hierarchical benchmark data and the hierarchical uploading mechanism, this solution achieves precise matching between the test benchmark and the hierarchical judgment mechanism of rapid initial location in Zone I and fine-tuning verification in Zone IV of the power distribution network. This ensures comprehensive and objective quantitative verification of the end-to-end performance of the traveling wave fault detection system under different testing scenarios.

[0015] This application utilizes a multi-constraint joint identification mechanism to lock the physical initial moment of the traveling wave and generate hierarchical benchmark data. By employing triple constraints of temporal gradient abrupt change, waveform kurtosis immunity, and baseline step offset, it effectively avoids the timing lag and false wavefront false detection problems caused by traditional methods that rely solely on the peak or maximum gradient moment. This mechanism reconstructs the true moment of the first effective abrupt change of the traveling wave from a physical mechanism perspective, providing a sub-microsecond high-precision unified timing origin for hardware-in-the-loop testing, significantly improving the reliability and reproducibility of the test benchmark.

[0016] This application realistically simulates the business logic of prioritizing feature uploads and delaying waveform uploads in a power distribution network by dividing lightweight feature data and complete waveform data according to terminal data upload permissions and uploading them to the first-level and second-level master stations respectively according to different response time limits. Combined with pre-processing quality screening and verification based on indicators such as signal-to-noise ratio and amplitude ratio, this not only verifies the terminal's data quality control capabilities but also ensures full coverage of the compliance of the hierarchical data transmission protocol in the testing process, filling the gap in traditional testing regarding data interaction timing verification.

[0017] This application constructs a dual-mode adaptive testing system using both absolute and relative time. This system enables high-precision dual-zone linkage verification in a laboratory environment based on a common time reference, while also allowing for compatibility and consistency evaluation through multi-terminal relative time sequence benchmarking in outdoor conditions without time synchronization. This dual-mode design overcomes the limitations of ideal operating condition testing, enabling it to adapt to complex and abnormal scenarios such as BeiDou signal interruptions. It provides comprehensive technical support for standardized network access calibration and operational performance evaluation of multi-brand traveling wave terminals across the entire domain. Attached Figure Description

[0018] Figure 1 This is a flowchart of the hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing particular embodiments only and is not intended to be construed as such. This is to limit the scope of this application.

[0021] Example 1 like Figure 1 As shown, this embodiment provides a hardware-in-the-loop testing method for a distribution network traveling wave fault detection system. This method is executed by a hardware-in-the-loop testing system or a server connected to it, and aims to perform comprehensive performance verification of the distribution network traveling wave fault detection system. Specifically, the method includes the following steps: Step S1: Obtain pre-built hierarchical benchmark data, which includes first-level benchmark data adapted to initial positioning requirements and second-level benchmark data adapted to fine positioning requirements.

[0022] In this embodiment, the hierarchical benchmark data is not generated in real time at the test site, but is a standardized dataset pre-constructed and stored in a database based on typical fault conditions and traveling wave propagation physical mechanisms of the distribution network. At the start of the test, the system directly calls or reads this dataset as the unified truth scale for this test. The first-level benchmark data mainly includes lightweight feature parameters such as traveling wavefront timescale, amplitude, and polarity, specifically used to support the performance verification of the first-level master station for rapid initial fault location. The second-level benchmark data includes complete waveform sampling sequences, fault feature templates, and refined timing thresholds, used to support the second-level master station for in-depth verification of precise fault location and secondary verification. Through this hierarchical design, the test benchmark can accurately match the differentiated requirements for data granularity and response speed at different levels under the dual-level master station architecture of the distribution network, avoiding the problem that traditional single benchmarks cannot simultaneously address speed and accuracy.

[0023] Step S2: Based on the access characteristics of the terminal under test, control the terminal under test to upload differentiated test data to the first-level master station and the second-level master station in a hierarchical manner; the test data is generated by the terminal under test in a hierarchical manner based on the collected traveling wave signal.

[0024] Specifically, after receiving the simulated traveling wave signal output by the hardware-in-the-loop signal generator, the terminal under test (DUT) will autonomously or in a controlled manner process the acquired raw traveling wave signal based on its own hardware capabilities, communication protocols, and access characteristics such as registration permissions in the master station system, generating test data at different levels. For example, for terminals with real-time interactive capabilities, lightweight data containing key wavefront features and detailed data containing complete waveform recordings can be generated. Subsequently, according to a preset hierarchical transmission strategy, the system prioritizes uploading the lightweight feature data to the first-level master station to trigger the initial positioning process, while simultaneously uploading the complete waveform data to the second-level master station in a specific time sequence to support fine positioning analysis. This process realistically simulates the "feature-first, waveform recording supplementary transmission" business logic of the distribution network, ensuring the consistency between the test environment and actual operating conditions.

[0025] Step S3: In the target test mode, based on the hierarchical benchmark data, combined with the uploaded data of the terminal under test and the judgment results of the two-level master station, the full-link performance of the distribution network traveling wave fault detection system is quantitatively evaluated.

[0026] In this step, the system uses pre-acquired hierarchical benchmark data as the standard truth value and performs multi-dimensional cross-comparison with the test data actually reported by the terminal under test, the initial positioning result of the first-level master station, and the fine positioning result of the second-level master station. The evaluation scope covers the entire business chain from terminal signal acquisition, wavefront identification, hierarchical data uploading, to master station reception, algorithm analysis, and result output. By calculating the deviation of each link from the benchmark data, key indicators such as terminal identification accuracy, data transmission compliance, and master station collaborative analysis capabilities can be accurately quantified, thereby achieving closed-loop verification of the system's end-to-end performance.

[0027] The target test modes include an absolute time test mode based on the same time reference or a relative time test mode based on the relative timing of multiple terminals.

[0028] These two modes are parallel optional technical solutions, which can be flexibly switched according to the actual testing environment and conditions. The absolute time testing mode is suitable for scenarios with high-precision time synchronization conditions, such as laboratories, and uses the same clock source as the signal source and the terminal to achieve microsecond-level absolute timing verification. The relative time testing mode is suitable for scenarios without an absolute time reference, such as outdoor field, and evaluates the consistency and compatibility of the system by comparing the relative timing relationships between multiple terminals. This dual-mode design effectively solves the limitations of traditional testing methods that rely on ideal time synchronization conditions and cannot adapt to complex field conditions, significantly improving the engineering practicality and scenario coverage of the testing solution.

[0029] Example 2 In this embodiment, the hierarchical reference data is generated by locking the physical initial moment of the traveling wave based on a multi-constraint joint identification mechanism. This mechanism aims to solve the timing lag and false wavefront detection problems caused by traditional testing methods that rely solely on the peak moment or the maximum gradient moment as the reference. By restoring the true physical abrupt change point of the traveling wave signal at the moment of fault occurrence, it provides a high-precision unified timing origin for hardware-in-the-loop testing. Specifically, the multi-constraint joint identification mechanism includes time-domain gradient abrupt change constraint, waveform kurtosis immunity constraint, and baseline step offset constraint. These three constraints are not simply superimposed, but rather a progressive screening logic built based on the physical propagation characteristics of the traveling wave signal in the distribution network: the time-domain gradient abrupt change constraint is used to quickly screen potential abrupt change candidate points based on the steady-state noise reference, solving the initial judgment of whether there is an abrupt change; the waveform kurtosis immunity constraint is used to filter out false abrupt change points caused by white noise, switching operations, etc., based on the local waveform kurtosis, solving the nature identification of true and false abrupt changes; the baseline step offset constraint is used to verify the continuous step abrupt change based on the baseline difference before and after the abrupt change, solving the status confirmation of transient interference and continuous faults. Finally, based on the mutation points verified through a multi-constraint joint identification mechanism, the physical initial moment of the traveling wave is locked, and first-level and second-level reference data are generated accordingly. This generation method ensures that the reference data has both sub-microsecond time accuracy and complete physical characteristics suitable for the analysis needs of the two-level master station.

[0030] To more clearly illustrate the implementation process of the above mechanism, this embodiment further provides the specific execution steps of the multi-constraint joint identification mechanism. First, the acquired raw traveling wave time series is preprocessed, and the steady-state waveform data before the fault is extracted to calculate the mean steady-state baseline, establishing a fault-free reference level. This step eliminates the interference of DC components or low-frequency drift during normal line operation on subsequent abrupt change detection, providing a stable reference zero point for subsequent differential calculations. For example, waveform data within 2 milliseconds before the fault trigger signal can be selected as the steady-state interval, and the reference level can be calculated after smoothing high-frequency noise using a moving average filtering algorithm.

[0031] Next, based on the system sampling interval, the first-order instantaneous abrupt change gradient of the waveform is calculated point by point to quantify the severity of waveform abrupt changes at each time-series node. Specifically, this calculation process involves: obtaining the amplitude of the traveling wave signal at the current sampling point and the amplitude of the traveling wave signal at the previous sampling point, calculating the amplitude difference between the two, and then dividing this amplitude difference by the system sampling interval to obtain the first-order instantaneous abrupt change gradient characterizing the waveform's rate of change. Compared to directly using the absolute value of the amplitude, gradient calculation can more sensitively capture the steep rising edge characteristics of the arriving traveling wavefront, effectively amplifying the difference between fault characteristics and background noise even in weak fault scenarios such as low-resistance grounding.

[0032] Subsequently, the maximum gradient value within the steady-state interval before the fault is used as the benchmark for system noise fluctuation, and an adaptive mutation threshold is constructed. When the gradient at a sampling point is greater than or equal to the adaptive mutation threshold, that point is determined to be a candidate mutation point. An adaptive threshold is used instead of a fixed threshold to adapt to the complex and variable operating environment of the distribution network. For example, the maximum gradient value within the steady-state interval can be multiplied by a preset amplification factor to serve as the adaptive mutation threshold. When the ambient noise is high, the maximum steady-state gradient value increases accordingly, and the threshold is automatically adjusted upwards to avoid false triggering; conversely, in a quiet environment, the threshold is automatically adjusted downwards to ensure sensitivity. This dynamic adjustment mechanism significantly improves the robustness of the algorithm under different signal-to-noise ratio conditions.

[0033] After identifying candidate mutation points, the ratio of the fourth-order central moment to the variance of the instantaneous signal amplitude (after removing the mean) is calculated for the waveform data within a short-time window adjacent to each candidate point. This yields the local waveform kurtosis, and only candidate points with a local waveform kurtosis greater than a preset kurtosis threshold are retained. Statistically, the kurtosis value of conventional white noise or sinusoidal steady-state signals typically approaches 3, while the kurtosis value of a true fault traveling wavefront is an impulsive pulse, significantly higher than 3. Therefore, by calculating this mathematical characteristic of the ratio of the fourth-order central moment to the variance, it is possible to effectively distinguish fault wavefronts with impulsive characteristics from random noise or steady-state oscillations without such characteristics. The preset kurtosis threshold can be set based on the statistical distribution of historical measured data, for example, set to 5, thereby accurately filtering out pseudo-mutation points that, although having large gradients, lack impulsive characteristics.

[0034] Furthermore, the mean of the short-time steady-state waveform after the sudden change is statistically analyzed, and the baseline difference before and after the change is calculated. When the baseline difference before and after the change is greater than or equal to a preset offset threshold, the change is determined to be a fault-induced sustained step change. This step utilizes another important physical characteristic of fault traveling waves: after a fault occurs, the line voltage or current usually jumps from one steady-state level to another new steady-state level and remains there for a period of time, rather than recovering instantaneously like lightning interference or switching glitches. By comparing the mean difference between the two steady-state intervals before and after the change, instantaneous spike interference can be completely eliminated, ensuring that the locked moment corresponds to the true fault initiation state.

[0035] Finally, all abrupt change points that pass the triple constraint verification are iterated through, and the earliest abrupt change point in the time-series dimension is selected as the physical initial moment of the traveling wave. The reason for choosing the earliest point rather than the largest or steepest point is that the traveling wave undergoes refraction and reflection during propagation, and the subsequent wavefronts are often superimposed waveforms after line impedance matching or node reflection. Only the first abrupt change point directly corresponds to the physical origin of the fault. Based on this precisely locked physical initial moment of the traveling wave, the system can generate first-level reference data containing parameters such as precise time scale, amplitude, and polarity, as well as second-level reference data containing a complete waveform sequence and fault feature templates. This reference generation method based on physical mechanisms fundamentally ensures the accuracy of the true value of subsequent full-link performance quantitative evaluation and avoids test result distortion caused by the deviation of the reference itself.

[0036] Example 3 In this embodiment, to ensure that the data uploaded to the master station has effective testing value and truly simulates the edge computing capabilities of the distribution network terminal, the method also includes a pre-processing data quality control step before uploading the test data to the first-level master station and the second-level master station in a differentiated manner. Specifically, this step scores the waveforms acquired by the terminal under test based on the signal-to-noise ratio (SNR), amplitude ratio, and rise edge steepness. This scoring mechanism is not a simple threshold comparison, but rather simulates the preprocessing logic of the original signal by a real terminal in a resource-constrained environment. For example, the SNR can be calculated by extracting the ratio of the noise power of the steady-state window before the fault to the effective power of the signal window after the fault is triggered, which is used to characterize the degree to which the signal is submerged by noise; the amplitude ratio can be defined as the multiple relationship between the peak value of the traveling wavefront and the amplitude of the background noise baseline fluctuation, which is used to judge the significance of the fault characteristics; the rise edge steepness reflects the transient severity when the wavefront arrives, and is usually quantified by calculating the ratio of the amplitude difference between the wavefront start point and the peak point to the corresponding time interval. These three dimensions construct a comprehensive quality evaluation system from the aspects of signal purity, feature strength, and physical authenticity, which can effectively identify invalid data that, although triggered by waveform recording, are actually caused by interference.

[0037] Based on the aforementioned quality scores, the system further performs filtering and verification operations: filtering invalid waveforms based on the quality scores and verifying the compliance of valid waveform cache and upload. Specifically, when the overall quality score of a waveform segment is lower than a preset qualified threshold, it is determined to be an invalid waveform and discarded to avoid consuming valuable communication bandwidth and main station storage resources; for waveforms with scores higher than the threshold, they are marked as valid waveforms, and their integrity in the terminal's local cache and whether they conform to predefined message format specifications are further checked. Only the data corresponding to valid waveforms that pass the verification will be used as the data source for subsequent lightweight feature data and complete waveform data. This pre-screening mechanism not only verifies the performance of the data quality control algorithm of the terminal under test, but also ensures that the data entering the transmission stage are all high-confidence samples, thereby guaranteeing the objectivity and accuracy of the subsequent end-to-end performance evaluation results.

[0038] After completing the preliminary quality screening, this embodiment further describes the differentiated data interaction process based on the access characteristics of the terminal under test. Based on the access characteristics of the terminal under test, the test data is uploaded to the first-level master station and the second-level master station in a differentiated manner. These access characteristics include, but are not limited to, the terminal's hardware processing capabilities, communication module type, registration permission level in the master station system, and the bandwidth of the connected communication network. According to the data upload permissions of the terminal under test, the test data is divided into lightweight feature data and complete waveform data. Lightweight feature data refers to the set of key parameters extracted from the original waveform recording that is sufficient to support preliminary fault assessment, such as wavefront arrival time, polarity, peak amplitude, frequency feature vector, or compressed waveform summary information. Complete waveform data refers to a high-sampling-rate original sequence or lossless compressed sequence containing the entire process before and after the fault, used to support refined waveform analysis and secondary verification. It should be understood that although this embodiment lists the specific data types mentioned above, in other embodiments, lightweight feature data can also be any other form of low-dimensional feature representation, as long as its data volume is significantly smaller than that of complete waveform data and can meet the initial localization requirements.

[0039] For these two types of data, the system employs differentiated timing strategies for transmission: According to the first timing requirement, lightweight feature data is uploaded to the first-level master station to trigger initial fault location; according to the second timing requirement, complete waveform data is uploaded to the second-level master station to trigger precise fault location and secondary verification. The response time limit for the first timing requirement is shorter than that for the second timing requirement. For example, the first timing requirement can be set to milliseconds to ensure that the first-level master station can obtain critical information and make rapid regional isolation decisions within a very short time after a fault occurs; while the second timing requirement can be relaxed to seconds or even minutes, allowing reliable transmission of large amounts of data during network congestion or idle periods. This hierarchical timing design perfectly aligns with the two-layer business architecture of rapid response in Zone I and in-depth analysis in Zone IV of the distribution network, ensuring both real-time fault handling and the completeness of data analysis while making rational use of communication resources. In this way, hardware-in-the-loop testing not only verifies the terminal's data generation function, but also verifies its data scheduling and protocol adaptation capabilities under complex business constraints, enhancing the non-obviousness and practical value of the test solution for actual engineering scenarios.

[0040] Example 4 In this embodiment, the specific execution logic for quantitatively evaluating the end-to-end performance of the power distribution network traveling wave fault detection system based on the uploaded data from the terminal under test and the judgment results of the dual-level master station is further detailed in the target test mode. As mentioned above, the target test mode includes an absolute time test mode based on a common time reference or a relative time test mode based on the relative timing of multiple terminals. These two modes are not mutually exclusive fixed configurations, but rather parallel technical solutions that can be flexibly selected or dynamically switched according to the satellite signal quality, communication network conditions, and test objectives at the test site, thereby ensuring that the test method can cover all scenarios from standardized laboratory verification to complex outdoor field operation and maintenance.

[0041] Specifically, in absolute time testing mode, based on the same absolute time reference of the signal generator and the terminal under test (DUT), a hierarchical analysis mechanism linking the first-level master station and the second-level master station is used to simultaneously verify the terminal wavefront identification accuracy, hierarchical upload compliance, fault algorithm matching degree, and dual-zone analysis consistency. This mode is mainly suitable for laboratory network access verification scenarios with high-precision time synchronization capabilities. During execution, the hardware-in-the-loop signal generator, the DUT cluster, and the dual-level master station system are all connected to the same high-precision time source, establishing a unified absolute time coordinate system. While outputting standardized fault waveform signals, the signal generator records the absolute timestamp of the physical initial moment of the traveling wave in the waveform as the truth reference. After acquiring the signal, the DUT generates test data with absolute timestamps based on its own local clock and uploads it hierarchically. The test system compares the wavefront time reported by the terminal, the initial positioning time of the first-level master station, and the fine positioning time of the second-level master station with the truth reference of the signal source at the microsecond level. By cross-checking the absolute times of the signal source, terminal, and master station, not only can the terminal's own wavefront identification error and clock drift error be accurately quantified, but the compliance of the hierarchical data transmission protocol under strict timing constraints can also be verified, as well as the accuracy of algorithm matching by the two-level master station under different fault types. This closed-loop verification mechanism based on absolute truth provides the highest level of accuracy assurance for terminal network access.

[0042] As another parallel implementation method, in the relative time test mode, based on the relative timing of a multi-terminal cluster, the system verifies the consistency of parameters for terminals of the same model and the compatibility of terminals of different brands or types, to evaluate the system's compatibility in multi-terminal mixed access scenarios. This mode is designed for on-site maintenance scenarios such as outdoor substations and remote lines where BeiDou signals are interrupted or a stable absolute time reference cannot be obtained. The advantage of this implementation method is that, in the absence of a unified external clock, it utilizes the differences in the responses of multiple terminals to the same physical event to evaluate system performance.

[0043] To more clearly illustrate the execution details of the relative time test mode, this embodiment further provides the following specific implementation process. First, multiple brands and types of terminals under test are simultaneously connected to the hardware-in-the-loop test circuit to uniformly receive standardized fault waveform signals. For example, an FTU terminal from manufacturer A, a fault indicator from manufacturer B, and a new sensor terminal from manufacturer C can be simultaneously connected to the same test bus, with all terminals experiencing the same traveling wave excitation at the same physical moment. Subsequently, each terminal under test independently completes wavefront identification, quality screening, and graded data uploading, generating its own relative judgment timing sequence and test data. Since there is no absolute time reference, the test system no longer focuses on the absolute accuracy of individual terminal data, but rather on the relative relationship between data.

[0044] Based on this, the testing system performs differential verification in two dimensions. The first dimension compares the waveform timing, quality score, and upload timing deviation of terminals of the same model to assess the stability of the equipment hardware and algorithms. For example, for three terminals of the same model from Manufacturer A, if the dispersion of their identification time for the same fault waveform exceeds a preset threshold, or if the quality score results are inconsistent, it indicates that the hardware consistency or algorithm robustness of this model of terminal has defects. Even if it may barely pass in absolute time mode, there are still hidden dangers in batch deployment. The second dimension quantifies the timing deviation and data interaction compatibility of terminals of different brands or types in mixed access scenarios. In actual distribution networks, there are often differences in the sampling rate, filtering algorithm, and communication protocol stack implementation of equipment from different manufacturers. These differences are difficult to expose when testing a single device, and only cause timing alignment difficulties or data parsing conflicts when mixed networking occurs. By comparing the relative timing deviation and data interaction success rate between heterogeneous terminals, it is possible to effectively identify those isolated devices that meet their respective enterprise standards but cannot be integrated into the multi-vendor ecosystem.

[0045] Finally, based on the differentiated data uploaded from multiple terminals, the stability of the initial and fine positioning results of the two-level master station is verified. In relative time mode, the master station receives asynchronous data streams from a heterogeneous terminal cluster, which places higher demands on the master station's data fusion algorithm. The test system evaluates the master station's algorithm's fault tolerance capability for complex on-site conditions by observing whether the master station can still output stable and convergent positioning results even when the data timing is not completely aligned. It should be understood that although this embodiment uses no time synchronization as a typical application scenario for the relative time test mode, under certain specific test requirements, even with an absolute time reference, it is possible to actively switch to relative time mode to specifically assess the system's compatibility and robustness under extreme conditions. This demonstrates the flexibility and completeness of the technical solution in this application in terms of testing strategy.

[0046] Example 5 This embodiment further details the specific implementation mechanism for a full-link performance quantitative evaluation of a distribution network traveling wave fault detection system. This mechanism not only focuses on whether the final test passes or fails, but also emphasizes locating system performance bottlenecks through refined error tracing. Specifically, the process of a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system includes: distinguishing between terminal wavefront identification errors, quality screening misjudgment errors, hierarchical upload timing offset errors, fault algorithm matching deviations, master station data interaction delay errors, and terminal clock drift errors. These six types of errors cover the entire business chain from signal perception, edge processing, data transmission to master station analysis, and the distinction of each type of error relies on the hierarchical benchmark data obtained in the aforementioned embodiment as a truth reference.

[0047] Specifically, the terminal wavefront identification error is obtained by comparing the wavefront arrival time reported by the terminal under test with the physical initial time of the traveling wave locked in the hierarchical reference data at the microsecond level. This error directly reflects the accuracy of the terminal hardware sampling and the underlying identification algorithm. The quality screening misjudgment error is determined by comparing the valid / invalid wavefront status marked by the terminal with the standard wavefront attributes marked in the hierarchical reference data. If the terminal filters out a valid faulty wavefront in the reference or marks a noisy pseudo wavefront as valid, it is counted as a misjudgment. This indicator characterizes the reliability of the terminal's front-end data quality control capability. The hierarchical upload timing offset error refers to the difference between the actual upload time of lightweight feature data and complete waveform data and the standard response time limit specified by the protocol. It is used to verify whether the terminal's real-time scheduling capability under communication congestion or high load meets the service timeliness requirements of the dual-level master station. Algorithm matching deviation is determined by analyzing whether the fault type, location, and other information output by the master station are consistent with the preset standard fault scenarios in the hierarchical benchmark data. If the initial or fine positioning result deviates from the benchmark true value by more than the allowable range, it indicates that the fault inversion algorithm of the terminal or master station has an adaptation defect under the current working conditions. The master station data interaction delay error focuses on the communication link and master station processing links. It is obtained by recording the difference between the time when the data leaves the terminal and the time when the master station completes the parsing and storage, and subtracting the theoretical transmission time. It is used to evaluate the impact of network channel and master station concurrent processing performance on the overall judgment timeliness. The terminal clock drift error is mainly quantified in the absolute time test mode. It is calculated by continuously monitoring the phase difference change rate between the terminal's local clock and the same source standard time benchmark. This error is a key basis for evaluating the long-term operating stability of the terminal and the cross-device time synchronization capability.

[0048] Based on the accurate differentiation of the aforementioned types of errors, this embodiment further generates a quantitative score for the terminal's overall network access performance by combining preset thresholds in the hierarchical benchmark data with the differentiated errors. These preset thresholds are not a single global threshold, but rather a tiered evaluation scale set for different error types. For example, for wavefront identification errors, the threshold might be set at a sub-microsecond accuracy tolerance; while for tiered upload timing offsets, the threshold might be a millisecond or second-level service tolerance window. The generation logic for the quantitative score can employ methods such as weighted summation, radar chart area integration, or multi-dimensional vector magnitude calculation to map the error performance across the six dimensions into a comparable comprehensive value. This quantitative scoring mechanism avoids the one-sidedness of traditional testing where passing a single indicator equates to overall success, and can more comprehensively characterize the terminal's overall performance level under complex operating conditions.

[0049] Ultimately, the compliance of terminal network access is determined based on a comprehensive performance quantitative score, and differentiated access adaptation suggestions are generated based on the terminal type. This process transforms test results into engineering application strategies. Specifically, when the comprehensive score is higher than the preset access baseline, the terminal is deemed compliant; otherwise, it is deemed non-compliant. For compliant terminals, the system will automatically generate differentiated access configuration suggestions based on their scores and the terminal's own functional positioning. For example, for terminals with excellent comprehensive scores and advanced functions, it is recommended to connect them to the first-level master station and assign them to rapid initial positioning and regional protection linkage tasks; for terminals with qualified scores but at a critical state in certain dimensions, or existing terminals with relatively simple functions, it is recommended to connect them to the second-level master station, serving only as a data supplementary source for fine-tuning verification or an offline analysis node, without participating in real-time control decisions. This differentiated adaptation suggestion based on quantitative evaluation results not only improves the overall operational reliability of the distribution network traveling wave fault detection system, but also provides maintenance personnel with a scientific basis for equipment deployment and hierarchical management, fully demonstrating the practical value and guiding significance of the technical solution in actual engineering applications.

[0050] Example 6 This embodiment provides a hardware-in-the-loop testing device for a distribution network traveling wave fault detection system. This device is applicable to the hardware-in-the-loop testing method for the distribution network traveling wave fault detection system described in any of the preceding embodiments 1 to 5. As a virtual device embodiment, it characterizes the logical execution body of the method flow through the division of functional modules. Each module can be implemented as a program unit running on software code, firmware logic, application-specific integrated circuits, or a general-purpose processor, or it can be a distributed processing node composed of multiple physical devices working together. Specifically, the device includes a data acquisition module, a data upload module, and a performance evaluation module. The modules interact and logically link with each other through an internal bus, shared memory, or network interface, together forming a complete test execution closed loop.

[0051] The data acquisition module is used to acquire pre-constructed hierarchical benchmark data, which includes first-level benchmark data adapted to initial positioning requirements and second-level benchmark data adapted to fine positioning requirements. In specific implementation, this module can manifest as a database access interface or file reading service, configured to load standardized benchmark datasets from local storage media or a remote cloud server during the test startup phase. It should be understood that although this embodiment describes it as an acquisition action, in other embodiments, this module can also have the ability to generate benchmark data in real time. For example, it can integrate the multi-constraint joint recognition algorithm engine described in Embodiment 2 above, which can dynamically calculate and cache hierarchical benchmark data based on the input original simulation waveform, thereby adapting to two different application scenarios: offline testing and online calibration. Regardless of the implementation method, the first-level and second-level benchmark data output by this module serve as a unified truth reference for subsequent evaluation stages, ensuring the consistency between the test benchmark and the methodology.

[0052] The data upload module, based on the access characteristics of the terminal under test (DUT), uploads differentiated test data to the first-level and second-level master stations. The test data is generated hierarchically by the DUT based on the collected traveling wave signals. Logically, this module acts as the communication scheduling hub between the DUT and the two master stations. In actual deployment, it can be directly embedded into the embedded system of the DUT as part of the terminal firmware to control data encapsulation and transmission; alternatively, in a pure software simulation testing environment, it can be an independent protocol stack simulator to reproduce the data interaction behavior of real terminals under different access permissions. Specifically, this module maintains a terminal type mapping table and a hierarchical transmission strategy library, which can automatically match the corresponding upload timing and data format requirements according to the identity of the currently accessing terminal, thereby accurately triggering the initial positioning process of the first-level master station and the fine positioning process of the second-level master station. This design allows the device to not only verify the functional correctness of the terminal itself, but also verify the adaptability of the entire data transmission link to the hierarchical business architecture of the power distribution network.

[0053] The performance evaluation module is used to perform a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system based on the uploaded data from the terminal under test and the judgment results from the two-level master station under the target test mode. This module is the core analysis unit of the device, integrating multi-dimensional error calculation algorithms and a comprehensive scoring model. During execution, the module receives the baseline truth value from the data acquisition module, the terminal measured data from the data upload module, and the feedback results from the two-level master station, and performs comparative analysis under the selected target test mode. The target test modes include an absolute time test mode based on the same time reference or a relative time test mode based on the relative timing of multiple terminals. This means that the performance evaluation module has the ability to adaptively switch modes: when an external high-precision clock source is detected to be valid, the absolute time verification logic is automatically enabled to evaluate the accuracy index; when in an environment without time synchronization, it automatically switches to the relative timing benchmarking logic to evaluate the compatibility and consistency index. Through this flexible architecture design, the device can cover two completely different testing needs—laboratory verification and on-site operation and maintenance—on a unified hardware and software platform. This avoids the waste of resources in developing multiple independent testing tools for different scenarios and significantly improves the versatility and engineering practical value of the testing system.

[0054] Example 7 This embodiment provides a hardware-in-the-loop testing system for traveling wave fault detection in distribution networks. This system serves as the physical carrier of the aforementioned method embodiments and aims to provide complete testing environment support for the network access verification and on-site operation and maintenance of traveling wave fault detection equipment in distribution networks. Specifically, the testing system includes a hardware-in-the-loop signal generator, a cluster of terminals under test, a first-level master station, and a second-level master station. These components are interconnected through electrical lines and communication networks, forming a closed-loop simulation testing platform capable of realistically reproducing the electromagnetic transient processes and data interaction flow during distribution network faults.

[0055] The hardware-in-the-loop signal generator is configured to output standardized fault waveform signals. In its implementation, this device typically consists of a real-time simulator, a power amplifier, and a high-precision time synchronization module. The real-time simulator runs a pre-built electromagnetic transient model of the power distribution network, capable of calculating instantaneous voltage and current values ​​under specific fault types, fault distances, and transition resistances in real time according to test requirements. These values ​​are then converted into strong electrical signals recognizable by the terminal under test (DUT) via a digital-to-analog converter board and the power amplifier. Simultaneously, the device has a time synchronization interface with BeiDou or GPS satellites, enabling it to mark precise absolute timestamps while outputting signals, thus providing a unified physical time reference for the aforementioned absolute time test modes. It should be understood that although this embodiment describes a typical configuration including a power amplifier, in scenarios where only the terminal communication protocol or data processing logic needs to be tested, the device can also output only weak electrical signals or directly inject them into the DUT in the form of digital messages, as long as it can provide standardized fault excitation.

[0056] The terminal under test (DUT) cluster is configured to acquire traveling wave signals and generate test data in a hierarchical manner. Here, "cluster" does not refer to a single type of device collection, but rather a mixed access environment encompassing various heterogeneous devices such as modifiable FTU terminals, existing non-modifiable FTU terminals, and fault indication terminals. Physically, these terminals are connected in parallel to the output bus of a hardware-in-the-loop signal generator, simulating the multi-point deployment on an actual line. Each terminal independently completes signal acquisition, wavefront identification, and data encapsulation according to its own hardware capabilities and preset access permissions, and sends lightweight feature data and complete waveform data to the corresponding upper-level system according to the hierarchical transmission strategy described in the aforementioned embodiments. This clustered design enables the test system to simultaneously verify the response consistency and interoperability of multiple manufacturers and models of equipment under the same fault event, effectively supporting the evaluation of system compatibility under relative time testing mode.

[0057] The first and second level master stations are configured to receive test data and perform hierarchical fault analysis. Physically, these two master stations are typically deployed in different security zones to match the actual network security protection requirements of the distribution network. The first level master station is generally located in the production control zone (Zone I), connected to the terminal under test (DUT) with real-time interactive capabilities via a low-latency secure access channel, focusing on receiving lightweight characteristic data and performing millisecond-level initial fault location. The second level master station is located in the management information zone (Zone IV), acquiring complete waveform data and redundant supplementary data through forward and reverse isolation devices or a dedicated data exchange platform, responsible for performing refined waveform analysis and secondary verification. The two master stations exchange results and cross-verify through standardized data interfaces, thus physically reproducing the business architecture of dual-zone collaborative analysis of the distribution network. It should be understood that in some simplified testing scenarios, the first and second level master stations can also be logically integrated into the same server or cloud platform, as long as their software functional modules maintain a clear hierarchical division and data flow isolation, which is considered to comply with the technical solution of this application.

[0058] More importantly, the system is configured to execute the hardware-in-the-loop testing method for the distribution network traveling wave fault detection system described in any of the preceding embodiments 1 to 6. This means that the aforementioned physical components are not isolated general-purpose devices, but are specifically adapted into an organic whole through pre-built software programs, firmware logic, or configuration files. For example, the hardware-in-the-loop signal generator loads a fault scenario script corresponding to the hierarchical reference data; the terminal under test is programmed with an edge computing algorithm that supports quality scoring and hierarchical uploading; and the dual-level master station deploys an error tracing and quantitative evaluation module that matches the testing method. It is this deep coupling and functional binding of hardware and software that enables the testing system to automatically execute the entire process from reference acquisition and differentiated data interaction to end-to-end performance evaluation, thereby providing system integrators or equipment manufacturers with a complete product-level solution that cannot be circumvented by simply replacing a single component. Through the synergy of the above physical architecture and functional configuration, the testing system provided in this embodiment not only verifies the performance indicators of a single terminal, but also verifies the end-to-end reliability and engineering practicality of the entire distribution network traveling wave fault detection system under complex operating conditions.

[0059] Example 8 To more clearly illustrate the application value of the technical solution of this application in practical engineering, this embodiment combines the technical features of the foregoing embodiments and provides a comprehensive demonstration using two typical business scenarios: laboratory network access verification and outdoor field operation and maintenance. It should be understood that the following scenarios are merely illustrative application examples and not limitations on the scope of protection of this application. In practical applications, testers can flexibly combine or adjust the execution parameters of each technical feature according to specific needs.

[0060] In the first application scenario, standardized laboratory network access verification was conducted for FTU terminals with real-time interactive capabilities. This scenario adopted an absolute time testing mode based on a common time reference. Specifically, on a 10kV distribution network hardware-in-the-loop simulation platform, testers simultaneously connected a high-precision signal generator, the FTU terminal under test, and a reference terminal on the opposite side to the BeiDou time service system to establish a unified microsecond-level absolute time coordinate system. The test system pre-acquired first-level reference data adapted to initial positioning requirements and second-level reference data adapted to fine positioning requirements. These reference data were generated based on a multi-constraint joint identification mechanism to lock the physical initial moment of the traveling wave, ensuring the physical accuracy of the test truth. Subsequently, the signal generator output a standardized traveling wave signal simulating a phase-to-phase short-circuit fault, for example, setting the fault distance to 3km and the transition resistance to 10Ω, and simultaneously recorded the absolute timestamp of the physical initial moment of the traveling wave signal as the truth reference.

[0061] After acquiring the traveling wave signal, the FTU terminal under test first performs a pre-quality control step. Based on the signal-to-noise ratio, amplitude ratio, and rise edge steepness, it scores the acquired waveform for quality, automatically filtering out invalid wavefronts caused by simulation noise or equipment background noise, retaining only valid wavefronts that pass verification as subsequent data sources. Next, the terminal uploads test data in a differentiated manner according to its access characteristics and upload permissions: Following the first timing requirement, lightweight characteristic data containing wavefront timestamps, amplitude, polarity, and other parameters is uploaded to the first-level master station to trigger initial fault location; simultaneously, following the second timing requirement, complete waveform data is uploaded to the second-level master station to support precise fault location and secondary verification. Throughout this process, the test system monitors the data transmission link in real time, quantifying and classifying the upload timing offset error and the master station data interaction delay error.

[0062] After receiving the lightweight feature data, the first-level master station quickly completes the fault area assessment; the second-level master station, after acquiring complete waveform data, performs refined analysis in conjunction with the fault feature template. The test system cross-compares the terminal's reporting time and the assessment time of the two-level master stations with the true value benchmark of the signal source, distinguishing multi-dimensional indicators such as terminal wavefront identification error, fault algorithm matching deviation, and terminal clock drift error. Finally, based on the preset thresholds in the hierarchical benchmark data, a quantitative score for the FTU terminal's comprehensive network access performance is generated, and its compliance with the Zone I access standard is determined accordingly. This high-precision closed-loop verification based on absolute time not only assesses the terminal's own hardware and algorithm performance but also verifies the timeliness and accuracy of its collaborative work with the two-level master stations, providing authoritative performance endorsement for the terminal's network access.

[0063] In the second application scenario, multi-terminal hybrid operation and maintenance testing is conducted in outdoor substations under conditions of BeiDou signal interruption or weak signal. This scenario employs a relative time testing mode based on the relative timing of multiple terminals. Unlike laboratory scenarios, a unified absolute time reference cannot be obtained here, and the testing focus shifts to evaluating the system's compatibility and robustness under non-ideal operating conditions. Specifically, testers simultaneously connect terminals under test from different manufacturers and of different types to the hardware-in-the-loop test loop, uniformly receiving standardized fault waveform signals that include the characteristics of noise measured on-site.

[0064] Without an absolute time reference, each terminal under test independently completes wavefront identification, quality screening, and graded data upload, generating its own relative judgment timing and test data. The testing system no longer focuses on the absolute accuracy of individual data points, but rather on the relative relationships between data: on the one hand, it compares the wavefront timing dispersion, quality score consistency, and upload timing deviation among terminals of the same model to assess hardware consistency and algorithm stability during mass production; on the other hand, it quantifies the timing deviation and data interaction compatibility of terminals of different brands or types in mixed access scenarios, identifying isolated devices that, while conforming to their respective enterprise standards, face interoperability barriers in multi-vendor ecosystems. Furthermore, based on the differentiated data uploaded by multiple terminals, the testing system verifies the stability of the initial and fine positioning results of the dual-level master station under conditions of incomplete data timing alignment, and evaluates the master station algorithm's fault tolerance capability for complex heterogeneous data in the field.

[0065] Through this relative timing benchmarking mechanism, even under extreme conditions lacking an external high-precision clock source, a comprehensive evaluation of the overall performance of the power distribution network traveling wave fault detection system can still be achieved. This scenario fully demonstrates the adaptability of the proposed technical solution to complex and abnormal on-site operating conditions, solving the technical problem that traditional testing methods over-rely on ideal timing conditions and cannot support standardized operation and maintenance of multi-brand terminals across the entire domain. It should be understood that the above two scenarios do not exist in isolation. In actual engineering practice, testers can also proactively switch to relative time mode when an absolute time reference is available, specifically for stress testing or compatibility verification, thereby achieving flexible configuration and full coverage of testing strategies.

[0066] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations, substitutions, or improvements that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application, based on the aforementioned technical concepts regarding the hierarchical benchmark construction, differentiated data interaction, dual-mode performance evaluation, and error tracing of the hardware-in-the-loop testing method for traveling wave fault detection systems in distribution networks, such as adjustments to specific parameters in the multi-constraint joint identification mechanism, equivalent transformations of the hierarchical upload timing strategy, or combined applications of absolute time and relative time test modes, should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A hardware-in-the-loop testing method for a traveling wave fault detection system in a power distribution network, characterized in that, include: Acquire pre-constructed hierarchical reference data, which includes first-level reference data adapted to initial positioning requirements and second-level reference data adapted to fine positioning requirements; Based on the access characteristics of the terminal under test, control the terminal under test to upload differentiated test data to the first-level master station and the second-level master station in a hierarchical manner; The test data is generated by the terminal under test in stages based on the collected traveling wave signals; In the target test mode, based on the hierarchical benchmark data, combined with the uploaded data of the terminal under test and the judgment results of the dual-level master station, the full-link performance quantitative evaluation of the distribution network traveling wave fault detection system is carried out. The target test modes include an absolute time test mode based on a common time reference or a relative time test mode based on the relative timing of multiple terminals.

2. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 1, characterized in that, The hierarchical reference data is generated based on a multi-constraint joint identification mechanism that locks the physical initial moment of the traveling wave. The multi-constraint joint identification mechanism includes: Temporal gradient mutation constraint is used to screen mutation candidate points based on steady-state noise benchmark; Waveform kurtosis disturbance rejection constraint is used to filter out pseudo-abrupt points based on local waveform kurtosis; Baseline step offset constraint is used to verify persistent step abrupt changes based on the baseline difference before and after the abrupt change. Based on the mutation points verified by the multi-constraint joint identification mechanism, the physical initial moment of the traveling wave is locked, and the first-level reference data and the second-level reference data are generated accordingly.

3. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 2, characterized in that, The execution process of the multi-constraint joint identification mechanism includes: The acquired raw traveling wave time sequence is preprocessed, and the steady-state waveform data before the fault is extracted to calculate the mean steady-state baseline and establish the fault-free reference level. Based on the system sampling interval, the first-order instantaneous gradient of the waveform is calculated point by point to quantify the severity of waveform changes at each time-series node; The maximum gradient value within the steady-state interval before the fault is used as the benchmark for system noise fluctuation. An adaptive mutation threshold is constructed. When the gradient of a sampling point is greater than or equal to the adaptive mutation threshold, the point is determined to be a candidate mutation point. The ratio of the fourth central moment to the variance of the instantaneous signal amplitude after removing the mean is calculated for the waveform data within the short-time window near the candidate mutation point to obtain the local waveform kurtosis. Only candidate points whose local waveform kurtosis is greater than a preset kurtosis threshold are retained. The mean of the short-time steady-state waveform after the mutation occurs is statistically analyzed, and the baseline difference before and after the mutation is calculated. When the baseline difference before and after the mutation is greater than or equal to a preset offset threshold, the mutation is determined to be a continuous step mutation induced by a fault. Traverse all the mutation points that pass the verification, and select the earliest mutation point in the time dimension as the physical initial time of the traveling wave.

4. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 1, characterized in that, The method of controlling the terminal under test to upload differentiated test data to the first-level master station and the second-level master station based on the access characteristics of the terminal under test includes: Based on the data upload permissions of the terminal under test, the test data is divided into lightweight feature data and complete waveform data; According to the first timing requirement, the lightweight feature data is uploaded to the first-level master station to trigger the first-level master station to perform initial fault location. According to the second timing requirements, the complete waveform data is uploaded to the second-level master station to trigger the second-level master station to perform fault fine location and secondary verification; Wherein, the response time limit of the first timing requirement is shorter than the response time limit of the second timing requirement.

5. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 4, characterized in that, Before uploading differentiated test data from the terminal under test to the first-level master station and the second-level master station, the method further includes: The quality of the waveforms acquired by the terminal under test is scored based on the signal-to-noise ratio, amplitude ratio, and rise edge steepness. Invalid wavefronts are filtered based on the quality score, and the compliance of the caching and uploading of valid wavefronts is verified. The data corresponding to the valid wavefronts that pass the verification are used as the data source for the lightweight feature data and the complete waveform data.

6. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 1, characterized in that, In the target test mode, based on the hierarchical benchmark data, combined with the uploaded data of the terminal under test and the judgment results of the two-level master station, a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system is performed, including: In the absolute time test mode, based on the same absolute time reference of the signal generator and the terminal under test, the hierarchical analysis mechanism of the first-level master station and the second-level master station is linked to simultaneously verify the terminal wavefront identification accuracy, hierarchical uploading compliance, fault algorithm matching degree and dual-zone analysis consistency. or, In the relative time test mode, based on the relative timing of the multi-terminal cluster, the system verifies the consistency of parameters of terminals of the same model and the compatibility of terminals of different brands or types, in order to evaluate the system compatibility in multi-terminal mixed access scenarios.

7. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 6, characterized in that, In the relative time test mode, based on the relative timing of multiple terminal clusters, the consistency verification of parameters for terminals of the same model and the compatibility verification of terminals of different brands or types are performed, including: Simultaneously connect multiple brands and types of terminals under test to the hardware-in-the-loop test loop and uniformly receive standardized fault waveform signals; Each terminal under test independently completes wavefront identification, quality screening, and graded data upload, generating its own relative judgment timing and test data; By comparing the wavefront timing, quality score, and upload timing deviation of terminals of the same model, the stability of the equipment hardware and algorithm is assessed. Quantify the timing deviation and data interaction compatibility of terminals of different brands or types in mixed access scenarios; Based on differentiated data uploaded from multiple terminals, the stability of the initial and fine positioning results of the two-level master station is verified.

8. The hardware-in-the-loop testing method for the power distribution network traveling wave fault detection system according to claim 1, characterized in that, The full-link performance quantitative evaluation of the power distribution network traveling wave fault detection system also includes: Distinguish between terminal wavefront identification error, quality screening misjudgment error, hierarchical upload timing offset error, fault algorithm matching deviation, main station data interaction delay error, and terminal clock drift error; Based on the preset threshold in the hierarchical benchmark data, and combined with the identified types of errors, a quantitative score for the overall performance of the terminal's network access is generated. The compliance of the terminal's network access is determined based on the comprehensive performance quantitative score, and differentiated access adaptation suggestions are generated in combination with the terminal type.

9. A hardware-in-the-loop testing device for a distribution network traveling wave fault detection system, applicable to the hardware-in-the-loop testing method for a distribution network traveling wave fault detection system as described in any one of claims 1 to 8, characterized in that, include: The data acquisition module is used to acquire pre-constructed hierarchical reference data, which includes first-level reference data adapted to initial positioning requirements and second-level reference data adapted to fine positioning requirements. The data upload module is used to upload test data to the first-level master station and the second-level master station in a differentiated manner based on the access characteristics of the terminal under test. The test data is generated by the terminal under test in stages based on the collected traveling wave signal. The performance evaluation module is used to perform a full-link performance quantitative evaluation of the distribution network traveling wave fault detection system based on the uploaded data of the terminal under test and the judgment results of the two-level master station in the target test mode. The target test modes include an absolute time test mode based on a common time reference or a relative time test mode based on the relative timing of multiple terminals.

10. A hardware-in-the-loop test system for traveling wave fault detection in power distribution networks, characterized in that, include: Hardware-in-the-loop signal generator, terminal under test cluster, first-level master station and second-level master station; The hardware-in-the-loop signal generator is configured to output a standardized fault waveform signal; The terminal cluster under test is configured to collect traveling wave signals and generate test data in stages; The first-level master station and the second-level master station are configured to receive the test data and perform hierarchical fault analysis. The system is configured to perform a hardware-in-the-loop test method for a distribution network traveling wave fault detection system as described in any one of claims 1 to 8.

Images